Version bump

Fixes #20

.gitignore

@@ -15,3 +15,4 @@ bundle*.css
 bundle.js
 *.a
 vgcore.*
+build/

.gitmodules

@@ -4,24 +4,42 @@
 [submodule "cJSON"]
 	path = cJSON
 	url = https://github.com/DaveGamble/cJSON
-[submodule "lib/mupdf"]
-	path = lib/mupdf
-	url = git://git.ghostscript.com/mupdf.git
-[submodule "lib/onion"]
-	path = lib/onion
-	url = https://github.com/davidmoreno/onion
-[submodule "lib/ffmpeg"]
-	path = lib/ffmpeg
-	url = https://git.ffmpeg.org/ffmpeg.git
 [submodule "lmdb"]
 	path = lmdb
 	url = https://github.com/LMDB/lmdb
 [submodule "utf8.h"]
 	path = utf8.h
 	url = https://github.com/sheredom/utf8.h
-[submodule "lib/openjpeg"]
-	path = lib/openjpeg
-	url = https://github.com/uclouvain/openjpeg
+[submodule "lib/bzip2-1.0.6"]
+	path = lib/bzip2-1.0.6
+	url = https://github.com/enthought/bzip2-1.0.6
+[submodule "lib/libmagic"]
+	path = lib/libmagic
+	url = https://github.com/threatstack/libmagic
 [submodule "lib/harfbuzz"]
 	path = lib/harfbuzz
 	url = https://github.com/harfbuzz/harfbuzz
+[submodule "lib/openjpeg"]
+	path = lib/openjpeg
+	url = https://github.com/uclouvain/openjpeg
+[submodule "lib/ffmpeg"]
+	path = lib/ffmpeg
+	url = https://git.ffmpeg.org/ffmpeg.git
+[submodule "lib/onion"]
+	path = lib/onion
+	url = https://github.com/davidmoreno/onion
+[submodule "lib/mupdf"]
+	path = lib/mupdf
+	url = git://git.ghostscript.com/mupdf.git
+[submodule "lib/tesseract"]
+	path = lib/tesseract
+	url = https://github.com/tesseract-ocr/tesseract
+[submodule "lib/leptonica"]
+	path = lib/leptonica
+	url = https://github.com/danbloomberg/leptonica
+[submodule "lib/libtiff"]
+	path = lib/libtiff
+	url = https://gitlab.com/libtiff/libtiff
+[submodule "lib/libpng"]
+	path = lib/libpng
+	url = https://github.com/glennrp/libpng

.teamcity/settings.kts

@@ -0,0 +1,69 @@
+import jetbrains.buildServer.configs.kotlin.v2019_2.*
+import jetbrains.buildServer.configs.kotlin.v2019_2.buildSteps.ExecBuildStep
+import jetbrains.buildServer.configs.kotlin.v2019_2.buildSteps.exec
+import jetbrains.buildServer.configs.kotlin.v2019_2.triggers.vcs
+import jetbrains.buildServer.configs.kotlin.v2019_2.vcs.GitVcsRoot
+
+/*
+The settings script is an entry point for defining a TeamCity
+project hierarchy. The script should contain a single call to the
+project() function with a Project instance or an init function as
+an argument.
+
+VcsRoots, BuildTypes, Templates, and subprojects can be
+registered inside the project using the vcsRoot(), buildType(),
+template(), and subProject() methods respectively.
+
+To debug settings scripts in command-line, run the
+
+    mvnDebug org.jetbrains.teamcity:teamcity-configs-maven-plugin:generate
+
+command and attach your debugger to the port 8000.
+
+To debug in IntelliJ Idea, open the 'Maven Projects' tool window (View
+-> Tool Windows -> Maven Projects), find the generate task node
+(Plugins -> teamcity-configs -> teamcity-configs:generate), the
+'Debug' option is available in the context menu for the task.
+*/
+
+version = "2019.2"
+
+project {
+
+    vcsRoot(HttpsGithubComSimon987sist2refsHeadsMaster)
+
+    buildType(Build)
+}
+
+object Build : BuildType({
+    name = "Build"
+
+    artifactRules = """
+        sist2
+        sist2_scan
+    """.trimIndent()
+
+    vcs {
+        root(HttpsGithubComSimon987sist2refsHeadsMaster)
+    }
+
+    steps {
+        exec {
+            name = "Build"
+            path = "./ci/build.sh"
+            dockerImage = "simon987/general_ci"
+            dockerImagePlatform = ExecBuildStep.ImagePlatform.Linux
+            dockerPull = true
+        }
+    }
+
+    triggers {
+        vcs {
+        }
+    }
+})
+
+object HttpsGithubComSimon987sist2refsHeadsMaster : GitVcsRoot({
+    name = "https://github.com/simon987/sist2#refs/heads/master"
+    url = "https://github.com/simon987/sist2"
+})

CMakeLists.txt

@@ -1,94 +1,56 @@
 cmake_minimum_required(VERSION 3.7)
 set(CMAKE_C_STANDARD 11)
 
-option(WITH_SIST2 "Build main executable" ON)
-option(WITH_SIST2_SCAN "Build scan executable" ON)
-
 project(sist2 C)
 list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_LIST_DIR}/CMakeModules")
 
-if (WITH_SIST2)
-    add_executable(
-            sist2
-            src/main.c
-            src/sist.h
-            src/io/walk.h src/io/walk.c
-            src/parsing/media.h src/parsing/media.c
-            src/parsing/pdf.h src/parsing/pdf.c
-            src/io/store.h src/io/store.c
-            src/tpool.h src/tpool.c
-            src/parsing/parse.h src/parsing/parse.c
-            src/io/serialize.h src/io/serialize.c
-            src/parsing/mime.h src/parsing/mime.c src/parsing/mime_generated.c
-            src/parsing/text.h src/parsing/text.c
-            src/index/web.c src/index/web.h
-            src/web/serve.c src/web/serve.h
-            src/index/elastic.c src/index/elastic.h
-            src/util.c src/util.h
-            src/ctx.h src/types.h src/parsing/font.c src/parsing/font.h
+add_executable(
+        sist2
+        src/main.c
+        src/sist.h
+        src/io/walk.h src/io/walk.c
+        src/parsing/media.h src/parsing/media.c
+        src/parsing/pdf.h src/parsing/pdf.c
+        src/io/store.h src/io/store.c
+        src/tpool.h src/tpool.c
+        src/parsing/parse.h src/parsing/parse.c
+        src/io/serialize.h src/io/serialize.c
+        src/parsing/mime.h src/parsing/mime.c src/parsing/mime_generated.c
+        src/parsing/text.h src/parsing/text.c
+        src/index/web.c src/index/web.h
+        src/web/serve.c src/web/serve.h
+        src/web/auth_basic.h src/web/auth_basic.c
+        src/index/elastic.c src/index/elastic.h
+        src/util.c src/util.h
+        src/ctx.h src/types.h src/parsing/font.c src/parsing/font.h
+        src/parsing/arc.c src/parsing/arc.h
+        src/parsing/doc.c src/parsing/doc.h
+        src/log.c src/log.h
 
-            # argparse
-            argparse/argparse.h argparse/argparse.c
+        # argparse
+        argparse/argparse.h argparse/argparse.c
 
-            # cJSON
-            cJSON/cJSON.h cJSON/cJSON.c
+        # cJSON
+        cJSON/cJSON.h cJSON/cJSON.c
 
-            # LMDB
-            lmdb/libraries/liblmdb/lmdb.h lmdb/libraries/liblmdb/mdb.c
-            lmdb/libraries/liblmdb/midl.h lmdb/libraries/liblmdb/midl.c
-            src/cli.c src/cli.h
+        # LMDB
+        lmdb/libraries/liblmdb/lmdb.h lmdb/libraries/liblmdb/mdb.c
+        lmdb/libraries/liblmdb/midl.h lmdb/libraries/liblmdb/midl.c
+        src/cli.c src/cli.h
 
-            # utf8.h
-            utf8.h/utf8.h
-    )
-endif ()
-
-if (WITH_SIST2_SCAN)
-    add_executable(
-            sist2_scan
-            src/main.c
-            src/sist.h
-            src/io/walk.h src/io/walk.c
-            src/parsing/media.h src/parsing/media.c
-            src/parsing/pdf.h src/parsing/pdf.c
-            src/io/store.h src/io/store.c
-            src/tpool.h src/tpool.c
-            src/parsing/parse.h src/parsing/parse.c
-            src/io/serialize.h src/io/serialize.c
-            src/parsing/mime.h src/parsing/mime.c src/parsing/mime_generated.c
-            src/parsing/text.h src/parsing/text.c
-            src/util.c src/util.h
-            src/ctx.h src/types.h src/parsing/font.c src/parsing/font.h
-
-            # argparse
-            argparse/argparse.h argparse/argparse.c
-
-            # cJSON
-            cJSON/cJSON.h cJSON/cJSON.c
-
-            # LMDB
-            lmdb/libraries/liblmdb/lmdb.h lmdb/libraries/liblmdb/mdb.c
-            lmdb/libraries/liblmdb/midl.h lmdb/libraries/liblmdb/midl.c
-            src/cli.c src/cli.h
-
-            # utf8.h
-            utf8.h/utf8.h
-    )
-endif ()
+        # utf8.h
+        utf8.h/utf8.h
+)
 
 find_package(PkgConfig REQUIRED)
 set(ENV{PKG_CONFIG_PATH} "$ENV{PKG_CONFIG_PATH}:/usr/local/lib/pkgconfig/")
 
-find_package(LibMagic REQUIRED)
-find_package(FFmpeg REQUIRED)
-find_package(OpenSSL REQUIRED)
 find_package(Freetype REQUIRED)
 
 pkg_check_modules(GLIB REQUIRED glib-2.0)
 pkg_check_modules(GOBJECT REQUIRED gobject-2.0)
 pkg_check_modules(UUID REQUIRED uuid)
 
-add_definitions(${LIBMAGIC_CFLAGS_OTHER})
 add_definitions(${UUID_CFLAGS_OTHER})
 add_definitions(${GLIB_CFLAGS_OTHER})
 add_definitions(${GOBJECT_CFLAGS_OTHER})
@@ -98,145 +60,87 @@ list(REMOVE_ITEM GLIB_LIBRARIES pcre)
 list(REMOVE_ITEM GOBJECT_LIBRARIES pcre)
 list(REMOVE_ITEM UUID_LIBRARIES pcre)
 
-if (WITH_SIST2)
-    target_include_directories(
-            sist2 PUBLIC
-            ${LIBMAGIC_INCLUDE_DIRS}
-            ${GOBJECT_INCLUDE_DIRS}
-            ${OPENSSL_INCLUDE_DIR}
-            ${FFMPEG_INCLUDE_DIRS}
-            ${GLIB_INCLUDE_DIRS}
-            ${FREETYPE_INCLUDE_DIRS}
-            ${UUID_INCLUDE_DIRS}
-            ${PROJECT_SOURCE_DIR}/
-            ${PROJECT_SOURCE_DIR}/lmdb/libraries/liblmdb/
-            ${PROJECT_SOURCE_DIR}/lib/onion/src/
-            ${PROJECT_SOURCE_DIR}/lib/mupdf/include/
-    )
-    target_link_directories(
-            sist2 PUBLIC
-            ${UUID_LIBRARY_DIRS}
-            ${FFMPEG_LIBRARY_DIRS}
-    )
+target_include_directories(
+        sist2 PUBLIC
+        ${GOBJECT_INCLUDE_DIRS}
+        ${GLIB_INCLUDE_DIRS}
+        ${PROJECT_SOURCE_DIR}/lib/ffmpeg/
+        ${FREETYPE_INCLUDE_DIRS}
+        ${UUID_INCLUDE_DIRS}
+        ${PROJECT_SOURCE_DIR}/
+        ${PROJECT_SOURCE_DIR}/lmdb/libraries/liblmdb/
+        ${PROJECT_SOURCE_DIR}/lib/onion/src/
+        ${PROJECT_SOURCE_DIR}/lib/mupdf/include/
+        ${PROJECT_SOURCE_DIR}/include/
+        /usr/include/libxml2/
+        ${PROJECT_SOURCE_DIR}/lib/tesseract/include/
+)
+target_link_directories(
+        sist2 PUBLIC
+        ${UUID_LIBRARY_DIRS}
+)
 
+target_compile_options(sist2
+        PRIVATE
+        -Ofast
+#        -march=native
+        -fPIC
+        -fno-stack-protector
+        -fomit-frame-pointer
+        )
 
-    target_compile_options(sist2
-            PRIVATE
-            -Ofast
-#                    -march=native
-            -fno-stack-protector
-            -fomit-frame-pointer
-            )
+TARGET_LINK_LIBRARIES(
+        sist2
 
-    TARGET_LINK_LIBRARIES(
-            sist2
+        ${GLIB_LIBRARIES}
+        ${GOBJECT_LIBRARIES}
+        ${UUID_LIBRARIES}
 
-            ${GLIB_LIBRARIES}
-            ${GOBJECT_LIBRARIES}
-            ${UUID_LIBRARIES}
+        # ffmpeg
+        ${PROJECT_SOURCE_DIR}/lib/libavcodec.a
+        ${PROJECT_SOURCE_DIR}/lib/libavformat.a
+        ${PROJECT_SOURCE_DIR}/lib/libavutil.a
+        ${PROJECT_SOURCE_DIR}/lib/libswscale.a
+        ${PROJECT_SOURCE_DIR}/lib/libswresample.a
 
-            # ffmpeg
-            ${PROJECT_SOURCE_DIR}/lib/libavcodec.a
-            ${PROJECT_SOURCE_DIR}/lib/libavformat.a
-            ${PROJECT_SOURCE_DIR}/lib/libavutil.a
-            ${PROJECT_SOURCE_DIR}/lib/libswscale.a
-            ${PROJECT_SOURCE_DIR}/lib/libswresample.a
-            #        ${FFMPEG_LIBRARIES}
-            #        swscale
+        # mupdf
+        ${PROJECT_SOURCE_DIR}/lib/libmupdf.a
+        ${PROJECT_SOURCE_DIR}/lib/libmupdf-third.a
 
-            # mupdf
-            ${PROJECT_SOURCE_DIR}/lib/libmupdf.a
-            ${PROJECT_SOURCE_DIR}/lib/libmupdf-third.a
+        # onion
+        ${PROJECT_SOURCE_DIR}/lib/libonion_static.a
 
-            # onion
-            ${PROJECT_SOURCE_DIR}/lib/libonion_static.a
+        pthread
 
-            pthread
-            curl
-            m
-            bz2
-            magic
-            harfbuzz
-            openjp2
-            freetype
-    )
+        m
+        bz2
+        ${PROJECT_SOURCE_DIR}/lib/libmagic.a
+        ${PROJECT_SOURCE_DIR}/lib/libharfbuzz.a
+        ${PROJECT_SOURCE_DIR}/lib/libopenjp2.a
+        freetype
+        archive
 
-endif ()
+        xml2
+        ${PROJECT_SOURCE_DIR}/lib/libopc/libmce.a
+        ${PROJECT_SOURCE_DIR}/lib/libopc/libopc.a
+        ${PROJECT_SOURCE_DIR}/lib/libopc/libplib.a
 
-if (WITH_SIST2_SCAN)
+        ${PROJECT_SOURCE_DIR}/lib/libtesseract.a
+        ${PROJECT_SOURCE_DIR}/lib/liblept.a
+        ${PROJECT_SOURCE_DIR}/lib/libtiff.a
+        ${PROJECT_SOURCE_DIR}/lib/libpng16.a
+        stdc++
 
-    set_target_properties(
-            sist2_scan
-            PROPERTIES COMPILE_DEFINITIONS SIST_SCAN_ONLY
-    )
-    set_target_properties(
-            sist2_scan
-            PROPERTIES
-            COMPILE_DEFINITIONS SIST_SCAN_ONLY
-            LINK_FLAGS -static
-    )
-    target_include_directories(
-            sist2_scan PUBLIC
-            ${LIBMAGIC_INCLUDE_DIRS}
-            ${GOBJECT_INCLUDE_DIRS}
-            ${OPENSSL_INCLUDE_DIR}
-            ${FFMPEG_INCLUDE_DIRS}
-            ${GLIB_INCLUDE_DIRS}
-            ${UUID_INCLUDE_DIRS}
-            ${FREETYPE_INCLUDE_DIRS}
-            ${PROJECT_SOURCE_DIR}/
-            ${PROJECT_SOURCE_DIR}/lmdb/libraries/liblmdb/
-            ${PROJECT_SOURCE_DIR}/lib/onion/src/
-            ${PROJECT_SOURCE_DIR}/lib/mupdf/include/
-    )
-    target_link_directories(
-            sist2_scan PUBLIC
-            ${UUID_LIBRARY_DIRS}
-            ${FFMPEG_LIBRARY_DIRS}
-    )
-    target_compile_options(sist2_scan
-            PRIVATE
-            -Ofast
-            #        -march=native
-            -fno-stack-protector
-            -fomit-frame-pointer
-            )
-    TARGET_LINK_LIBRARIES(
-            sist2_scan
-
-            ${GLIB_LIBRARIES}
-            ${GOBJECT_LIBRARIES}
-            ${UUID_LIBRARIES}
-
-            # ffmpeg
-            ${PROJECT_SOURCE_DIR}/lib/libavcodec.a
-            ${PROJECT_SOURCE_DIR}/lib/libavformat.a
-            ${PROJECT_SOURCE_DIR}/lib/libavutil.a
-            ${PROJECT_SOURCE_DIR}/lib/libswscale.a
-            ${PROJECT_SOURCE_DIR}/lib/libswresample.a
-
-            # mupdf
-            ${PROJECT_SOURCE_DIR}/lib/libmupdf.a
-            ${PROJECT_SOURCE_DIR}/lib/libmupdf-third.a
-
-            ${PROJECT_SOURCE_DIR}/lib/libbz2.a
-            ${PROJECT_SOURCE_DIR}/lib/libmagic.a
-
-            pthread
-            m
-            ${PROJECT_SOURCE_DIR}/lib/libharfbuzz.a
-            ${PROJECT_SOURCE_DIR}/lib/libopenjp2.a
-            freetype
-    )
-endif ()
+        # curl
+        ${PROJECT_SOURCE_DIR}/lib/libcurl.a
+        ${PROJECT_SOURCE_DIR}/lib/libcrypto.a
+        ${PROJECT_SOURCE_DIR}/lib/libssl.a
+        dl
+)
 
 add_custom_target(
         before_sist2
         COMMAND ${CMAKE_CURRENT_SOURCE_DIR}/scripts/before_build.sh
 )
 
-IF (WITH_SIST2)
-    add_dependencies(sist2 before_sist2)
-else ()
-    add_dependencies(sist2_scan before_sist2)
-endif ()
+add_dependencies(sist2 before_sist2)

Docker/Dockerfile

@@ -0,0 +1,22 @@
+FROM ubuntu:19.10
+MAINTAINER simon987 <me@simon987.net>
+
+RUN apt update
+RUN apt install -y libglib2.0-0 libcurl4 libmagic1 libharfbuzz-bin libopenjp2-7 libarchive13 liblzma5 libzstd1 liblz4-1 \
+ curl libtiff5 libpng16-16
+
+RUN mkdir -p /usr/share/tessdata && \
+    cd /usr/share/tessdata/ && \
+    curl -o /usr/share/tessdata/hin.traineddata https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/hin.traineddata &&\
+    curl -o /usr/share/tessdata/jpn.traineddata https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/jpn.traineddata &&\
+    curl -o /usr/share/tessdata/eng.traineddata https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/eng.traineddata &&\
+    curl -o /usr/share/tessdata/fra.traineddata https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/fra.traineddata &&\
+    curl -o /usr/share/tessdata/rus.traineddata https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/rus.traineddata &&\
+    curl -o /usr/share/tessdata/spa.traineddata https://raw.githubusercontent.com/tesseract-ocr/tessdata/master/spa.traineddata && ls -lh
+
+ADD sist2 /root/sist2
+
+ENV LANG C.UTF-8
+ENV LC_ALL C.UTF-8
+
+ENTRYPOINT ["/root/sist2"]

Docker/build.sh

@@ -0,0 +1,12 @@
+rm ./sist2
+cp ../sist2 .
+strip sist2
+
+version=$(./sist2 --version)
+
+echo "Version ${version}"
+docker build . -t simon987/sist2:${version} -t simon987/sist2:latest
+docker push simon987/sist2:${version}
+docker push simon987/sist2:latest
+
+docker run --rm -it simon987/sist2 -v

README.md

@@ -1,5 +1,6 @@
 ![GitHub](https://img.shields.io/github/license/simon987/sist2.svg)
 [![CodeFactor](https://www.codefactor.io/repository/github/simon987/sist2/badge?s=05daa325188aac4eae32c786f3d9cf4e0593f822)](https://www.codefactor.io/repository/github/simon987/sist2)
+[![Development snapshots](https://ci.simon987.net/app/rest/builds/buildType(Sist2_Build)/statusIcon)](https://files.simon987.net/artifacts/Sist2/Build/)
 
 # sist2
 
@@ -11,25 +12,34 @@ sist2 (Simple incremental search tool)
 
 * Fast, low memory usage, multi-threaded
 * Portable (all its features are packaged in a single executable)
-* Extracts text from common file types\*
-* Generates thumbnails\*
+* Extracts text from common file types \*
+* Generates thumbnails \*
 * Incremental scanning
+* Automatic tagging from file attributes via [user scripts](scripting/README.md)
+* Recursive scan inside archive files \*\*
+* OCR support with tesseract \*\*\*
 
 
-\* See [format support](#format-support)
+\* See [format support](#format-support)    
+\*\* See [Archive files](#archive-files)    
+\*\*\* See [OCR](#ocr)    
 
 ## Getting Started
 
 1. Have an [Elasticsearch](https://www.elastic.co/downloads/elasticsearch) instance running
-1. Download the [latest sist2 release](https://github.com/simon987/sist2/releases)
+1. 
+    1. Download the [latest sist2 release](https://github.com/simon987/sist2/releases) *
+    1. *(or)* Download a [development snapshot](https://files.simon987.net/artifacts/Sist2/Build/) *(Not recommended!)*
+    1. *(or)* `docker pull simon987/sist2:latest`
+   
 
-*Windows users*: `sist2` runs under [WSL](https://en.wikipedia.org/wiki/Windows_Subsystem_for_Linux)
-
-*Mac users*: See [#1](https://github.com/simon987/sist2/issues/1)
+\* *Windows users*: **sist2** runs under [WSL](https://en.wikipedia.org/wiki/Windows_Subsystem_for_Linux)    
+\* *Mac users*: See [#1](https://github.com/simon987/sist2/issues/1)
 
 
 ## Example usage
 
+
 ![demo](demo.gif)
 
 See help page `sist2 --help` for more details.
@@ -52,19 +62,76 @@ sist2 index --print ./my_idx > raw_documents.ndjson
 sist2 web --bind 0.0.0.0 --port 4321 ./my_idx1 ./my_idx2 ./my_idx3
 ```
 
+### Use sist2 with docker
+
+**scan**
+```bash
+docker run -it \
+    -v /path/to/files/:/files \
+    -v $PWD/out/:/out \
+    simon987/sist2 scan -t 4 /files -o /out/my_idx1
+```
+**index**
+```bash
+docker run -it --network host\
+    -v $PWD/out/:/out \
+    simon987/sist2 index /out/my_idx1
+```
+
+**web**
+```bash
+docker run --rm --network host -d --name sist2\
+    -v $PWD/out/my_idx:/idx \
+    -v $PWD/my/files:/files
+    simon987/sist2 web --bind 0.0.0.0 /idx
+docker stop sist2
+```
+
+
 ## Format support
 
 File type | Library | Content | Thumbnail | Metadata
 :---|:---|:---|:---|:---
-pdf,xps,cbz,fb2,epub | MuPDF | yes | yes, `png` | title |
+pdf,xps,cbz,fb2,epub | MuPDF | text+ocr | yes, `png` | title |
 `audio/*` | ffmpeg | - | yes, `jpeg` | ID3 tags |
 `video/*` | ffmpeg | - | yes, `jpeg` | title, comment, artist |
 `image/*` | ffmpeg | - | yes, `jpeg` | `EXIF:Artist`, `EXIF:ImageDescription` |
 ttf,ttc,cff,woff,fnt,otf | Freetype2 | - | yes, `bmp` | Name & style |
 `text/plain` | *(none)* | yes | no | - |
-docx, xlsx, pptx |  | *planned* | no | *planned* |
+tar, zip, rar, 7z, ar ...  | Libarchive | yes\* | - | no |
+docx, xlsx, pptx | libOPC | yes | no | no |
 
+\* *See [Archive files](#archive-files)*
+ 
+### Archive files
+**sist2** will scan files stored into archive files (zip, tar, 7z...) as if
+they were directly in the file system. Recursive (archives inside archives)
+scan is also supported.
 
+**Limitations**:
+* Parsing media files with formats that require
+*seek* (e.g. `.gif`, `.mp4` w/ fragmented metadata etc.) is not supported.
+* Archive files are scanned sequentially, by a single thread. On systems where
+**sist2** is not I/O bound, scans might be faster when larger archives are split
+ into smaller parts.
+
+To check if a media file can be parsed without *seek*, execute `cat file.mp4 | ffprobe -`
+ 
+ 
+### OCR
+
+You can enable OCR support for pdf,xps,cbz,fb2,epub file types with the
+`--ocr <lang>` option. Download the language data files with your
+package manager (`apt install tesseract-ocr-eng`) or directly [from Github](https://github.com/tesseract-ocr/tesseract/wiki/Data-Files).
+
+The `simon987/sist2` github image comes with common languages 
+(hin, jpn, eng, fra, rus, spa) pre-installed.
+
+Examples
+```bash
+sist2 scan --ocr jpn ~/Books/Manga/
+sist2 scan --ocr eng ~/Books/Textbooks/
+```
 
 
 ## Build from source
@@ -76,15 +143,10 @@ binaries.
 
     *(Debian)*
     ```bash
-    apt install git cmake pkg-config libglib2.0-dev\
-        libssl-dev uuid-dev libavformat-dev libswscale-dev \
-        python3 libmagic-dev libfreetype6-dev libcurl-dev \
-        libbz2-dev yasm libharfbuzz-dev ragel
-   ```
-    *(FreeBSD)*
-    ```bash
-   pkg install cmake gcc yasm gmake bash ffmpeg e2fsprogs-uuid\
-        autotools ragel
+    apt install git cmake pkg-config libglib2.0-dev \
+        libssl-dev uuid-dev python3 libmagic-dev libfreetype6-dev \
+        libcurl-dev libbz2-dev yasm libharfbuzz-dev ragel \
+        libarchive-dev libtiff5 libpng16-16 libpango1.0-dev
    ```
 
 2. Build
@@ -93,4 +155,4 @@ binaries.
     ./scripts/get_static_libs.sh
     cmake .
     make
-    ```
+    ```

ci/build.sh

@@ -0,0 +1,7 @@
+#!/usr/bin/env bash
+
+./scripts/get_static_libs.sh
+
+cmake .
+make
+strip sist2

include/mce/config.h

@@ -0,0 +1,53 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+ notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+ notice, this list of conditions and the following disclaimer in 
+ the documentation and/or other materials provided with the 
+ distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+ may be used to endorse or promote products derived from this 
+ software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/**@file config/mce/config.h
+ */
+#ifndef MCE_CONFIG_H
+#define MCE_CONFIG_H
+
+#include <libxml/xmlstring.h>
+#include <stdio.h>
+#include <plib/plib.h>
+#include <assert.h>
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#define MCE_NAMESPACE_SUBSUMPTION_ENABLED 0
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif  
+
+#endif /* MCE_CONFIG_H */

include/mce/helper.h

@@ -0,0 +1,189 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file mce/helper.h
+Helper functions needed by mce/textreader.h and mce/textwriter.h to implement MCE:
+- mceQNameLevelAdd(), mceQNameLevelLookup() and mceQNameLevelCleanup() maintain a set of mceQNameLevel_t tuples.
+- mceQNameLevelPush() and mceQNameLevelPopIfMatch() maintain a stack of mceQNameLevel_t tuples.
+- mceCtxInit(), mceCtxCleanup() and mceCtxUnderstandsNamespace() manage a context which holds all information needed to do MCE proprocessing.
+ */
+#include <mce/config.h>
+
+#ifndef MCE_HELPER_H
+#define MCE_HELPER_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+    /**
+      Tiple (ns, ln, level).
+    */
+    typedef struct MCE_QNAME_LEVEL {
+        xmlChar *ns;
+        xmlChar *ln;
+        puint32_t level;
+        puint32_t flag; // used by mceTextWriter
+    } mceQNameLevel_t;
+
+    /**
+     */
+    typedef enum MCE_SKIP_STATE_ENUM {
+        MCE_SKIP_STATE_IGNORE,
+        MCE_SKIP_STATE_ALTERNATE_CONTENT,
+        MCE_SKIP_STATE_CHOICE_MATCHED
+    } mceSkipState_t;
+
+    /**
+     Represents an intervall of levels which are "skipped" i.e. ignored.
+     */
+    typedef struct MCE_SKIP_ITEM {
+        puint32_t level_start;
+        puint32_t level_end;
+        mceSkipState_t state;
+    } mceSkipItem_t;
+
+    /**
+      Either represents a set of (ns, ln, level) triples.
+    */
+    typedef struct MCE_QNAME_LEVEL_SET {
+        mceQNameLevel_t *list_array;
+        puint32_t list_items;
+        puint32_t max_level;
+    } mceQNameLevelSet_t;
+
+    /**
+     The skip stack.
+     */
+    typedef struct MCE_SKIP_STACK {
+        mceSkipItem_t *stack_array;
+        puint32_t stack_items;
+    } mceSkipStack_t;
+
+
+    typedef enum MCE_ERROR_ENUM {
+        MCE_ERROR_NONE,
+        MCE_ERROR_XML,
+        MCE_ERROR_MUST_UNDERSTAND,
+        MCE_ERROR_VALIDATION,
+        MCE_ERROR_MEMORY
+    } mceError_t;
+
+    /**
+      Holds all information to do MCE preprocessing.
+    */
+    typedef struct MCE_CONTEXT {
+        mceQNameLevelSet_t ignorable_set;
+        mceQNameLevelSet_t understands_set;
+        mceQNameLevelSet_t processcontent_set;
+        mceQNameLevelSet_t suspended_set;
+#if (MCE_NAMESPACE_SUBSUMPTION_ENABLED)
+        mceQNameLevelSet_t subsume_namespace_set;
+        mceQNameLevelSet_t subsume_exclude_set;
+        mceQNameLevelSet_t subsume_prefix_set;
+#endif
+        mceSkipStack_t skip_stack;
+        mceError_t error;
+        pbool_t mce_disabled;        
+        puint32_t suspended_level;
+    } mceCtx_t;
+
+    /**
+      Add a new tiple (ns, ln, level) to the triple set \c qname_level_set.
+      The \c ns_sub string is optional and will not be touched.
+    */
+    pbool_t mceQNameLevelAdd(mceQNameLevelSet_t *qname_level_set, const xmlChar *ns, const xmlChar *ln, puint32_t level);
+
+    /**
+      Lookup a tiple (ns, ln, level) via \c ns and \c ln. If \c ignore_ln is PTRUE then the first tiple matching \c ns will be returned.
+    */
+    mceQNameLevel_t* mceQNameLevelLookup(mceQNameLevelSet_t *qname_level_set, const xmlChar *ns, const xmlChar *ln, pbool_t ignore_ln);
+
+    /**
+      Remove all triples (ns, ln, level) where the level greater or equal to \c level.
+    */
+    pbool_t mceQNameLevelCleanup(mceQNameLevelSet_t *qname_level_set, puint32_t level);
+
+    /**
+      Push a new skip intervall (level_start, level_end, state) on the stack \c skip_stack.
+    */
+    pbool_t mceSkipStackPush(mceSkipStack_t *skip_stack, puint32_t level_start, puint32_t level_end, mceSkipState_t state);
+
+    /**
+      Pop the intervall (ns, ln, level) from the stack \c qname_level_array.
+    */
+    void mceSkipStackPop(mceSkipStack_t *skip_stack);
+
+    /**
+     Returns top item or NULL.
+     */
+    mceSkipItem_t *mceSkipStackTop(mceSkipStack_t *skip_stack);
+
+    /**
+     Returns TRUE, if the \c level is in the top skip intervall.
+     */
+    pbool_t mceSkipStackSkip(mceSkipStack_t *skip_stack, puint32_t level);
+
+    /**
+      Initialize the mceCtx_t \c ctx.
+    */
+    pbool_t mceCtxInit(mceCtx_t *ctx);
+
+    /**
+      Cleanup, i.e. release all resourced from the mceCtx_t \c ctx.
+    */
+    pbool_t mceCtxCleanup(mceCtx_t *ctx);
+
+    /**
+      Register the namespace \ns in \c ctx.
+    */
+    pbool_t mceCtxUnderstandsNamespace(mceCtx_t *ctx, const xmlChar *ns);
+
+    /**
+     Register the namespace \ns in \c ctx.
+     */
+    pbool_t mceCtxSuspendProcessing(mceCtx_t *ctx, const xmlChar *ns, const xmlChar *ln);
+    
+
+
+#if (MCE_NAMESPACE_SUBSUMPTION_ENABLED)
+    /**
+    Subsume namespace \c ns_new with \c ns_old.
+     */
+    pbool_t mceCtxSubsumeNamespace(mceCtx_t *ctx, const xmlChar *prefix_new, const xmlChar *ns_new, const xmlChar *ns_old);
+#endif
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif
+
+#endif /* MCE_HELPER_H */

include/mce/textreader.h

@@ -0,0 +1,464 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file mce/textreader.h
+ 
+ */
+#ifndef MCE_TEXTREADER_H
+#define MCE_TEXTREADER_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+    /**
+      A handle to an MCE-aware libxml2 xmlTextReader.
+    */
+    typedef struct MCE_TEXTREADER mceTextReader_t;
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif
+
+
+#include <mce/config.h>
+#include <opc/opc.h>
+#include <mce/helper.h>
+#include <libxml/xmlwriter.h>
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+    struct MCE_TEXTREADER {
+        xmlTextReaderPtr reader;
+        mceCtx_t mceCtx;
+    };
+
+    /**
+      Wrapper around an libxml2 xmlTextReaderRead function.
+      \see http://xmlsoft.org/html/libxml-xmlreader.html#xmlTextReaderRead
+    */
+    int mceTextReaderRead(mceTextReader_t *mceTextReader);
+
+    /**
+      Wrapper around a libxml2 xmlTextReaderNext function.
+      \see http://xmlsoft.org/html/libxml-xmlreader.html#xmlTextReaderNext
+    */
+    int mceTextReaderNext(mceTextReader_t *mceTextReader);
+
+    /** 
+      Creates an mceTextReader from an XmlTextReader. 
+      \code
+      mceTextReader reader;
+      mceTextReaderInit(&reader, xmlNewTextReaderFilename("sample.xml"));
+      // reader is ready to use.
+      mceTextReaderCleanup(&reader);
+      \endcode
+      \see http://xmlsoft.org/html/libxml-xmlreader.html#xmlNewTextReaderFilename
+    */
+    int mceTextReaderInit(mceTextReader_t *mceTextReader, xmlTextReaderPtr reader);
+
+    /**
+      Cleanup MCE reader, i.e. free all resources. Also calls xmlTextReaderClose and xmlFreeTextReader.
+      \see http://xmlsoft.org/html/libxml-xmlreader.html#xmlTextReaderClose
+      \see http://xmlsoft.org/html/libxml-xmlreader.html#xmlFreeTextReader
+    */
+    int mceTextReaderCleanup(mceTextReader_t *mceTextReader);
+
+    /** 
+      Reads all events \c mceTextReader and pipes them to \writer.
+      \code
+      mceTextReader reader;
+      mceTextReaderInit(&reader, xmlNewTextReaderFilename("sample.xml"));
+      mceTextReaderUnderstandsNamespace(&reader, _X("http://myextension"));
+      xmlTextWriterPtr writer=xmlNewTextWriterFilename("out.xml", 0);
+      mceTextReaderDump(&reader, writer, P_FALSE);
+      xmlFreeTextWriter(writer);
+      mceTextReaderCleanup(&reader);
+      \endcode
+      */
+    int mceTextReaderDump(mceTextReader_t *mceTextReader, xmlTextWriter *writer, pbool_t fragment);
+
+    /**
+      Registers an MCE namespace.
+      \see mceTextReaderDump()
+      */
+    int mceTextReaderUnderstandsNamespace(mceTextReader_t *mceTextReader, const xmlChar *ns);
+
+    /**
+     Disable MCE processing.
+     \return Returns old value.
+     */
+    pbool_t mceTextReaderDisableMCE(mceTextReader_t *mceTextReader, pbool_t flag);
+
+
+    /**
+     Signal an error to the MCE processor.
+     */
+    void mceRaiseError(xmlTextReader *reader, mceCtx_t *ctx, mceError_t error, const xmlChar *str, ...);
+
+    /**
+        Internal function which does the MCE postprocessing. E.g. mceTextReaderRead() is implemented as
+        \code
+        mceTextReaderPostprocess(mceTextReader->reader, &mceTextReader->mceCtx, xmlTextReaderRead(mceTextReader->reader))
+        \endcode
+        This function is exposed to make existing libxm2 xmlTextReader MCE aware.
+    */
+    int mceTextReaderPostprocess(xmlTextReader *reader, mceCtx_t *ctx, int ret);
+
+    /**
+     Get the error code.
+     */
+    mceError_t mceTextReaderGetError(mceTextReader_t *mceTextReader);
+
+/**
+ Helper macro to declare a start/end document block in a declarative way:
+ \code
+  mce_start_document(reader) {
+  } mce_end_document(reader);
+  \endcode
+  \hideinitializer
+*/
+#define mce_start_document(_reader_) \
+    if (NULL!=(_reader_)) {            \
+        mceTextReaderRead(_reader_); \
+        if (0)                     
+
+/**
+  \see mce_start_document.
+  \hideinitializer
+*/
+#define mce_end_document(_reader_)   \
+    } /* if (NULL!=reader) */        \
+
+
+/**
+  Container for mce_start_element and mce_start_attribute declarations.
+  \see mce_match_element
+  \see mce_match_attribute
+  \hideinitializer
+  */
+#define mce_start_choice(_reader_)  \
+    if (0)                          
+
+/**
+  \see mce_start_choice
+  \hideinitializer
+  */
+#define mce_end_choice(_reader_) 
+
+
+/**
+  Skips the attributes. 
+  \see mce_match_element.
+  \hideinitializer
+*/
+#define mce_skip_attributes(_reader_) \
+    mce_start_attributes(_reader_) {  \
+    } mce_end_attributes(_reader_);   
+
+
+/**
+  Skips the attributes. 
+  \see mce_match_attribute.
+  \hideinitializer
+*/
+#define mce_skip_children(_reader_) \
+    mce_start_children(_reader_) {  \
+    } mce_end_children(_reader_);   
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_start_children(_reader_)                  \
+if (!xmlTextReaderIsEmptyElement((_reader_)->reader)) { \
+    mceTextReaderRead(_reader_); do {                 \
+        if (0)                                        
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_end_children(_reader_)                                                      \
+        else {                                                                          \
+            if (XML_READER_TYPE_END_ELEMENT!=xmlTextReaderNodeType((_reader_)->reader)) { \
+                mceTextReaderNext(_reader_); /*skip unhandled element */                \
+            }                                                                           \
+        }                                                                               \
+    } while(XML_READER_TYPE_END_ELEMENT!=xmlTextReaderNodeType((_reader_)->reader) &&     \
+            XML_READER_TYPE_NONE!=xmlTextReaderNodeType((_reader_)->reader));             \
+} /* if (!xmlTextReaderIsEmptyElement(reader->reader)) */                               
+
+
+/**
+  Helper macro to match an element. Usefull for calling code in a seperate function:
+
+  \code
+  void handleElement(reader) {
+    mce_start_choice(reader) {
+        mce_start_element(reader, _X("ns"), _X("element")) {
+            
+        } mce_end_element(reader)
+    } mce_end_choice(reader);
+  }
+
+  void parse(reader) {
+    mce_start_document(reader) {
+      mce_start_element(reader, _X("ns"), _X("ln")) {
+        mce_skip_attributes(reader);
+        mce_start_children(reader) {
+           mce_match_element(reader, _X("ns"), _X("element")) {
+             handleElement(reader);
+           }
+        } mce_end_children(reader);
+      } mce_end_element();
+    } mce_end_document(reader);
+  }
+  \endcode
+  \hideinitializer
+*/
+#define mce_match_element(_reader_, ns, ln)                                                       \
+    } else if (XML_READER_TYPE_ELEMENT==xmlTextReaderNodeType((_reader_)->reader)                 \
+            && (NULL==ns || 0==xmlStrcmp(ns, xmlTextReaderConstNamespaceUri((_reader_)->reader))) \
+            && (NULL==ln || 0==xmlStrcmp(ln, xmlTextReaderConstLocalName((_reader_)->reader)))) { 
+
+
+/**
+ Helper macro to declare a element block in a declarative way:
+ \code
+  mce_start_element(reader) {
+    mce_start_attributes(reader) {
+      mce_start_attribute(reader, _X("ns"), _X("lnA")) {
+         // code for handling lnA.
+      } mce_end_attribute(reader);
+      mce_start_attribute(reader, _X("ns"), _X("lnB")) {
+         // code for handling lnB.
+      } mce_end_attribute(reader);
+    } mce_end_attributes(reader);
+    mce_start_children(reader) {
+        mce_start_element(reader, _X("ns"), _X("lnA")) {
+         // code for handling lnA.
+        } mce_end_element(reader);
+        mce_start_element(reader, _X("ns"), _X("lnB")) {
+         // code for handling lnB.
+        } mce_end_element(reader);
+        mce_start_text(reader) {
+         // code for handling text.
+        } mce_end_text(reader);
+    } mce_end_children(reader);
+  } mce_end_element(reader);
+  \endcode
+  \hideinitializer
+*/
+#define mce_start_element(_reader_, ns, ln) \
+    mce_match_element(_reader_, ns, ln)     
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_end_element(_reader_) \
+    mceTextReaderNext(_reader_)   
+
+/**
+  Matches #TEXT without consuming it.
+  \hideinitializer
+*/
+#define mce_match_text(_reader_)                                                                   \
+    } else if (XML_READER_TYPE_TEXT==xmlTextReaderNodeType((_reader_)->reader)                     \
+            || XML_READER_TYPE_SIGNIFICANT_WHITESPACE==xmlTextReaderNodeType((_reader_)->reader)) {
+
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_start_text(_reader_) \
+    mce_match_text(_reader_)      
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_end_text(_reader_) \
+    mceTextReaderNext(_reader_)
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_start_attributes(_reader_)                            \
+    if (1==xmlTextReaderMoveToFirstAttribute((_reader_)->reader)) { \
+        do {                                                      \
+            if (0)                                                
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_end_attributes(_reader_)                                    \
+            else { /* skipped attribute */ }                            \
+        } while(1==xmlTextReaderMoveToNextAttribute((_reader_)->reader)); \
+    xmlTextReaderMoveToElement((_reader_)->reader); }                     
+
+/**
+  Helper macro to match an attribute. Usefull for calling code in a seperate function:
+
+  \code
+  void handleA(reader) {
+    mce_start_choice(reader) {
+        mce_start_attribute(reader, _X("ns"), _X("attr")) {
+
+        } mce_end_attribute(reader);
+    } mce_end_choice(reader);
+  }
+
+  void parse(reader) {
+    mce_start_document(reader) {
+      mce_start_element(reader, _X("ns"), _X("ln")) {
+        mce_start_attributes(reader) {
+           mce_match_attribute(reader, _X("ns"), _X("attr")) {
+             handleA(reader);
+           }
+        } mce_end_attributes(reader);
+        mce_skip_children(reader);
+      } mce_end_element();
+    } mce_end_document(reader);
+  }
+  \endcode
+  \hideinitializer
+*/
+#define mce_match_attribute(_reader_, ns, ln)                                                   \
+    } else if ((NULL==ns || 0==xmlStrcmp(ns, xmlTextReaderConstNamespaceUri((_reader_)->reader))) \
+            && (NULL==ln || 0==xmlStrcmp(ln, xmlTextReaderConstLocalName((_reader_)->reader)))) { 
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_start_attribute(_reader_, ns, ln) \
+    mce_match_attribute(_reader_, ns, ln) 
+
+/**
+  \see mce_start_element.
+  \hideinitializer
+*/
+#define mce_end_attribute(_reader_)
+
+
+/**
+  Error handling for MCE parsers.
+  \code
+   mce_start_element(&reader, NULL, _X("Default")) {
+       const xmlChar *ext=NULL;
+       const xmlChar *type=NULL;
+       mce_start_attributes(&reader) {
+           mce_start_attribute(&reader, NULL, _X("Extension")) {
+               ext=xmlTextReaderConstValue(reader.reader);
+           } mce_end_attribute(&reader);
+           mce_start_attribute(&reader, NULL, _X("ContentType")) {
+               type=xmlTextReaderConstValue(reader.reader);
+           } mce_end_attribute(&reader);
+       } mce_end_attributes(&reader);
+       mce_error_guard_start(&reader) {
+           mce_error(&reader, NULL==ext || ext[0]==0, MCE_ERROR_VALIDATION, "Missing @Extension attribute!");
+           mce_error(&reader, NULL==type || type[0]==0, MCE_ERROR_VALIDATION, "Missing @ContentType attribute!");
+           opcContainerType *ct=insertType(c, type, OPC_TRUE);
+           mce_error(&reader, NULL==ct, MCE_ERROR_MEMORY, NULL);
+           opcContainerExtension *ce=opcContainerInsertExtension(c, ext, OPC_TRUE);
+           mce_error(&reader, NULL==ce, MCE_ERROR_MEMORY, NULL);
+           mce_errorf(&reader, NULL!=ce->type && 0!=xmlStrcmp(ce->type, type), MCE_ERROR_VALIDATION, "Extension \"%s\" is mapped to type \"%s\" as well as \"%s\"", ext, type, ce->type);
+           ce->type=ct->type;
+       } mce_error_guard_end(&reader);
+       mce_skip_children(&reader);
+   } mce_end_element(&reader);
+  \endcode
+  \hideinitializer
+*/
+#define mce_error_guard_start(_reader_) if (MCE_ERROR_NONE==(_reader_)->mceCtx.error) do {
+
+/**
+  \see mce_error_guard_start
+  \hideinitializer
+*/
+#define mce_error_guard_end(_reader_)  } while(0)
+
+/**
+  Signal an error if guard if false.
+  \hideinitializer
+*/
+#define mce_error(_reader_, guard, err, msg) if (guard) { (_reader_)->mceCtx.error=(err); fprintf(stderr, (NULL!=msg?msg:#err));  continue; }
+
+/**
+  Signal an error if guard if false.
+  \hideinitializer
+*/
+#if defined(__GNUC__)
+#define mce_errorf(_reader_, guard, err, msg, ...) if (guard) { mceRaiseError((_reader_)->reader, &(_reader_)->mceCtx, err, _X((NULL!=msg?msg:#err)), ##__VA_ARGS__ );  continue; }
+#else
+#define mce_errorf(_reader_, guard, err, msg, ...) if (guard) { mceRaiseError((_reader_)->reader, &(_reader_)->mceCtx, err, _X((NULL!=msg?msg:#err)), __VA_ARGS__ );  continue; }
+#endif
+
+/**
+  Only issues the error when in "strict mode".
+  \hideinitializer
+*/
+#define mce_error_strict mce_error
+
+/**
+  \see mce_error_strict
+  \hideinitializer
+*/
+#define mce_error_strictf mce_errorf
+
+
+/**
+  Marker for a MCE defintion.
+  \hideinitializer
+*/
+#define mce_def
+
+/**
+  Marker for a MCE reference.
+  \hideinitializer
+*/
+#define mce_ref(r) (r)
+
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif
+
+#endif /* MCE_TEXTREADER_H */

include/mce/textwriter.h

@@ -0,0 +1,176 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file mce/textwriter.h
+
+*/
+#include <mce/config.h>
+#include <libxml/xmlwriter.h>
+#include <mce/helper.h>
+
+#ifndef MCE_TEXTWRITER_H
+#define MCE_TEXTWRITER_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+
+/**
+  Default flags for an MCE namespace declaration.
+  */
+#define MCE_DEFAULT 0x0
+
+/**
+  Flags MCE namespace declaration "ignorable".
+  */
+#define MCE_IGNORABLE 0x1
+
+/**
+  Flags MCE namespace declaration "must understand".
+  */
+#define MCE_MUSTUNDERSTAND 0x2
+
+    /**
+      The MCE text writer context.
+      */
+    typedef struct MCE_TEXTWRITER_STRUCT mceTextWriter;
+
+    /**
+      Create a new MCE text writer.
+      \see http://xmlsoft.org/html/libxml-xmlIO.html#xmlOutputBufferCreateIO
+      \see http://xmlsoft.org/html/libxml-xmlwriter.html#xmlNewTextWriter
+      */
+    mceTextWriter *mceTextWriterCreateIO(xmlOutputWriteCallback iowrite, xmlOutputCloseCallback  ioclose, void *ioctx, xmlCharEncodingHandlerPtr encoder);
+
+    /**
+      Helper which create a new MCE text writer for a FILE handle.
+      */
+    mceTextWriter *mceNewTextWriterFile(FILE *file);
+
+    /**
+      Free all resources for \w.
+      */
+    int mceTextWriterFree(mceTextWriter *w);
+
+    /**
+      \see http://xmlsoft.org/html/libxml-xmlwriter.html#xmlTextWriterStartDocument
+      */
+    int mceTextWriterStartDocument(mceTextWriter *w);
+
+    /**
+      \see http://xmlsoft.org/html/libxml-xmlwriter.html#xmlTextWriterEndDocument
+      */
+    int mceTextWriterEndDocument(mceTextWriter *w);
+
+    /**
+      Start a new XML element. If ns==NULL then there is no namespace and ""==ns means the default namespace.
+      \see http://xmlsoft.org/html/libxml-xmlwriter.html#xmlTextWriterStartElement
+      \see http://xmlsoft.org/html/libxml-xmlwriter.html#xmlTextWriterStartElementNS
+      */
+    int mceTextWriterStartElement(mceTextWriter *w, const xmlChar *ns, const xmlChar *ln);
+
+    /**
+      \see http://xmlsoft.org/html/libxml-xmlwriter.html#xmlTextWriterEndElement
+      */
+    int mceTextWriterEndElement(mceTextWriter *w, const xmlChar *ns, const xmlChar *ln);
+
+    /**
+      \see http://xmlsoft.org/html/libxml-xmlwriter.html#xmlTextWriterWriteString
+      */
+    int mceTextWriterWriteString(mceTextWriter *w, const xmlChar *content);
+
+    /**
+      Register a namespace. Must be called before mceTextWriterStartElement.
+      \see MCE_DEFAULT
+      \see MCE_IGNORABLE
+      \see MCE_MUSTUNDERSTAND
+      */
+    const xmlChar *mceTextWriterRegisterNamespace(mceTextWriter *w, const xmlChar *ns, const xmlChar *prefix, int flags);
+
+    /**
+      Register qname (ns, ln) as a "process content" element wrt. MCE. Must be called before mceTextWriterStartElement.
+      */
+    int mceTextWriterProcessContent(mceTextWriter *w, const xmlChar *ns, const xmlChar *ln);
+
+    /**
+      Writes a formatted attribute.
+      \see http://xmlsoft.org/html/libxml-xmlwriter.html#xmlTextWriterWriteFormatAttribute
+      */
+    int mceTextWriterAttributeF(mceTextWriter *w, const xmlChar *ns, const xmlChar *ln, const char *value, ...);
+
+    /**
+      Starts an MCE alternate content section.
+      */
+    int mceTextWriterStartAlternateContent(mceTextWriter *w);
+
+    /**
+      Ends an MCE alternate content section.
+      */
+    int mceTextWriterEndAlternateContent(mceTextWriter *w);
+
+    /**
+      Start an MCE choice.
+      */
+    int mceTextWriterStartChoice(mceTextWriter *w, const xmlChar *ns);
+
+    /**
+      Ends an MCE choice.
+      */
+    int mceTextWriterEndChoice(mceTextWriter *w);
+
+    /**
+      Start an MCE fallback.
+      */
+    int mceTextWriterStartFallback(mceTextWriter *w);
+
+    /**
+      Ends an MCE fallback.
+      */
+    int mceTextWriterEndFallback(mceTextWriter *w);
+
+
+    /**
+      Returns the underlying xmlTextWriter.
+      */
+    xmlTextWriterPtr mceTextWriterIntern(mceTextWriter *w);
+
+    /**
+      Helper which create a new xmlTextWriterPtr for a FILE handle.
+      */
+    xmlTextWriterPtr xmlNewTextWriterFile(FILE *file);
+
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+
+#endif /* MCE_TEXTWRITER_H */

include/opc/config.h

@@ -0,0 +1,189 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+ notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+ notice, this list of conditions and the following disclaimer in 
+ the documentation and/or other materials provided with the 
+ distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+ may be used to endorse or promote products derived from this 
+ software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+/**@file config/opc/config.h
+ */
+#ifndef OPC_CONFIG_H
+#define OPC_CONFIG_H
+
+#include <libxml/xmlstring.h>
+#include <plib/plib.h>
+#include <assert.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif   
+
+
+/**
+  Assert expression e is true. Will be removed entirely in release mode.
+  \hideinitializer
+ */
+#define OPC_ASSERT(e) assert(e)
+
+/**
+  Assert expression e is true. Expression will be executed in release mode too.
+  \hideinitializer
+ */
+#ifdef NDEBUG
+#define OPC_ENSURE(e) (void)(e)
+#else
+#define OPC_ENSURE(e) assert(e)
+#endif
+
+
+/**
+  Constant for boolean true.
+  \hideinitializer
+ */
+#define OPC_TRUE (0==0)
+
+/**
+  Constant for boolean false.
+  \hideinitializer
+ */
+#define OPC_FALSE (0==1)
+
+    /** 
+      Boolean type.
+      \hideinitializer
+      */
+    typedef pbool_t opc_bool_t;
+
+    /** 
+      Type which represents an offset in e.g. a file.
+      \hideinitializer
+      */
+    typedef pofs_t opc_ofs_t;
+
+    /** 
+      8-bit unsigned integer.
+      \hideinitializer
+      */
+    typedef puint8_t opc_uint8_t;
+
+    /** 
+      16-bit unsigned integer.
+      \hideinitializer
+      */
+    typedef puint16_t opc_uint16_t;
+
+    /** 
+      32-bit unsigned integer.
+      \hideinitializer
+      */
+    typedef puint32_t opc_uint32_t;
+
+    /** 
+      64-bit unsigned integer.
+      \hideinitializer
+      */
+    typedef puint64_t opc_uint64_t;
+
+    /** 
+      8-bit signed integer.
+      \hideinitializer
+      */
+    typedef pint8_t opc_int8_t;
+
+    /** 
+      16-bit signed integer.
+      \hideinitializer
+      */
+    typedef pint16_t opc_int16_t;
+
+    /** 
+      32-bit signed integer.
+      \hideinitializer
+      */
+    typedef pint32_t opc_int32_t;
+
+    /** 
+      64-bit signed integer.
+      \hideinitializer
+      */
+    typedef pint64_t opc_int64_t;
+
+/**
+  Default size fo the deflate buffer used by zlib.
+  */
+#define OPC_DEFLATE_BUFFER_SIZE 4096
+
+/**
+  Max system path len.
+  */
+#define OPC_MAX_PATH 512
+
+    /**
+      Error codes for the OPC module.
+      */
+    typedef enum OPC_ERROR_ENUM {
+        OPC_ERROR_NONE,
+        OPC_ERROR_STREAM,
+        OPC_ERROR_SEEK, // can't seek
+        OPC_ERROR_UNSUPPORTED_DATA_DESCRIPTOR,
+        OPC_ERROR_UNSUPPORTED_COMPRESSION,
+        OPC_ERROR_DEFLATE,
+        OPC_ERROR_HEADER,
+        OPC_ERROR_MEMORY,
+        OPC_ERROR_XML, 
+        OPC_ERROR_USER // user triggered an abort
+    } opc_error_t;
+    
+    /**
+      Compression options for OPC streams.
+      */
+    typedef enum OPC_COMPRESSIONOPTION_ENUM {
+        OPC_COMPRESSIONOPTION_NONE,
+        OPC_COMPRESSIONOPTION_NORMAL,
+        OPC_COMPRESSIONOPTION_MAXIMUM,
+        OPC_COMPRESSIONOPTION_FAST,
+        OPC_COMPRESSIONOPTION_SUPERFAST
+    } opcCompressionOption_t;
+
+
+/**
+  Helper for debug logs.
+  \hideinitializer
+  */
+#define opc_logf printf
+
+/**
+  Abstraction for memset(m, 0, s).
+  \hideinitializer
+ */
+#define opc_bzero_mem(m,s) memset(m, 0, s)
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif  
+
+#endif /* OPC_CONFIG_H */

include/opc/container.h

@@ -0,0 +1,300 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/container.h
+
+ The container.h module has the fundamental methods for dealing with ZIP-based OPC container. 
+ 
+ OPC container can be opened in READ-ONLY mode, WRITE-ONLY mode, READ/WRITE mode, TEMPLATE mode and TRANSITION mode. 
+ The most notable mode is the READ/WRITE mode, which gives you concurrent stream-based READ and WRITE access to a 
+ single ZIP-based OPC container. This is achieved without the use of temporary files by taking advantage of the 
+ OPC specific “interleave” mode. \see http://standards.iso.org/ittf/PubliclyAvailableStandards/c051459_ISOIEC_29500-2_2008(E).zip
+ 
+ The TEMPLATE mode allows very fast customized "cloning" of ZIP-based OPC container by using "RAW access" to the ZIP streams. 
+ The TRANSITION mode is a special version of the TEMPLATE mode, which allows transition-based READ/WRITE access to the 
+ ZIP-based OPC container using a temporary file.
+ 
+ */
+#include <opc/config.h>
+#include <opc/file.h>
+
+#ifndef OPC_CONTAINER_H
+#define OPC_CONTAINER_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+    /**
+     Handle to an OPC container created by \ref opcContainerOpen.
+     \see opcContainerOpen.
+     */
+    typedef struct OPC_CONTAINER_STRUCT opcContainer;
+    
+    /**
+     Modes for opcContainerOpen();
+     \see opcContainerOpen
+     */
+    typedef enum {
+        /**
+         Opens the OPC container denoted by \a fileName in READ-ONLY mode. The \a destName parameter must be \a NULL.
+         \hideinitializer
+         */
+        OPC_OPEN_READ_ONLY=0, 
+        /**
+         Opens the OPC container denoted by \a fileName in WRITE-ONLY mode. The \a destName parameter must be \a NULL.
+         \hideinitializer
+         */
+        OPC_OPEN_WRITE_ONLY=1,
+        /**
+         Opens the OPC container denoted by \a fileName in READ/WRITE mode. The \a destName parameter must be \a NULL.
+         \hideinitializer
+         */
+        OPC_OPEN_READ_WRITE=2,
+        /**
+         This mode will open the container denoted by \a fileName in READ-ONLY mode and the container denoted by 
+         \a destName in write-only mode. Any modifications will be written to the container denoted by \a destName 
+         and the unmodified streams from \a fileName will be written to \a destName on closing.
+         \warning Currently not implemented.
+         \hideinitializer
+         */
+        OPC_OPEN_TEMPLATE=3,
+        /**
+         Like the OPC_OPEN_TEMPLATE mode, but the \a destName will be renamed to the \a fileName on closing. If \a destName 
+         is \a NULL, then the name of the temporary file will be generated automatically.
+         \warning Currently not implemented.
+         \hideinitializer
+         */
+        OPC_OPEN_TRANSITION=4
+    } opcContainerOpenMode; 
+    
+    /** Modes for opcContainerClose.
+     \see opcContainerClose.
+     */
+    typedef enum {
+        /**
+         Close the OPC container without any further postprocessing.
+         \hideinitializer
+         */
+        OPC_CLOSE_NOW = 0,
+        /**
+         Close the OPC container and trim the file by removing unused fragments like e.g. 
+         deleted parts.
+         \hideinitializer
+         */
+        OPC_CLOSE_TRIM = 1,
+        /**
+         Close the OPC container like in \a OPC_CLOSE_TRIM mode, but additionally remove any 
+         "interleaved" parts by reordering them.
+         \warning Currently not implemented. Same semantic as OPC_CLOSE_TRIM.       
+         \hideinitializer
+         */
+        OPC_CLOSE_DEFRAG = 2
+    } opcContainerCloseMode;
+    
+    /**
+     Opens a ZIP-based OPC container.
+     @param[in] fileName. For more details see \ref opcContainerOpenMode.
+     @param[in] mode. For more details see \ref opcContainerOpenMode.
+     @param[in] userContext. Will not be modified by libopc. Can be used to e.g. store the "this" pointer for C++ bindings.
+     @param[in] destName. For more details see \ref opcContainerOpenMode.
+     @return \a NULL if failed. 
+     \see opcContainerOpenMode
+     \see opcContainerDump
+     */
+    opcContainer* opcContainerOpen(const xmlChar *fileName, 
+                                   opcContainerOpenMode mode, 
+                                   void *userContext, 
+                                   const xmlChar *destName);
+
+    /**
+     Opens a ZIP-based OPC container from memory.
+     @param[in] data. 
+     @param[in] data_len.
+     @param[in] userContext. Will not be modified by libopc. Can be used to e.g. store the "this" pointer for C++ bindings.
+     @param[in] mode. For more details see \ref opcContainerOpenMode.
+     @return \a NULL if failed. 
+     */
+    opcContainer* opcContainerOpenMem(const opc_uint8_t *data, opc_uint32_t data_len,
+                                      opcContainerOpenMode mode, 
+                                      void *userContext);
+
+    /**
+     Opens a ZIP-based OPC container from memory.
+     @param[in] ioread. 
+     @param[in] iowrite. 
+     @param[in] ioclose. 
+     @param[in] ioseek. 
+     @param[in] iotrim. 
+     @param[in] ioflush. 
+     @param[in] iocontext. 
+     @param[in] file_size. 
+     @param[in] userContext. Will not be modified by libopc. Can be used to e.g. store the "this" pointer for C++ bindings.
+     @param[in] mode. For more details see \ref opcContainerOpenMode.
+     @return \a NULL if failed. 
+     */
+    opcContainer* opcContainerOpenIO(opcFileReadCallback *ioread,
+                                     opcFileWriteCallback *iowrite,
+                                     opcFileCloseCallback *ioclose,
+                                     opcFileSeekCallback *ioseek,
+                                     opcFileTrimCallback *iotrim,
+                                     opcFileFlushCallback *ioflush,
+                                     void *iocontext,
+                                     pofs_t file_size,
+                                     opcContainerOpenMode mode, 
+                                     void *userContext);
+    
+    /**
+     Close an OPC container.
+     @param[in] c. \ref opcContainer openered by \ref opcContainerOpen.
+     @param[in] mode. For more information see \ref opcContainerCloseMode.
+     @return Non-zero if successful.
+     \see opcContainerOpen
+     \see opcContainerCloseMode
+     */
+    opc_error_t opcContainerClose(opcContainer *c, opcContainerCloseMode mode);
+    
+    /**
+     Returns the unmodified user context passed to \ref opcContainerOpen.
+     \see opcContainerOpen
+     */
+    void *opcContainerGetUserContext(opcContainer *c);
+    
+    /**
+     List all types, relations and parts of the container \a c to \a out.
+     \par Sample:
+     \include opc_dump.c
+     */
+    opc_error_t opcContainerDump(opcContainer *c, FILE *out);
+    
+    /**
+     Exports the OPC container to "Flat OPC" (http://blogs.msdn.com/b/ericwhite/archive/2008/09/29/the-flat-opc-format.aspx).
+     The flat versions of an OPC file are very important when dealing with e.g XSL(T)-based or Javascript-based transformations.
+     \see opcContainerFlatImport.
+     \todo Implementation needed.
+     */
+    int opcContainerFlatExport(opcContainer *c, const xmlChar *fileName);
+    
+    /**
+     Imports the flat version of an OPC container. 
+     \see opcContainerFlatExport.
+     \todo Implementation needed.
+     */
+    int opcContainerFlatImport(opcContainer *c, const xmlChar *fileName);
+    
+    /**
+     Iterate all types.
+     \code
+     for(xmlChar *type=opcContentTypeFirst(c);
+         NULL!=type;
+         type=opcContentTypeNext(c, type)) {
+        printf("%s\n", type);
+     }
+     \endcode
+    */
+    const xmlChar *opcContentTypeFirst(opcContainer *container);
+    
+    /**
+     \see opcContentTypeNext()
+    */
+    const xmlChar *opcContentTypeNext(opcContainer *container, const xmlChar *type);
+
+    /**
+     Iterate extensions.
+     \code
+     for(const xmlChar *ext=opcExtensionFirst(c);
+         NULL!=ext;
+         ext=opcExtensionNext(ext)) {
+        printf("%s\n", ext);
+     }
+     \endcode
+    */
+    const xmlChar *opcExtensionFirst(opcContainer *container);
+    
+    /**
+     \see opcExtensionFirst()
+     */
+    const xmlChar *opcExtensionNext(opcContainer *container, const xmlChar *ext);
+    
+    /**
+     Get registered type for extension.
+     \see opcExtensionRegister()
+     */
+    const xmlChar *opcExtensionGetType(opcContainer *container, const xmlChar *ext);
+
+    /**
+     Register a mime-type and and extension.
+     \see opcExtensionGetType()
+     */
+    const xmlChar *opcExtensionRegister(opcContainer *container, const xmlChar *ext, const xmlChar *type);
+
+
+    /**
+     Iterator through all relation types of the container:
+     \code
+     for(xmlChar *type=opcRelationTypeFirst(c);
+         NULL!=type;
+         type=opcRelationTypeNext(c, type)) {
+        printf("%s\n", type);
+     }
+     \endcode
+     */
+    const xmlChar *opcRelationTypeFirst(opcContainer *container);
+
+    /**
+     \see opcRelationTypeFirst()
+    */
+    const xmlChar *opcRelationTypeNext(opcContainer *container, const xmlChar *type);
+
+
+    /**
+     Iterator through all relation types of the container:
+     \code
+     for(xmlChar *target=opcExternalTargetFirst(c);
+         NULL!=target;
+         type=opcExternalTargetNext(c, target)) {
+        printf("%s\n", target);
+     }
+     \endcode
+     */
+    const xmlChar *opcExternalTargetFirst(opcContainer *container);
+
+    /**
+     \see opcExternalTargetFirst()
+    */
+    const xmlChar *opcExternalTargetNext(opcContainer *container, const xmlChar *target);
+
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_CONTAINER_H */

include/opc/file.h

@@ -0,0 +1,200 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/file.h
+ The opc module contains the file library functions.
+*/
+#include <opc/config.h>
+
+#ifndef OPC_FILE_H
+#define OPC_FILE_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+
+/**
+  Flag for READ access.
+  \hideinitializer
+*/
+#define OPC_FILE_READ  (1<<0)
+
+/**
+  Flag for WRITE access.
+  \hideinitializer
+*/
+#define OPC_FILE_WRITE (1<<1)
+
+/**
+  Flag indicates that file will be truncated when opened.
+  \hideinitializer
+*/
+#define OPC_FILE_TRUNC  (1<<2)
+
+
+    /** 
+      Abstraction for see modes.
+      */
+    typedef enum OPC_FILESEEKMODE_ENUM {
+        opcFileSeekSet = SEEK_SET,
+        opcFileSeekCur = SEEK_CUR,
+        opcFileSeekEnd = SEEK_END
+    } opcFileSeekMode;
+
+     /**
+      Callback to read a file. E.g. for a FILE * context this can be implemented as
+      \code
+      static int opcFileRead(void *iocontext, char *buffer, int len) {
+          return fread(buffer, sizeof(char), len, (FILE*)iocontext);
+      }
+      \endcode
+      */
+    typedef int opcFileReadCallback(void *iocontext, char *buffer, int len);
+
+     /**
+      Callback to write a file. E.g. for a FILE * context this can be implemented as
+      \code
+      static int opcFileWrite(void *iocontext, const char *buffer, int len) {
+          return fwrite(buffer, sizeof(char), len, (FILE*)iocontext);
+      }
+      \endcode
+      */
+    typedef int opcFileWriteCallback(void *iocontext, const char *buffer, int len);
+
+     /**
+      Callback to close a file. E.g. for a FILE * context this can be implemented as
+      \code
+      static int opcFileClose(void *iocontext) {
+          return fclose((FILE*)iocontext);
+      }
+      \endcode
+      */
+    typedef int opcFileCloseCallback(void *iocontext);
+
+     /**
+      Callback to seek a file. E.g. for a FILE * context this can be implemented as
+      \code
+      static opc_ofs_t opcFileSeek(void *iocontext, opc_ofs_t ofs) {
+          int ret=fseek((FILE*)iocontext, ofs, SEEK_SET);
+          if (ret>=0) {
+              return ftell((FILE*)iocontext);
+          } else {
+              return ret;
+          }
+      }
+      \endcode
+      */
+    typedef opc_ofs_t opcFileSeekCallback(void *iocontext, opc_ofs_t ofs);
+
+     /**
+      Callback to trim a file. E.g. for a FILE * context this can be implemented as
+      \code
+      static int opcFileTrim(void *iocontext, opc_ofs_t new_size) {
+      #ifdef WIN32
+          return _chsize(fileno((FILE*)iocontext), new_size);
+      #else
+          return ftruncate(fileno((FILE*)iocontext), new_size);
+      #endif
+      }
+      \endcode
+      */
+    typedef int opcFileTrimCallback(void *iocontext, opc_ofs_t new_size);
+
+     /**
+      Callback to flush a file. E.g. for a FILE * context this can be implemented as
+      \code
+      static int opcFileFlush(void *iocontext) {
+          return fflush((FILE*)iocontext);
+      }
+      \endcode
+      */
+    typedef int opcFileFlushCallback(void *iocontext);
+
+    /**
+      Represents a state of a file, i.e. file position (buf_pos) and error status (err).
+      */
+    typedef struct OPC_FILERAWSTATE_STRUCT {
+        opc_error_t err;
+        opc_ofs_t   buf_pos; // current pos in file
+    } opcFileRawState;
+
+    /**
+     File IO context.
+     */
+    typedef struct OPC_IO_STRUCT {
+        opcFileReadCallback *_ioread;
+        opcFileWriteCallback *_iowrite;
+        opcFileCloseCallback *_ioclose;
+        opcFileSeekCallback *_ioseek;
+        opcFileTrimCallback *_iotrim;
+        opcFileFlushCallback *_ioflush;
+        void *iocontext;
+        int flags;
+        opcFileRawState state;
+        opc_ofs_t file_size;
+    } opcIO_t;
+
+    /**
+      Initialize an IO context.
+      */
+    opc_error_t opcFileInitIO(opcIO_t *io,
+                              opcFileReadCallback *ioread,
+                              opcFileWriteCallback *iowrite,
+                              opcFileCloseCallback *ioclose,
+                              opcFileSeekCallback *ioseek,
+                              opcFileTrimCallback *iotrim,
+                              opcFileFlushCallback *ioflush,
+                              void *iocontext,
+                              pofs_t file_size,
+                              int flags);
+
+    /**
+      Initialize an IO context for a file.
+      */
+    opc_error_t opcFileInitIOFile(opcIO_t *io, const xmlChar *filename, int flags);
+
+    /**
+      Initialize an IO for memory.
+      \warning Currently supports READ-ONLY file access.
+      */
+    opc_error_t opcFileInitIOMemory(opcIO_t *io, const opc_uint8_t *data, opc_uint32_t data_len, int flags);
+
+    /**
+      Cleanup an IO context, i.e. release all system resources.
+      */
+    opc_error_t opcFileCleanupIO(opcIO_t *io);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_FILE_H */

include/opc/helper.h

@@ -0,0 +1,60 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/helper.h
+  Contains helper functions for the opc module.
+*/
+#include <opc/config.h>
+
+#ifndef OPC_HELPER_H
+#define OPC_HELPER_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+
+    /**
+      Constructs a segment name.
+      */
+    opc_uint16_t opcHelperAssembleSegmentName(char *out, opc_uint16_t out_size, const xmlChar *name, opc_uint32_t segment_number, opc_uint32_t next_segment_id, opc_bool_t rels_segment, opc_uint16_t *out_max);
+
+    /**
+      Splits a filename into the segment informations.
+      */
+    opc_error_t opcHelperSplitFilename(opc_uint8_t *filename, opc_uint32_t filename_length, opc_uint32_t *segment_number, opc_bool_t *last_segment, opc_bool_t *rel_segment);
+
+#endif /* OPC_HELPER_H */

include/opc/inputstream.h

@@ -0,0 +1,74 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/inputstream.h
+ 
+ */
+#include <opc/config.h>
+
+#ifndef OPC_INPUTSTREAM_H
+#define OPC_INPUTSTREAM_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+    /**
+      Internal type which represents a binary input stream.
+      */
+    typedef struct OPC_CONTAINER_INPUTSTREAM_STRUCT opcContainerInputStream;
+
+    /**
+      Opens the part \c name of the \c container for reading.
+      */
+    opcContainerInputStream* opcContainerOpenInputStream(opcContainer *container, const xmlChar *name);
+
+    /**
+     Reads maximal \c buffer_len bytes from the input \c stream to \c buffer. 
+     \return The number of byes read or "0" in case of an error or end-of-stream.
+     */
+    opc_uint32_t opcContainerReadInputStream(opcContainerInputStream* stream, opc_uint8_t *buffer, opc_uint32_t buffer_len);
+
+    /**
+      Closes the input stream and releases all system resources.
+      */
+    opc_error_t opcContainerCloseInputStream(opcContainerInputStream* stream);
+
+    /**
+      Returns the type of compression used for the stream.
+      */
+    opcCompressionOption_t opcContainerGetInputStreamCompressionOption(opcContainerInputStream* stream);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_INPUTSTREAM_H */

include/opc/opc.h

@@ -0,0 +1,73 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/opc.h
+ The opc module contains the basic library functions.
+*/
+#include <opc/config.h>
+#include <opc/container.h>
+#include <opc/part.h>
+#include <opc/relation.h>
+#include <opc/inputstream.h>
+#include <opc/outputstream.h>
+#include <opc/zip.h>
+#include <opc/xmlreader.h>
+#include <opc/xmlwriter.h>
+#include <opc/properties.h>
+
+#ifndef OPC_OPC_H
+#define OPC_OPC_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+    
+    /**
+     * Initialize libopc.
+     * Sample:
+     * \include opc_helloworld.c
+     * @return Non-zero if successful.
+     */
+    opc_error_t opcInitLibrary();
+
+    /**
+     * Free libopc. Clean up all resources.
+     * @return Non-zero if successful.
+     * \see opcInitLibrary.
+     */
+    opc_error_t opcFreeLibrary();
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_OPC_H */

include/opc/outputstream.h

@@ -0,0 +1,71 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/outputstream.h
+ 
+ */
+#include <opc/config.h>
+
+#ifndef OPC_OUTPUTSTREAM_H
+#define OPC_OUTPUTSTREAM_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+    /**
+      Internal type which represents a binary output stream.
+      */
+    typedef struct OPC_CONTAINER_OUTPUTSTREAM_STRUCT opcContainerOutputStream;
+
+    /** 
+      Open the part \c name or writing in \c container with compression \c compression_option.
+      \note Make sure the part exists! 
+      \see opcPartCreate.
+      */
+    opcContainerOutputStream* opcContainerCreateOutputStream(opcContainer *container, const xmlChar *name, opcCompressionOption_t compression_option);
+
+    /**
+      Write \c buffer_len bytes from \c buffer to \c stream. 
+      \return Returns the number of bytes written.
+      */
+    opc_uint32_t opcContainerWriteOutputStream(opcContainerOutputStream* stream, const opc_uint8_t *buffer, opc_uint32_t buffer_len);
+
+    /**
+      Close the \c stream and free all associated resources.
+      */
+    opc_error_t opcContainerCloseOutputStream(opcContainerOutputStream* stream);
+        
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_OUTPUTSTREAM_H */

include/opc/part.h

@@ -0,0 +1,118 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/part.h
+ 
+ */
+#include <opc/config.h>
+
+#ifndef OPC_PART_H
+#define OPC_PART_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+    /**
+     Handle to an OPC part created by \ref opcPartOpen.
+     \see opcPartOpen.
+     */
+    typedef xmlChar* opcPart;
+
+/**
+  Represents an invalid (resp. NULL) part.
+  In releations OPC_PART_INVALID also represents the root part.
+  \hideinitializer
+  */
+#define OPC_PART_INVALID NULL
+
+    /**
+     Find a part in a \ container by \c absolutePath and/or \c type.
+     Currently no flags are supported.
+     */
+    opcPart opcPartFind(opcContainer *container, 
+                        const xmlChar *absolutePath, 
+                        const xmlChar *type,
+                        int flags);
+
+    /**
+     Creates a part in a \ container with \c absolutePath and \c type.
+     Currently no flags are supported.
+     */
+    opcPart opcPartCreate(opcContainer *container, 
+                          const xmlChar *absolutePath, 
+                          const xmlChar *type,
+                          int flags);
+
+    /**
+      Returns the type of the container.
+      The string is interned and must not be freed.
+      */
+    const xmlChar *opcPartGetType(opcContainer *c, opcPart part);
+
+    /**
+      Returns the type of the container. 
+      If \c override_only then the return value will be NULL for parts not having an override type.
+      The string is interned and must not be freed.
+      */
+    const xmlChar *opcPartGetTypeEx(opcContainer *c, opcPart part, opc_bool_t override_only);
+
+    /**
+     Deleted that part \c absolutePath in the \c container.
+     */
+    opc_error_t opcPartDelete(opcContainer *container, const xmlChar *absolutePath);
+
+    /**
+      Get the first part.
+      \code
+      for(opcPart part=opcPartGetFirst(c);OPC_PART_INVALID!=part;part=opcPartGetNext(c, part)) {
+        printf("%s; \n", part, opcPartGetType(c, part));
+      }
+      \endcode 
+      */
+    opcPart opcPartGetFirst(opcContainer *container);
+
+    /**
+     Get the next part.
+     \see opcPartGetFirst
+      */
+    opcPart opcPartGetNext(opcContainer *container, opcPart part);
+
+    /**
+      Returns the size in bytes of the \c part.
+      */
+    opc_ofs_t opcPartGetSize(opcContainer *c, opcPart part);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_PART_H */

include/opc/properties.h

@@ -0,0 +1,121 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/properties.h
+ 
+ */
+#include <opc/config.h>
+#include <opc/container.h>
+
+#ifndef OPC_PROPERTIES_H
+#define OPC_PROPERTIES_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+
+    /** 
+      Represents a simple Dublin Core type.
+      */
+    typedef struct OPC_DC_SIMPLE_TYPE {
+        xmlChar *str;
+        xmlChar *lang;
+    } opcDCSimpleType_t;
+
+    /** 
+      Represents the core properties of an OPC container.
+      */
+    typedef struct OPC_PROPERTIES_STRUCT {
+        xmlChar *category;                /* xsd:string     */
+        xmlChar *contentStatus;           /* xsd:string     */
+        xmlChar *created;                 /* dc:date        */
+        opcDCSimpleType_t creator;        /* dc:any         */
+        opcDCSimpleType_t description;    /* dc:any         */
+        opcDCSimpleType_t identifier;     /* dc:any         */
+        opcDCSimpleType_t *keyword_array; /* cp:CT_Keywords */
+        opc_uint32_t keyword_items;
+        opcDCSimpleType_t language;       /* dc:any         */
+        xmlChar *lastModifiedBy;          /* xsd:string     */
+        xmlChar *lastPrinted;             /* xsd:dateTime   */
+        xmlChar *modified;                /* dc:date        */
+        xmlChar *revision;                /* xsd:string     */
+        opcDCSimpleType_t subject;        /* dc:any         */
+        opcDCSimpleType_t title;          /* dc:any         */
+        xmlChar *version;                 /* xsd:string     */
+    } opcProperties_t;
+
+    /**
+      Initialize the core properties \c cp.
+      \see opcCorePropertiesSetString
+      */
+    opc_error_t opcCorePropertiesInit(opcProperties_t *cp);
+
+    /**
+      Cleanup the core properties \c cp, i.e. release all resources.
+      \see opcCorePropertiesSetString
+      */
+    opc_error_t opcCorePropertiesCleanup(opcProperties_t *cp);
+
+    /**
+      Rease the core properties \c cp from the container \c.
+      */
+    opc_error_t opcCorePropertiesRead(opcProperties_t *cp, opcContainer *c);
+
+
+    /**
+      Write/Update the core properties \c cp in the container \c.
+      */
+    opc_error_t opcCorePropertiesWrite(opcProperties_t *cp, opcContainer *c);
+
+    /**
+      Update a string in the core properties the right way.
+      \code
+      opcProperties_t cp;
+      opcCorePropertiesInit(&cp);
+      opcCorePropertiesSetString(&cp.revision, "1");
+      opcCorePropertiesSetStringLang(&cp.creator, "Florian Reuter", NULL);
+      opcCorePropertiesCleanup(&cp);
+      \endcode
+      */
+    opc_error_t opcCorePropertiesSetString(xmlChar **prop, const xmlChar *str);
+
+    /** 
+      Update a core properties the right way.
+      \see opcCorePropertiesSetString
+      */
+    opc_error_t opcCorePropertiesSetStringLang(opcDCSimpleType_t *prop, const xmlChar *str, const xmlChar *lang);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_PROPERTIES_H */

include/opc/relation.h

@@ -0,0 +1,140 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/relation.h
+ 
+ */
+#include <opc/config.h>
+
+#ifndef OPC_RELATION_H
+#define OPC_RELATION_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+
+    /**
+     Indentifier for an OPC relation.
+     */
+    typedef opc_uint32_t opcRelation;
+
+/**
+  Constant which represents an invalid relation.
+*/
+#define OPC_RELATION_INVALID (-1)
+
+    /**
+      Find a relation originating from \c part in \c container with \c relationId and/or \c mimeType.
+      If \c part is OPC_PART_INVALID then part represents the root part.
+      @param[in] relationId The relationId (e.g. "rId1") or NULL.
+      @param[in] mimeType The mimeType or NULL.
+      */
+    opcRelation opcRelationFind(opcContainer *container, opcPart part, const xmlChar *relationId, const xmlChar *mimeType);
+
+    /**
+      Deleted the relation from the container.
+      \see opcRelationFind.
+      */
+    opc_error_t opcRelationDelete(opcContainer *container, opcPart part, const xmlChar *relationId, const xmlChar *mimeType);
+
+    /**
+      Returns the first relation.
+      The following code will dump all relations:
+      \code
+        for(opcPart part=opcPartGetFirst(c);OPC_PART_INVALID!=part;part=opcPartGetNext(c, part)) {
+           for(opcRelation rel=opcRelationFirst(part, c);
+               OPC_PART_INVALID!=rel;
+               rel=opcRelationNext(c, rel)) {
+               opcPart internal_target=opcRelationGetInternalTarget(c, part, rel);
+               const xmlChar *external_target=opcRelationGetExternalTarget(c, part, rel);
+               const xmlChar *target=(NULL!=internal_target?internal_target:external_target);
+               const xmlChar *prefix=NULL;
+               opc_uint32_t counter=-1;
+               const xmlChar *type=NULL;
+               opcRelationGetInformation(c, part, rel, &prefix, &counter, &type);        
+               if (-1==counter) { // no counter after prefix
+                  printf("%s;%s;%s;%s\n", part, prefix, target, type);
+               } else {
+                  printf("%s;%s%i;%s;%s\n", part, prefix, counter, target, type);
+               }
+           }
+        }
+      \endcode
+      */
+    opcRelation opcRelationFirst(opcContainer *container, opcPart part);
+
+    /**
+      \see opcRelationFirst
+      */
+    opcRelation opcRelationNext(opcContainer *container, opcPart part, opcRelation relation);
+    
+    /**
+      Returns the internal target.
+      \note To test for an external target use opcRelationGetExternalTarget.
+      \see opcRelationGetExternalTarget
+      */
+    opcPart opcRelationGetInternalTarget(opcContainer *container, opcPart part, opcRelation relation);
+
+    /**
+      Returns the external target or NULL if it is an internal target.
+      The string is interned. Must not be freed.
+      \see opcRelationGetExternalTarget
+      */
+    const xmlChar *opcRelationGetExternalTarget(opcContainer *container, opcPart part, opcRelation relation);
+
+    /**
+      Returns the relations type.
+      The string is interned. Must not be freed.
+      */
+    const xmlChar *opcRelationGetType(opcContainer *container, opcPart part, opcRelation relation);
+
+    /** 
+      Get information about a relation.
+      \see opcRelationFirst
+      */
+    void opcRelationGetInformation(opcContainer *container, opcPart part, opcRelation relation, const xmlChar **prefix, opc_uint32_t *counter, const xmlChar **type);
+
+    /** 
+      Add a relation to \c container from \c src part to \c dest part with id \c rid and type \c type.
+      */
+    opc_uint32_t opcRelationAdd(opcContainer *container, opcPart src, const xmlChar *rid, opcPart dest, const xmlChar *type);
+
+    /** 
+      Add an external relation to \c container from \c src part to \c target URL with id \c rid and type \c type.
+      */
+    opc_uint32_t opcRelationAddExternal(opcContainer *container, opcPart src, const xmlChar *rid, const xmlChar *target, const xmlChar *type);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+
+#endif /* OPC_RELATION_H */

include/opc/xmlreader.h

@@ -0,0 +1,69 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/xmlreader.h
+ 
+ */
+
+#ifndef OPC_XMLREADER_H
+#define OPC_XMLREADER_H
+
+#include <opc/config.h>
+#include <libxml/xmlreader.h>
+#include <mce/textreader.h>
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+
+    /** 
+      Open an MCE reader for \c partName. Parameters \c URL, \c encoding and \c options will be passed unmodified to 
+      http://xmlsoft.org/html/libxml-xmlreader.html#xmlReaderForIO and they can we NULL, NULL, 0.
+      \note Make sure the part exists.
+      \see opcPartFind
+      */
+    opc_error_t opcXmlReaderOpen(opcContainer *container, mceTextReader_t *mceTextReader, const xmlChar *partName, const char * URL, const char * encoding, int options);
+
+    /**
+      Returns an libxml DOM document. Parameters \c URL, \c encoding and \c options will be passed unmodified to 
+      http://xmlsoft.org/html/libxml-parser.html#xmlReadIO and they can we NULL, NULL, 0.
+      \note Make sure the part exists.
+      \see opcPartFind
+      */
+    xmlDocPtr opcXmlReaderReadDoc(opcContainer *container, const xmlChar *partName, const char * URL, const char * encoding, int options);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_XMLREADER_H */

include/opc/xmlwriter.h

@@ -0,0 +1,57 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/xmlwriter.h
+ 
+ */
+#include <opc/config.h>
+#include <mce/textwriter.h>
+
+#ifndef OPC_XMLWRITER_H
+#define OPC_XMLWRITER_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+
+    /**
+      Create an MCE text writer for \c part in \c container with compression \c compression_option.
+      \note Make sure the part exists.
+      \see opcPartFind
+      */
+    mceTextWriter *mceTextWriterOpen(opcContainer *c, opcPart part, opcCompressionOption_t compression_option);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_XMLWRITER_H */

include/opc/zip.h

@@ -0,0 +1,255 @@
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+/** @file opc/zip.h
+  The ZIP file backend of an OPC container.
+ */
+#include <opc/config.h>
+#include <opc/file.h>
+#include <opc/container.h>
+
+#ifndef OPC_ZIP_H
+#define OPC_ZIP_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif    
+
+    /**
+     Default growth hint of an OPC stream.
+     */
+    #define OPC_DEFAULT_GROWTH_HINT 512
+
+    /**
+     Handle to a ZIP archive.
+     \see internal.h
+     */
+    typedef struct OPC_ZIP_STRUCT opcZip;
+
+    /**
+      Handle to a raw ZIP input stream.
+     \see internal.h
+      */
+    typedef struct OPC_ZIPINPUTSTREAM_STRUCT opcZipInputStream;
+
+    /**
+      Handle to a raw ZIP output stream.
+     \see internal.h
+      */
+    typedef struct OPC_ZIPOUTPUTSTREAM_STRUCT opcZipOutputStream;
+
+    /**
+     Holds all information of a ZIP segment.
+     */
+    typedef struct OPC_ZIP_SEGMENT_INFO_STRUCT {
+        xmlChar name[OPC_MAX_PATH]; 
+        opc_uint32_t name_len;
+        opc_uint32_t segment_number;
+        opc_bool_t   last_segment;
+        opc_bool_t   rels_segment;
+        opc_uint32_t header_size;
+        opc_uint32_t min_header_size;
+        opc_uint32_t trailing_bytes;
+        opc_uint32_t compressed_size;
+        opc_uint32_t uncompressed_size;
+        opc_uint16_t bit_flag;
+        opc_uint32_t data_crc;
+        opc_uint16_t compression_method;
+        opc_ofs_t    stream_ofs;
+        opc_uint16_t growth_hint;
+    } opcZipSegmentInfo_t;
+
+    /**
+      \see opcZipLoader
+      */
+    typedef int opcZipLoaderOpenCallback(void *iocontext);
+    /**
+      \see opcZipLoader
+      */
+    typedef int opcZipLoaderSkipCallback(void *iocontext);
+    /**
+      \see opcZipLoader
+      */
+    typedef int opcZipLoaderReadCallback(void *iocontext, char *buffer, int len);
+    /**
+      \see opcZipLoader
+      */
+    typedef int opcZipLoaderCloseCallback(void *iocontext);
+
+    /**
+      \see opcZipLoader
+      */
+    typedef opc_error_t (opcZipLoaderSegmentCallback_t)(void *iocontext, void *userctx, opcZipSegmentInfo_t *info, opcZipLoaderOpenCallback *open, opcZipLoaderReadCallback *read, opcZipLoaderCloseCallback *close, opcZipLoaderSkipCallback *skip);
+
+    /**
+      Walks every segment in a ZIP archive and calls the \c segmentCallback callback method.
+      The implementer \c segmentCallback method must then eiher use the passed \c open, \c read and \c close methods
+      to read the stream or the passed \c skip methods to skip the stream.
+      This method can be used to e.g. read ZIP file in stream mode.
+      */
+    opc_error_t opcZipLoader(opcIO_t *io, void *userctx, opcZipLoaderSegmentCallback_t *segmentCallback);
+
+    /**
+      \see opcZipClose
+     */
+    typedef opc_error_t (opcZipSegmentReleaseCallback)(opcZip *zip, opc_uint32_t segment_id);
+
+    /** 
+     Closes the ZIP archive \c zip and will call \c releaseCallback for every segment to give the implementer a chance
+     to free user resources.
+     */
+    void opcZipClose(opcZip *zip, opcZipSegmentReleaseCallback* releaseCallback);
+
+    /**
+      Creates an empty ZIP archive with the given \c io.
+      */
+    opcZip *opcZipCreate(opcIO_t *io);
+
+    /**
+      Commits all buffers and writes the ZIP archives local header directories.
+      if \c trim is true then padding bytes will be removed, i.e. the ZIP file size fill be minimalized.
+     */
+    opc_error_t opcZipCommit(opcZip *zip, opc_bool_t trim);
+
+    /**
+      Garbage collection on the passed \c zip archive. This will e.g. make deleted files available as free space.
+      */
+    opc_error_t opcZipGC(opcZip *zip);
+
+    /**
+      Load segment information into \c info.
+      If \c rels_segment is -1 then load the info for part with name \c partName.
+      Otherwise load the segment information for the ".rels." segment of \c partName.
+      \return Returns the segment_id.
+      */
+    opc_uint32_t opcZipLoadSegment(opcZip *zip, const xmlChar *partName, opc_bool_t rels_segment, opcZipSegmentInfo_t *info);
+
+    /**
+      Create a segment with the given parameters.
+      \return Returns the segment_id.
+      */
+    opc_uint32_t opcZipCreateSegment(opcZip *zip, 
+                                     const xmlChar *partName, 
+                                     opc_bool_t relsSegment, 
+                                     opc_uint32_t segment_size, 
+                                     opc_uint32_t growth_hint,
+                                     opc_uint16_t compression_method,
+                                     opc_uint16_t bit_flag);
+
+    /**
+      Creates an input stream for the segment with \c segment_id.
+      \see opcZipLoadSegment
+      \see opcZipCreateSegment
+      */
+    opcZipInputStream *opcZipOpenInputStream(opcZip *zip, opc_uint32_t segment_id);
+
+    /**
+     Free all resources of the input stream.
+     */
+    opc_error_t opcZipCloseInputStream(opcZip *zip, opcZipInputStream *stream);
+
+    /**
+     Read maximal \c buf_len bytes from the input stream into \buf. 
+     \return Returns the number of bytes read.
+     */
+    opc_uint32_t opcZipReadInputStream(opcZip *zip, opcZipInputStream *stream, opc_uint8_t *buf, opc_uint32_t buf_len);
+
+
+    /**
+      Creates an output stream for the segment with \c segment_id.
+      If \c *segment_id is -1 then a new segment will be created. 
+      Otherwise the segment with \c *segment_id will be overwritten.
+     */
+    opcZipOutputStream *opcZipCreateOutputStream(opcZip *zip, 
+                                             opc_uint32_t *segment_id, 
+                                             const xmlChar *partName, 
+                                             opc_bool_t relsSegment, 
+                                             opc_uint32_t segment_size, 
+                                             opc_uint32_t growth_hint,
+                                             opc_uint16_t compression_method,
+                                             opc_uint16_t bit_flag);
+
+    /**
+      Opens an existing ouput stream for reading.
+      The \c *segment_id will be set to -1 and reset on opcZipCloseOutputStream.
+      \see opcZipCloseOutputStream
+     */
+    opcZipOutputStream *opcZipOpenOutputStream(opcZip *zip, opc_uint32_t *segment_id);
+
+    /** 
+      Will close the stream and free all resources. Additionally the new segment id will be stored in \c *segment_id.
+      \see opcZipOpenOutputStream
+      */
+    opc_error_t opcZipCloseOutputStream(opcZip *zip, opcZipOutputStream *stream, opc_uint32_t *segment_id);
+
+    /**
+     Write \c buf_len bytes to \c buf. 
+     \return Returns the number of bytes written.
+     */
+    opc_uint32_t opcZipWriteOutputStream(opcZip *zip, opcZipOutputStream *stream, const opc_uint8_t *buf, opc_uint32_t buf_len);
+
+    /**
+     Returns the first segment id or -1.
+     Use the following code to iterarte through all segments.
+     \code 
+     for(opc_uint32_t segment_id=opcZipGetFirstSegmentId(zip);
+         -1!=segment_id;
+         segment_id=opcZipGetNextSegmentId(zip, segment_id) {
+        ...
+     }
+     \endcode
+     \see opcZipGetNextSegmentId
+     */
+    opc_uint32_t opcZipGetFirstSegmentId(opcZip *zip);
+
+    /**
+     Returns the next segment id or -1.
+     \see opcZipGetFirstSegmentId
+     */
+    opc_uint32_t opcZipGetNextSegmentId(opcZip *zip, opc_uint32_t segment_id);
+
+    /**
+     Returns info about the given segment id.
+     */
+    opc_error_t opcZipGetSegmentInfo(opcZip *zip, opc_uint32_t segment_id, const xmlChar **name, opc_bool_t *rels_segment, opc_uint32_t *crc);
+
+    /**
+     Marks a given segments as deleted.
+     \see opcZipGC
+     */
+    opc_bool_t opcZipSegmentDelete(opcZip *zip, opc_uint32_t *first_segment, opc_uint32_t *last_segment, opcZipSegmentReleaseCallback* releaseCallback);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif    
+        
+#endif /* OPC_ZIP_H */

include/plib/plib.h

@@ -0,0 +1,168 @@
+/* include/plib/plib.h.  Generated from plib.h by configure.  */
+/*
+ Copyright (c) 2010, Florian Reuter
+ All rights reserved.
+ 
+ Redistribution and use in source and binary forms, with or without 
+ modification, are permitted provided that the following conditions 
+ are met:
+ 
+ * Redistributions of source code must retain the above copyright 
+   notice, this list of conditions and the following disclaimer.
+ * Redistributions in binary form must reproduce the above copyright 
+   notice, this list of conditions and the following disclaimer in 
+   the documentation and/or other materials provided with the 
+   distribution.
+ * Neither the name of Florian Reuter nor the names of its contributors 
+   may be used to endorse or promote products derived from this 
+   software without specific prior written permission.
+ 
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE 
+ COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, 
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, 
+ BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
+ LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER 
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED 
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+ 
+*/
+#ifndef _PLIB_PLIB_H_
+#define _PLIB_PLIB_H_
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#define HAVE_STDINT_H 1
+#define HAVE_STDDEF_H 1
+#define HAVE_STDIO_H 1
+#define HAVE_STRING_H 1
+#define HAVE_LIMITS_H 1
+#define HAVE_STDLIB_H 1
+/* #undef HAVE_IO_H */
+#define HAVE_UNISTD_H 1
+#define HAVE_SYS_TYPES_H 1
+#define IS_CONFIGURED 1
+
+#if !defined(IS_CONFIGURED)
+#if defined(WIN32)
+#define HAVE_STRING_H 1
+#define HAVE_STDINT_H 1
+#define HAVE_LIMITS_H 1
+#define HAVE_STDDEF_H 1
+#define HAVE_STDIO_H 1
+#define HAVE_STDLIB_H 1
+#define HAVE_IO_H
+#define snprintf _snprintf
+#else
+#error "configure not executed and we are not on a win32 machine? please run configure or define WIN32 is you are on a WIN32 platform."
+#endif
+#endif
+
+#ifdef HAVE_STDDEF_H
+#include <stddef.h>
+typedef size_t pofs_t; // maximum file offset for eg. read write ops
+#else
+#error "system types can not be determined"
+#endif
+
+#ifdef HAVE_STDIO_H
+#include <stdio.h>
+#else
+#error "system io can not be determined"
+#endif
+
+#ifdef HAVE_STDINT_H
+#include <stdint.h>
+
+typedef int8_t pint8_t;
+typedef uint8_t puint8_t;
+
+typedef int16_t pint16_t;
+typedef uint16_t puint16_t;
+
+typedef int32_t pint32_t;
+typedef uint32_t puint32_t;
+
+typedef int64_t pint64_t;
+typedef uint64_t puint64_t;
+
+typedef int pbool_t;
+
+typedef size_t psize_t;
+
+// INTN_MAX, INTN_MIN, UINTN_MAX
+#else
+#error "system types can not be determined"
+#endif
+
+#ifdef HAVE_STRING_H
+#include <string.h>
+#endif
+
+#ifdef HAVE_LIMITS_H
+#include <limits.h>
+#define PUINT8_MAX UCHAR_MAX 
+#define PINT32_MAX INT_MAX 
+#define PINT32_MIN INT_MIN 
+#define PUINT32_MAX UINT_MAX 
+#define PUINT32_MIN 0 
+#define PUINT16_MAX	USHRT_MAX 
+#define PUINT16_MIN 0 
+#else
+#error "limits can not be determined"
+#endif
+
+#ifdef HAVE_STDLIB_H
+#include <stdlib.h>
+#endif
+
+#ifdef HAVE_IO_H
+#include <io.h>
+#endif
+
+#ifdef HAVE_UNISTD_H
+#include <unistd.h>
+#endif
+
+#ifdef HAVE_SYS_TYPES_H
+#include <sys/types.h>
+#endif
+
+/**
+ Converts an ASCII string to a xmlChar string. This only works for ASCII strings.
+ */
+#ifndef _X
+#define _X(s) BAD_CAST(s) 
+#endif
+
+
+/**
+ Converts an xmlChar string to an ASCII string. This only works for ASCII charsets.
+ */
+#ifndef _X2C
+#define _X2C(s) ((char*)(s))
+#endif
+
+
+#define PASSERT(e) assert(e)
+#ifdef NDEBUG
+#define PENSURE(e) (void)(e)
+#else
+#define PENSURE(e) assert(e)
+#endif
+#define PTRUE (0==0)
+#define PFALSE (0==1)
+
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif
+
+
+#endif /* _PLIB_PLIB_H_ */

mime.csv

@@ -252,7 +252,7 @@ text/html, acgi|htm|html|htmls|htx|shtml
 text/javascript, js
 text/mcf, mcf
 text/pascal, pas
-text/plain, com|cmd|conf|def|g|idc|list|lst|mar|sdml|text|txt|md|groovy|license|properties|desktop|ini|rst|cmake|ipynb|readme|less|lo|go|yml|d|cs|hpp|srt
+text/plain, com|cmd|conf|def|g|idc|list|lst|mar|sdml|text|txt|md|groovy|license|properties|desktop|ini|rst|cmake|ipynb|readme|less|lo|go|yml|d|cs|hpp|srt|nfo|sfv|m3u|csv|eml
 text/richtext, rt|rtf|rtx
 text/rtf,
 text/scriplet, wsc
@@ -410,4 +410,9 @@ text/PGP,
 audio/x-hx-aac-adts,
 application/x-chrome-extension,
 image/heic, heic
-image/x-gem,
+image/x-gem,
+application/x-lzma, lzma
+application/warc, warc
+application/x-lz4, lz4
+application/x-lzip, lz
+application/x-lzop, lzo

schema/mappings.json

@@ -7,25 +7,30 @@
     },
     "suggest-path": {
       "type": "completion",
-      "analyzer": "keyword"
+      "analyzer": "case_insensitive_kw_analyzer"
     },
     "mime": {
       "type": "keyword"
     },
     "videoc": {
-      "type": "keyword"
+      "type": "keyword",
+      "index": false
     },
     "audioc": {
-      "type": "keyword"
+      "type": "keyword",
+      "index": false
     },
     "duration": {
-      "type": "float"
+      "type": "float",
+      "index": false
     },
     "width": {
-      "type": "integer"
+      "type": "integer",
+      "index": false
     },
     "height": {
-      "type": "integer"
+      "type": "integer",
+      "index": false
     },
     "mtime": {
       "type": "integer"
@@ -70,6 +75,23 @@
       "analyzer": "my_nGram",
       "type": "text"
     },
+    "_keyword.*":  {
+      "type": "keyword"
+    },
+    "_text.*":  {
+      "analyzer": "content_analyzer",
+      "type": "text",
+      "fields": {
+        "nGram": {
+          "type": "text",
+          "analyzer": "my_nGram"
+        }
+      }
+    },
+    "_url": {
+      "type": "keyword",
+      "index": false
+    },
     "content": {
       "analyzer": "content_analyzer",
       "type": "text",
@@ -80,6 +102,9 @@
           "analyzer": "my_nGram"
         }
       }
+    },
+    "tag": {
+      "type": "keyword"
     }
   }
 }

schema/settings.json

@@ -21,6 +21,12 @@
           "lowercase"
         ]
       },
+      "case_insensitive_kw_analyzer": {
+        "tokenizer": "keyword",
+        "filter": [
+          "lowercase"
+        ]
+      },
       "my_nGram": {
         "tokenizer": "my_nGram_tokenizer",
         "filter": [

scripting/README.md

@@ -0,0 +1,117 @@
+## User scripts
+
+*This document is under construction, more in-depth guide coming soon*
+
+During the `index` step, you can use the `--script-file <script>` option to
+modify documents or add user tags. This option is mainly used to
+implement automatic tagging based on file attributes.
+
+The scripting language used 
+([Painless Scripting Language](https://www.elastic.co/guide/en/elasticsearch/painless/7.4/index.html)) 
+is very similar to Java, but you should be able to create user scripts
+without programming experience at all if you're somewhat familiar with
+regex.
+
+This is the base structure of the documents we're working with:
+```json
+{
+  "_id": "e171405c-fdb5-4feb-bb32-82637bc32084",
+  "_index": "sist2",
+  "_type": "_doc",
+  "_source": {
+    "index": "206b3050-e821-421a-891d-12fcf6c2db0d",
+    "mime": "application/json",
+    "size": 1799,
+    "mtime": 1545443685,
+    "extension": "md",
+    "name": "README",
+    "path": "sist2/scripting",
+    "content": "..."
+  }
+}
+```
+
+**Example script**
+
+This script checks if the `genre` attribute exists, if it does
+it adds the `genre.<genre>` tag. 
+```Java
+ArrayList tags = ctx._source.tag = new ArrayList();
+
+if (ctx._source?.genre != null) {
+    tags.add("genre." + ctx._source.genre.toLowerCase())
+}
+```
+
+You can use `.` to create a hierarchical tag tree:
+
+![scripting/genre_example](genre_example.png)
+
+
+To use regular expressions, you need to add this line in `/etc/elasticsearch/elasticsearch.yml`
+```yaml
+script.painless.regex.enabled: true
+```
+Or, if you're using docker add `-e "script.painless.regex.enabled=true"`
+
+### Examples
+
+If `(20XX)` is in the file name, add the `year.<year>` tag:
+```Java
+ArrayList tags = ctx._source.tag = new ArrayList();
+
+Matcher m = /[\(\.+](20[0-9]{2})[\)\.+]/.matcher(ctx._source.name);
+if (m.find()) {
+    tags.add("year." + m.group(1))
+}
+```
+
+Use default *Calibre* folder structure to infer author.
+```Java
+ArrayList tags = ctx._source.tag = new ArrayList();
+
+// We expect the book path to look like this:
+//  /path/to/Calibre Library/Author/Title/Title - Author.pdf
+
+if (ctx._source.name.contains("-") && ctx._source.extension == "pdf") {
+    String[] names = ctx._source.name.splitOnToken('-');
+    tags.add("author." + names[1].strip());
+}
+```
+
+If the file matches a specific pattern `AAAA-000 fName1 lName1, <fName2 lName2>...`, add the `actress.<actress>` and 
+`studio.<studio>` tag:
+```Java
+ArrayList tags = ctx._source.tag = new ArrayList();
+
+Matcher m = /([A-Z]{4})-[0-9]{3} (.*)/.matcher(ctx._source.name);
+if (m.find()) {
+    tags.add("studio." + m.group(1));
+
+    // Take the matched group (.*), and add a tag for
+    //  each name, separated by comma
+    for (String name : m.group(2).splitOnToken(',')) {
+        tags.add("actress." + name);
+    }
+}
+```
+
+Set the name of the last folder (`/path/to/<studio>/file.mp4`) to `studio.<studio>` tag
+```Java
+ArrayList tags = ctx._source.tag = new ArrayList();
+
+if (ctx._source.path != "") {
+    String[] names = ctx._source.path.splitOnToken('/');
+    tags.add("studio." + names[names.length-1]);
+}
+```
+
+Set the name of the last folder (`/path/to/<studio>/file.mp4`) to `studio.<studio>` tag
+```Java
+ArrayList tags = ctx._source.tag = new ArrayList();
+
+if (ctx._source.path != "") {
+    String[] names = ctx._source.path.splitOnToken('/');
+    tags.add("studio." + names[names.length-1]);
+}
+```

scripts/get_static_libs.sh

@@ -1,8 +1,11 @@
 #!/usr/bin/env bash
+
+THREADS=$(nproc)
+
 cd lib
 
 cd mupdf
-USE_SYSTEM_HARFBUZZ=yes USE_SYSTEM_OPENJPEG=yes HAVE_X11=no HAVE_GLUT=no make -j 4
+CFLAGS=-fPIC make USE_SYSTEM_HARFBUZZ=yes USE_SYSTEM_OPENJPEG=yes HAVE_X11=no HAVE_GLUT=no -j $THREADS
 cd ..
 
 mv mupdf/build/release/libmupdf.a .
@@ -10,17 +13,16 @@ mv mupdf/build/release/libmupdf-third.a .
 
 # openjp2
 cd openjpeg
-#cmake . -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS="-O3 -march=native -DNDEBUG"
-cmake . -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS="-O3"
-make -j 4
+cmake . -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS="-O3 -march=native -DNDEBUG -fPIC"
+make -j $THREADS
 cd ..
 mv openjpeg/bin/libopenjp2.a .
 
 # harfbuzz
 cd harfbuzz
 ./autogen.sh
-./configure --disable-shared --enable-static
-make -j 4
+CFLAGS=-fPIC ./configure --disable-shared --enable-static
+make -j $THREADS
 cd ..
 mv harfbuzz/src/.libs/libharfbuzz.a .
 
@@ -30,8 +32,9 @@ cd ffmpeg
  --disable-ffprobe --disable-doc\
  --disable-manpages --disable-postproc --disable-avfilter \
  --disable-alsa --disable-lzma --disable-xlib --disable-debug\
- --disable-vdpau --disable-vaapi --disable-sdl2 --disable-network
-make -j 4
+ --disable-vdpau --disable-vaapi --disable-sdl2 --disable-network\
+ --extra-cflags=-fPIC
+make -j $THREADS
 cd ..
 
 mv ffmpeg/libavcodec/libavcodec.a .
@@ -48,26 +51,78 @@ cmake -DONION_USE_SSL=false -DONION_USE_PAM=false -DONION_USE_PNG=false -DONION_
 -DONION_USE_JPEG=false -DONION_USE_XML2=false -DONION_USE_SYSTEMD=false -DONION_USE_SQLITE3=false \
 -DONION_USE_REDIS=false -DONION_USE_GC=false -DONION_USE_TESTS=false -DONION_EXAMPLES=false \
 -DONION_USE_BINDINGS_CPP=false ..
-make -j 4
+make -j $THREADS
 cd ../..
 
 mv onion/build/src/onion/libonion_static.a .
 
 #bzip2
-git clone https://github.com/enthought/bzip2-1.0.6
 cd bzip2-1.0.6
-make -j 4
+make -j $THREADS
 cd ..
 mv bzip2-1.0.6/libbz2.a .
 
 # magic
-git clone https://github.com/threatstack/libmagic
 cd libmagic
 ./autogen.sh
 ./configure --enable-static --disable-shared
-make -j 4
+make -j $THREADS
 cd ..
 mv libmagic/src/.libs/libmagic.a .
 
+# tesseract
+cd tesseract
+mkdir build
+cd build
+cmake -DSTATIC=on -DBUILD_TRAINING_TOOLS=off -DBUILD_TESTS=off -DCMAKE_BUILD_TYPE=Release \
+  -DCMAKE_CXX_FLAGS="-fPIC"  ..
+make -j $THREADS
+cd ../..
+mv tesseract/build/libtesseract.a .
 
+# leptonica
+cd leptonica
+./autogen.sh
+CFLAGS="-fPIC" ./configure --without-zlib --without-jpeg --without-giflib \
+  --without-giflib --without-libwebp --without-libwebpmux --without-libopenjpeg \
+  --enable-static --disable-shared
+make -j $THREADS
 cd ..
+mv leptonica/src/.libs/liblept.a .
+
+# tiff
+cd libtiff
+./autogen.sh
+CFLAGS="-fPIC" CXXFLAGS="-fPIC" CXX_FLAGS="-fPIC" ./configure --enable-static --disable-shared --disable-lzw --disable-jpeg --disable-webp \
+  --disable-lzma --disable-zstd --disable-jbig
+make -j $THREADS
+cd ..
+mv libtiff/libtiff/.libs/libtiff.a .
+
+# png
+cd libpng
+CFLAGS="-fPIC" ./configure --enable-static --disable-shared
+make -j $THREADS
+cd ..
+mv libpng/.libs/libpng16.a .
+
+# openssl...
+git clone --depth 1 -b OpenSSL_1_1_0-stable https://github.com/openssl/openssl
+cd openssl
+./config --prefix=$(pwd)/../ssl
+make depend
+make -j $THREADS
+make install
+cd ..
+mv ./openssl/libcrypto.a ./openssl/libssl.a .
+
+# curl
+wget -nc https://curl.haxx.se/download/curl-7.68.0.tar.gz
+tar -xzf curl-7.68.0.tar.gz
+cd curl-7.68.0
+./configure --disable-ldap --disable-ldaps --without-librtmp --disable-rtsp --disable-crypto-auth \
+  --disable-smtp --without-libidn2 --without-nghttp2 --without-brotli --enable-static --disable-shared \
+  --without-libpsl --with-ssl=$(pwd)/../ssl
+make -j $THREADS
+cd ..
+mv curl-7.68.0/lib/.libs/libcurl.a .

scripts/get_static_libs_freebsd.sh

@@ -1,60 +0,0 @@
-#!/usr/bin/env bash
-cd lib
-
-# mupdf
-cd mupdf
-HAVE_X11=no HAVE_GLUT=no gmake -j 4
-cd ..
-
-mv mupdf/build/release/libmupdf.a .
-mv mupdf/build/release/libmupdf-third.a .
-
-# openjp2
-cd openjpeg
-#cmake . -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS="-O3 -march=native -DNDEBUG"
-cmake . -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS="-O3"
-gmake -j 4
-cd ..
-mv openjpeg/bin/libopenjp2.a .
-
-# harfbuzz
-cd harfbuzz
-./autogen.sh
-./configure --disable-shared --enable-static
-gmake -j 4
-cd ..
-mv harfbuzz/src/.libs/libharfbuzz.a .
-
-# ffmpeg
-cd ffmpeg
-./configure --disable-shared --enable-static --disable-ffmpeg --disable-ffplay \
- --disable-ffprobe --disable-doc\
- --disable-manpages --disable-postproc --disable-avfilter \
- --disable-alsa --disable-lzma --disable-xlib --disable-debug\
- --disable-vdpau --disable-vaapi --disable-sdl2 --disable-network
-gmake -j 4
-cd ..
-
-mv ffmpeg/libavcodec/libavcodec.a .
-mv ffmpeg/libavformat/libavformat.a .
-mv ffmpeg/libavutil/libavutil.a .
-mv ffmpeg/libswresample/libswresample.a .
-mv ffmpeg/libswscale/libswscale.a .
-
-#bzip2
-git clone https://github.com/enthought/bzip2-1.0.6
-cd bzip2-1.0.6
-make -j 4
-cd ..
-mv bzip2-1.0.6/libbz2.a .
-
-# magic
-git clone https://github.com/threatstack/libmagic
-cd libmagic
-./autogen.sh
-./configure --enable-static --disable-shared
-make -j 4
-cd ..
-mv libmagic/src/.libs/libmagic.a .
-
-cd ..

scripts/mime.py

@@ -34,6 +34,34 @@ font = (
     "font/woff2"
 )
 
+# Archive "formats"
+archive = (
+    "application/x-tar",
+    "application/zip",
+    "application/x-rar",
+    "application/x-arc",
+    "application/x-warc",
+    "application/x-7z-compressed",
+)
+
+# Archive "filters"
+arc_filter = (
+    "application/gzip",
+    "application/x-bzip2",
+    "application/x-xz",
+    "application/x-zstd",
+    "application/x-lzma",
+    "application/x-lz4",
+    "application/x-lzip",
+    "application/x-lzop",
+)
+
+doc = (
+    "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+    "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+    "application/vnd.openxmlformats-officedocument.presentationml.presentation"
+)
+
 cnt = 1
 
 
@@ -48,6 +76,12 @@ def mime_id(mime):
         mime_id += " | 0x40000000"
     elif mime in font:
         mime_id += " | 0x20000000"
+    elif mime in archive:
+        mime_id += " | 0x10000000"
+    elif mime in arc_filter:
+        mime_id += " | 0x08000000"
+    elif mime in doc:
+        mime_id += " | 0x04000000"
     elif mime == "application/x-empty":
         return "1"
     return mime_id

src/cli.c

@@ -1,22 +1,59 @@
 #include "cli.h"
+#include "ctx.h"
+
+#include <tesseract/capi.h>
 
 #define DEFAULT_OUTPUT "index.sist2/"
-#define DEFAULT_CONTENT_SIZE 4096
-#define DEFAULT_QUALITY 15
-#define DEFAULT_SIZE 200
+#define DEFAULT_CONTENT_SIZE 32768
+#define DEFAULT_QUALITY 5
+#define DEFAULT_SIZE 500
 #define DEFAULT_REWRITE_URL ""
 
 #define DEFAULT_ES_URL "http://localhost:9200"
+#define DEFAULT_BATCH_SIZE 100
 
 #define DEFAULT_BIND_ADDR "localhost"
 #define DEFAULT_PORT "4090"
 
+const char* TESS_DATAPATHS[] = {
+        "/usr/share/tessdata/",
+        "/usr/share/tesseract-ocr/tessdata/",
+        "./",
+        NULL
+};
+
 
 scan_args_t *scan_args_create() {
     scan_args_t *args = calloc(sizeof(scan_args_t), 1);
+
+    args->depth = -1;
+
     return args;
 }
 
+void scan_args_destroy(scan_args_t *args) {
+    if (args->name != NULL) {
+        free(args->name);
+    }
+    if (args->path != NULL) {
+        free(args->path);
+    }
+    if (args->output != NULL) {
+        free(args->output);
+    }
+    free(args);
+}
+
+void index_args_destroy(index_args_t *args) {
+    //todo
+    free(args);
+}
+
+void web_args_destroy(web_args_t *args) {
+    //todo
+    free(args);
+}
+
 int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
     if (argc < 2) {
         fprintf(stderr, "Required positional argument: PATH.\n");
@@ -25,7 +62,7 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
 
     char *abs_path = abspath(argv[1]);
     if (abs_path == NULL) {
-        fprintf(stderr, "File not found: %s", argv[1]);
+        fprintf(stderr, "File not found: %s\n", argv[1]);
         return 1;
     } else {
         args->path = abs_path;
@@ -34,7 +71,7 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
     if (args->incremental != NULL) {
         abs_path = abspath(args->incremental);
         if (abs_path == NULL) {
-            fprintf(stderr, "File not found: %s", args->incremental);
+            fprintf(stderr, "File not found: %s\n", args->incremental);
             return 1;
         }
     }
@@ -48,16 +85,13 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
 
     if (args->size == 0) {
         args->size = DEFAULT_SIZE;
-    } else if (args->size <= 0) {
-        fprintf(stderr, "Invalid size: %d\n", args->size);
+    } else if (args->size > 0 && args->size < 32) {
+        printf("Invalid size: %d\n", args->content_size);
         return 1;
     }
 
     if (args->content_size == 0) {
         args->content_size = DEFAULT_CONTENT_SIZE;
-    } else if (args->content_size <= 0) {
-        fprintf(stderr, "Invalid content-size: %d\n", args->content_size);
-        return 1;
     }
 
     if (args->threads == 0) {
@@ -80,6 +114,12 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
         return 1;
     }
 
+    if (args->depth < 0) {
+        args->depth = G_MAXINT32;
+    } else {
+        args->depth += 1;
+    }
+
     if (args->name == NULL) {
         args->name = g_path_get_basename(args->output);
     }
@@ -87,12 +127,62 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
     if (args->rewrite_url == NULL) {
         args->rewrite_url = DEFAULT_REWRITE_URL;
     }
+
+    if (args->archive == NULL || strcmp(args->archive, "recurse") == 0) {
+        args->archive_mode = ARC_MODE_RECURSE;
+    } else if (strcmp(args->archive, "list") == 0) {
+        args->archive_mode = ARC_MODE_LIST;
+    } else if (strcmp(args->archive, "shallow") == 0) {
+        args->archive_mode = ARC_MODE_SHALLOW;
+    } else if (strcmp(args->archive, "skip") == 0) {
+        args->archive_mode = ARC_MODE_SKIP;
+    } else {
+        fprintf(stderr, "Archive mode must be one of (skip, list, shallow, recurse), got '%s'", args->archive);
+        return 1;
+    }
+
+    if (args->tesseract_lang != NULL) {
+        TessBaseAPI *api = TessBaseAPICreate();
+
+        char filename[128];
+        sprintf(filename, "%s.traineddata", args->tesseract_lang);
+        const char * path = find_file_in_paths(TESS_DATAPATHS, filename);
+        if (path == NULL) {
+            LOG_FATAL("cli.c", "Could not find tesseract language file!");
+        }
+
+        ret = TessBaseAPIInit3(api, path, args->tesseract_lang);
+        if (ret != 0) {
+            fprintf(stderr, "Could not initialize tesseract with lang '%s'\n", args->tesseract_lang);
+            return 1;
+        }
+        TessBaseAPIEnd(api);
+        TessBaseAPIDelete(api);
+
+        args->tesseract_path = path;
+    }
+
+    LOG_DEBUGF("cli.c", "arg quality=%f", args->quality)
+    LOG_DEBUGF("cli.c", "arg size=%d", args->size)
+    LOG_DEBUGF("cli.c", "arg content_size=%d", args->content_size)
+    LOG_DEBUGF("cli.c", "arg threads=%d", args->threads)
+    LOG_DEBUGF("cli.c", "arg incremental=%s", args->incremental)
+    LOG_DEBUGF("cli.c", "arg output=%s", args->output)
+    LOG_DEBUGF("cli.c", "arg rewrite_url=%s", args->rewrite_url)
+    LOG_DEBUGF("cli.c", "arg name=%s", args->name)
+    LOG_DEBUGF("cli.c", "arg depth=%d", args->depth)
+    LOG_DEBUGF("cli.c", "arg path=%s", args->path)
+    LOG_DEBUGF("cli.c", "arg archive=%s", args->archive)
+    LOG_DEBUGF("cli.c", "arg tesseract_lang=%s", args->tesseract_lang)
+    LOG_DEBUGF("cli.c", "arg tesseract_path=%s", args->tesseract_path)
+
     return 0;
 }
 
-#ifndef SIST_SCAN_ONLY
 int index_args_validate(index_args_t *args, int argc, const char **argv) {
 
+    LogCtx.verbose = 1;
+
     if (argc < 2) {
         fprintf(stderr, "Required positional argument: PATH.\n");
         return 1;
@@ -100,20 +190,62 @@ int index_args_validate(index_args_t *args, int argc, const char **argv) {
 
     char *index_path = abspath(argv[1]);
     if (index_path == NULL) {
-        fprintf(stderr, "File not found: %s", argv[1]);
+        fprintf(stderr, "File not found: %s\n", argv[1]);
         return 1;
     } else {
         args->index_path = argv[1];
+        free(index_path);
     }
 
     if (args->es_url == NULL) {
         args->es_url = DEFAULT_ES_URL;
     }
+
+    if (args->script_path != NULL) {
+        struct stat info;
+        int res = stat(args->script_path, &info);
+
+        if (res == -1) {
+            fprintf(stderr, "Error opening script file '%s': %s\n", args->script_path, strerror(errno));
+            return 1;
+        }
+
+        int fd = open(args->script_path, O_RDONLY);
+        if (fd == -1) {
+            fprintf(stderr, "Error opening script file '%s': %s\n", args->script_path, strerror(errno));
+            return 1;
+        }
+
+        args->script = malloc(info.st_size + 1);
+        res = read(fd, args->script, info.st_size);
+        if (res == -1) {
+            fprintf(stderr, "Error reading script file '%s': %s\n", args->script_path, strerror(errno));
+            return 1;
+        }
+
+        *(args->script + info.st_size) = '\0';
+        close(fd);
+    }
+
+    if (args->batch_size == 0) {
+        args->batch_size = DEFAULT_BATCH_SIZE;
+    }
+
+    LOG_DEBUGF("cli.c", "arg es_url=%s", args->es_url)
+    LOG_DEBUGF("cli.c", "arg index_path=%s", args->index_path)
+    LOG_DEBUGF("cli.c", "arg script_path=%s", args->script_path)
+    LOG_DEBUGF("cli.c", "arg script=%s", args->script)
+    LOG_DEBUGF("cli.c", "arg print=%d", args->print)
+    LOG_DEBUGF("cli.c", "arg batch_size=%d", args->batch_size)
+    LOG_DEBUGF("cli.c", "arg force_reset=%d", args->force_reset)
+
     return 0;
 }
 
 int web_args_validate(web_args_t *args, int argc, const char **argv) {
 
+    LogCtx.verbose = 1;
+
     if (argc < 2) {
         fprintf(stderr, "Required positional argument: PATH.\n");
         return 1;
@@ -131,16 +263,33 @@ int web_args_validate(web_args_t *args, int argc, const char **argv) {
         args->port = DEFAULT_PORT;
     }
 
+    if (args->credentials != NULL) {
+        args->b64credentials = onion_base64_encode(args->credentials, (int) strlen(args->credentials));
+        //Remove trailing newline
+        *(args->b64credentials + strlen(args->b64credentials) - 1) = '\0';
+    }
+
     args->index_count = argc - 1;
     args->indices = argv + 1;
 
     for (int i = 0; i < args->index_count; i++) {
         char *abs_path = abspath(args->indices[i]);
         if (abs_path == NULL) {
-            fprintf(stderr, "File not found: %s", abs_path);
+            fprintf(stderr, "File not found: %s\n", args->indices[i]);
             return 1;
         }
     }
+
+    LOG_DEBUGF("cli.c", "arg es_url=%s", args->es_url)
+    LOG_DEBUGF("cli.c", "arg bind=%s", args->bind)
+    LOG_DEBUGF("cli.c", "arg port=%s", args->port)
+    LOG_DEBUGF("cli.c", "arg credentials=%s", args->credentials)
+    LOG_DEBUGF("cli.c", "arg b64credentials=%s", args->b64credentials)
+    LOG_DEBUGF("cli.c", "arg index_count=%d", args->index_count)
+    for (int i = 0; i < args->index_count; i++) {
+        LOG_DEBUGF("cli.c", "arg indices[%d]=%s", i, args->indices[i])
+    }
+
     return 0;
 }
 
@@ -153,5 +302,4 @@ web_args_t *web_args_create() {
     web_args_t *args = calloc(sizeof(web_args_t), 1);
     return args;
 }
-#endif
 

src/cli.h

@@ -12,17 +12,25 @@ typedef struct scan_args {
     char *output;
     char *rewrite_url;
     char *name;
+    int depth;
     char *path;
+    char *archive;
+    archive_mode_t archive_mode;
+    char *tesseract_lang;
+    const char *tesseract_path;
 } scan_args_t;
 
 scan_args_t *scan_args_create();
+void scan_args_destroy(scan_args_t *args);
 int scan_args_validate(scan_args_t *args, int argc, const char **argv);
 
-#ifndef SIST_SCAN_ONLY
 typedef struct index_args {
     char *es_url;
     const char *index_path;
+    const char *script_path;
+    char *script;
     int print;
+    int batch_size;
     int force_reset;
 } index_args_t;
 
@@ -30,15 +38,19 @@ typedef struct web_args {
     char *es_url;
     char *bind;
     char *port;
+    char *credentials;
+    char *b64credentials;
     int index_count;
     const char **indices;
 } web_args_t;
 
 index_args_t *index_args_create();
+void index_args_destroy(index_args_t *args);
+
 web_args_t *web_args_create();
+void web_args_destroy(web_args_t *args);
 
 int index_args_validate(index_args_t *args, int argc, const char **argv);
 int web_args_validate(web_args_t *args, int argc, const char **argv);
-#endif
 
 #endif

src/ctx.h

@@ -15,6 +15,10 @@ struct {
     int threads;
     int content_size;
     float tn_qscale;
+    int depth;
+    archive_mode_t archive_mode;
+    int verbose;
+    int very_verbose;
 
     size_t stat_tn_size;
     size_t stat_index_size;
@@ -23,20 +27,27 @@ struct {
     GHashTable *copy_table;
 
     pthread_mutex_t mupdf_mu;
+    char * tesseract_lang;
+    char * tesseract_path;
 } ScanCtx;
 
+struct {
+    int verbose;
+    int very_verbose;
+    int no_color;
+} LogCtx;
 
-#ifndef SIST_SCAN_ONLY
 struct {
     char *es_url;
+    int batch_size;
 } IndexCtx;
 
 struct {
     char *es_url;
     int index_count;
+    char *b64credentials;
     struct index_t indices[16];
 } WebCtx;
-#endif
 
 
 #endif

src/index/elastic.c

@@ -6,11 +6,9 @@
 #include <stdio.h>
 #include <string.h>
 #include <cJSON/cJSON.h>
-#include <src/ctx.h>
 
 #include "static_generated.c"
 
-#define BULK_INDEX_SIZE 100
 
 typedef struct es_indexer {
     int queued;
@@ -29,13 +27,14 @@ void print_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
     cJSON_AddStringToObject(line, "_id", uuid_str);
     cJSON_AddStringToObject(line, "_index", "sist2");
     cJSON_AddStringToObject(line, "_type", "_doc");
-    cJSON_AddItemToObject(line, "_source", document);
+    cJSON_AddItemReferenceToObject(line, "_source", document);
 
     char *json = cJSON_PrintUnformatted(line);
 
     printf("%s\n", json);
 
-    cJSON_free(line);
+    cJSON_free(json);
+    cJSON_Delete(line);
 }
 
 void index_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
@@ -54,6 +53,40 @@ void index_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
     elastic_index_line(bulk_line);
 }
 
+void execute_update_script(const char *script, const char index_id[UUID_STR_LEN]) {
+
+    cJSON *body = cJSON_CreateObject();
+    cJSON *script_obj = cJSON_AddObjectToObject(body, "script");
+    cJSON_AddStringToObject(script_obj, "lang", "painless");
+    cJSON_AddStringToObject(script_obj, "source", script);
+
+    cJSON *query = cJSON_AddObjectToObject(body, "query");
+    cJSON *term_obj = cJSON_AddObjectToObject(query, "term");
+    cJSON_AddStringToObject(term_obj, "index", index_id);
+
+    char * str = cJSON_Print(body);
+
+    char bulk_url[4096];
+    snprintf(bulk_url, 4096, "%s/sist2/_update_by_query?pretty", Indexer->es_url);
+    response_t *r = web_post(bulk_url, str, "Content-Type: application/json");
+    LOG_INFOF("elastic.c", "Executed user script <%d>", r->status_code);
+    cJSON *resp = cJSON_Parse(r->body);
+
+    cJSON_free(str);
+    cJSON_Delete(body);
+    free_response(r);
+
+    cJSON *error = cJSON_GetObjectItem(resp, "error");
+    if (error != NULL) {
+        char *error_str = cJSON_Print(error);
+
+        LOG_ERRORF("elastic.c", "User script error: \n%s", error_str);
+        cJSON_free(error_str);
+    }
+
+    cJSON_Delete(resp);
+}
+
 void elastic_flush() {
 
     if (Indexer == NULL) {
@@ -98,7 +131,12 @@ void elastic_flush() {
     char bulk_url[4096];
     snprintf(bulk_url, 4096, "%s/sist2/_bulk", Indexer->es_url);
     response_t *r = web_post(bulk_url, buf, "Content-Type: application/x-ndjson");
-    printf("Indexed %3d documents (%zukB) <%d>\n", count, buf_cur / 1024, r->status_code);
+
+    if (r->status_code == 0) {
+        LOG_FATALF("elastic.c", "Could not connect to %s, make sure that elasticsearch is running!\n", IndexCtx.es_url)
+    }
+
+    LOG_INFOF("elastic.c", "Indexed %d documents (%zukB) <%d>", count, buf_cur / 1024, r->status_code);
 
     cJSON *ret_json = cJSON_Parse(r->body);
     if (cJSON_GetObjectItem(ret_json, "errors")->valueint != 0) {
@@ -106,7 +144,7 @@ void elastic_flush() {
         cJSON_ArrayForEach(err, cJSON_GetObjectItem(ret_json, "items")) {
             if (cJSON_GetObjectItem(cJSON_GetObjectItem(err, "index"), "status")->valueint != 201) {
                 char* str = cJSON_Print(err);
-                fprintf(stderr, "%s\n", str);
+                LOG_ERRORF("elastic.c", "%s\n", str);
                 cJSON_free(str);
             }
         }
@@ -115,6 +153,7 @@ void elastic_flush() {
     cJSON_Delete(ret_json);
 
     free_response(r);
+    free(buf);
 }
 
 void elastic_index_line(es_bulk_line_t *line) {
@@ -133,15 +172,14 @@ void elastic_index_line(es_bulk_line_t *line) {
 
     Indexer->queued += 1;
 
-    if (Indexer->queued >= BULK_INDEX_SIZE) {
+    if (Indexer->queued >= IndexCtx.batch_size) {
         elastic_flush();
     }
 }
 
 es_indexer_t *create_indexer(const char *url) {
 
-    size_t url_len = strlen(url);
-    char *es_url = malloc(url_len);
+    char *es_url = malloc(strlen(url) + 1);
     strcpy(es_url, url);
 
     es_indexer_t *indexer = malloc(sizeof(es_indexer_t));
@@ -154,18 +192,27 @@ es_indexer_t *create_indexer(const char *url) {
     return indexer;
 }
 
-void destroy_indexer() {
+void destroy_indexer(char * script, char index_id[UUID_STR_LEN]) {
 
     char url[4096];
 
     snprintf(url, sizeof(url), "%s/sist2/_refresh", IndexCtx.es_url);
     response_t *r = web_post(url, "", NULL);
-    printf("Refresh index <%d>\n", r->status_code);
+    LOG_INFOF("elastic.c", "Refresh index <%d>", r->status_code);
+    free_response(r);
+
+    if (script != NULL) {
+        execute_update_script(script, index_id);
+    }
+
+    snprintf(url, sizeof(url), "%s/sist2/_refresh", IndexCtx.es_url);
+    r = web_post(url, "", NULL);
+    LOG_INFOF("elastic.c", "Refresh index <%d>", r->status_code);
     free_response(r);
 
     snprintf(url, sizeof(url), "%s/sist2/_forcemerge", IndexCtx.es_url);
     r = web_post(url, "", NULL);
-    printf("Merge index <%d>\n", r->status_code);
+    LOG_INFOF("elastic.c", "Merge index <%d>", r->status_code);
     free_response(r);
 
     if (Indexer != NULL) {
@@ -185,32 +232,32 @@ void elastic_init(int force_reset) {
 
     if (!index_exists || force_reset) {
         r = web_delete(url);
-        printf("Delete index <%d>\n", r->status_code);
+        LOG_INFOF("elastic.c", "Delete index <%d>", r->status_code);
         free_response(r);
 
         snprintf(url, 4096, "%s/sist2", IndexCtx.es_url);
         r = web_put(url, "", NULL);
-        printf("Create index <%d>\n", r->status_code);
+        LOG_INFOF("elastic.c", "Create index <%d>", r->status_code);
         free_response(r);
 
         snprintf(url, 4096, "%s/sist2/_close", IndexCtx.es_url);
         r = web_post(url, "", NULL);
-        printf("Close index <%d>\n", r->status_code);
+        LOG_INFOF("elastic.c", "Close index <%d>", r->status_code);
         free_response(r);
 
         snprintf(url, 4096, "%s/sist2/_settings", IndexCtx.es_url);
         r = web_put(url, settings_json, "Content-Type: application/json");
-        printf("Update settings <%d>\n", r->status_code);
+        LOG_INFOF("elastic.c", "Update settings <%d>", r->status_code);
         free_response(r);
 
         snprintf(url, 4096, "%s/sist2/_mappings/_doc?include_type_name=true", IndexCtx.es_url);
         r = web_put(url, mappings_json, "Content-Type: application/json");
-        printf("Update mappings <%d>\n", r->status_code);
+        LOG_INFOF("elastic.c", "Update mappings <%d>", r->status_code);
         free_response(r);
 
         snprintf(url, 4096, "%s/sist2/_open", IndexCtx.es_url);
         r = web_post(url, "", NULL);
-        printf("Open index <%d>\n", r->status_code);
+        LOG_INFOF("elastic.c", "Open index <%d>", r->status_code);
         free_response(r);
     }
 }
@@ -227,3 +274,28 @@ cJSON *elastic_get_document(const char *uuid_str) {
     free_response(r);
     return json;
 }
+
+char *elastic_get_status() {
+    char url[4096];
+    snprintf(url, 4096,
+            "%s/_cluster/state/metadata/sist2?filter_path=metadata.indices.*.state", WebCtx.es_url);
+
+    response_t *r = web_get(url);
+    cJSON *json = NULL;
+    char *status = malloc(128 * sizeof(char));
+    status[0] = '\0';
+
+    if (r->status_code == 200) {
+        json = cJSON_Parse(r->body);
+        const cJSON *metadata = cJSON_GetObjectItem(json, "metadata");
+        if (metadata != NULL) {
+            const cJSON *indices = cJSON_GetObjectItem(metadata, "indices");
+            const cJSON *sist2 = cJSON_GetObjectItem(indices, "sist2");
+            const cJSON *state = cJSON_GetObjectItem(sist2, "state");
+            strcpy(status, state->valuestring);
+        }
+    }
+    free_response(r);
+    cJSON_Delete(json);
+    return status;
+}

src/index/elastic.h

@@ -24,10 +24,12 @@ void index_json(cJSON *document, const char uuid_str[UUID_STR_LEN]);
 
 es_indexer_t *create_indexer(const char* es_url);
 
-void destroy_indexer();
+void destroy_indexer(char *script, char index_id[UUID_STR_LEN]);
 
 void elastic_init(int force_reset);
 
 cJSON *elastic_get_document(const char *uuid_str);
 
+char *elastic_get_status();
+
 #endif

src/io/serialize.c

@@ -1,7 +1,7 @@
 #include "src/ctx.h"
 #include "serialize.h"
 
-static __thread int IndexFd = -1;
+static __thread int index_fd = -1;
 
 typedef struct {
     unsigned char uuid[16];
@@ -34,6 +34,7 @@ void write_index_descriptor(char *path, index_descriptor_t *desc) {
     cJSON_AddStringToObject(json, "version", desc->version);
     cJSON_AddStringToObject(json, "root", desc->root);
     cJSON_AddStringToObject(json, "name", desc->name);
+    cJSON_AddStringToObject(json, "type", desc->type);
     cJSON_AddStringToObject(json, "rewrite_url", desc->rewrite_url);
     cJSON_AddNumberToObject(json, "timestamp", (double) desc->timestamp);
 
@@ -54,6 +55,11 @@ index_descriptor_t read_index_descriptor(char *path) {
     struct stat info;
     stat(path, &info);
     int fd = open(path, O_RDONLY);
+
+    if (fd == -1) {
+        LOG_FATAL("serialize.c", "Invalid/corrupt index (Could not find descriptor)\n")
+    }
+
     char *buf = malloc(info.st_size + 1);
     read(fd, buf, info.st_size);
     *(buf + info.st_size) = '\0';
@@ -69,6 +75,11 @@ index_descriptor_t read_index_descriptor(char *path) {
     descriptor.root_len = (short) strlen(descriptor.root);
     strcpy(descriptor.version, cJSON_GetObjectItem(json, "version")->valuestring);
     strcpy(descriptor.uuid, cJSON_GetObjectItem(json, "uuid")->valuestring);
+    if (cJSON_GetObjectItem(json, "type") == NULL) {
+        strcpy(descriptor.type, INDEX_TYPE_BIN);
+    } else {
+        strcpy(descriptor.type, cJSON_GetObjectItem(json, "type")->valuestring);
+    }
 
     cJSON_Delete(json);
     free(buf);
@@ -105,6 +116,8 @@ char *get_meta_key_text(enum metakey meta_key) {
             return "title";
         case MetaFontName:
             return "font_name";
+        case MetaParent:
+            return "parent";
         default:
             return NULL;
     }
@@ -113,13 +126,13 @@ char *get_meta_key_text(enum metakey meta_key) {
 
 void write_document(document_t *doc) {
 
-    if (IndexFd == -1) {
+    if (index_fd == -1) {
         char dstfile[PATH_MAX];
         pthread_t self = pthread_self();
         snprintf(dstfile, PATH_MAX, "%s_index_%lu", ScanCtx.index.path, self);
-        IndexFd = open(dstfile, O_CREAT | O_WRONLY | O_APPEND, S_IRUSR | S_IWUSR);
+        index_fd = open(dstfile, O_CREAT | O_WRONLY | O_APPEND, S_IRUSR | S_IWUSR);
 
-        if (IndexFd == -1) {
+        if (index_fd == -1) {
             perror("open");
         }
     }
@@ -152,17 +165,20 @@ void write_document(document_t *doc) {
     }
     dyn_buffer_write_char(&buf, '\n');
 
-    write(IndexFd, buf.buf, buf.cur);
+    int res = write(index_fd, buf.buf, buf.cur);
+    if (res == -1) {
+        perror("write");
+    }
     ScanCtx.stat_index_size += buf.cur;
     dyn_buffer_destroy(&buf);
 }
 
-void serializer_cleanup() {
-    close(IndexFd);
+void thread_cleanup() {
+    close(index_fd);
 }
 
-void read_index(const char *path, const char index_id[UUID_STR_LEN], index_func func) {
 
+void read_index_bin(const char *path, const char *index_id, index_func func) {
     line_t line;
     dyn_buffer_t buf = dyn_buffer_create();
 
@@ -218,7 +234,7 @@ void read_index(const char *path, const char index_id[UUID_STR_LEN], index_func
                 case MetaMediaBitrate: {
                     long value;
                     fread(&value, sizeof(long), 1, file);
-                    cJSON_AddNumberToObject(document, get_meta_key_text(key), value);
+                    cJSON_AddNumberToObject(document, get_meta_key_text(key), (double) value);
                     break;
                 }
                 case MetaMediaAudioCodec:
@@ -238,6 +254,7 @@ void read_index(const char *path, const char index_id[UUID_STR_LEN], index_func
                 case MetaAlbumArtist:
                 case MetaGenre:
                 case MetaFontName:
+                case MetaParent:
                 case MetaTitle: {
                     buf.cur = 0;
                     while ((c = getc(file)) != 0) {
@@ -250,19 +267,102 @@ void read_index(const char *path, const char index_id[UUID_STR_LEN], index_func
                     break;
                 }
                 default:
-                    fprintf(stderr, "Invalid meta key (corrupt index): %x\n", key);
-                    break;
+                LOG_FATALF("serialize.c", "Invalid meta key (corrupt index): %x", key)
             }
 
             key = getc(file);
         }
 
         func(document, uuid_str);
-        cJSON_free(document);
+        cJSON_Delete(document);
+    }
+    dyn_buffer_destroy(&buf);
+    fclose(file);
+}
+
+const char *json_type_copy_fields[] = {
+        "mime", "name", "path", "extension", "index", "size", "mtime", "parent",
+
+        // Meta
+        "title", "content", "width", "height", "duration", "audioc", "videoc",
+        "bitrate", "artist", "album", "album_artist", "genre", "title", "font_name",
+
+        // Special
+        "tag", "_url"
+};
+
+const char *json_type_array_fields[] = {
+        "_keyword", "_text"
+};
+
+void read_index_json(const char *path, UNUSED(const char *index_id), index_func func) {
+
+    FILE *file = fopen(path, "r");
+    while (1) {
+        char *line = NULL;
+        size_t len;
+        size_t read = getline(&line, &len, file);
+        if (read == -1) {
+            if (line) {
+                free(line);
+            }
+            break;
+        }
+
+        cJSON *input = cJSON_Parse(line);
+        if (input == NULL) {
+            LOG_FATALF("serialize.c", "Could not parse JSON line: \n%s", line)
+        }
+        if (line) {
+            free(line);
+        }
+
+        cJSON *document = cJSON_CreateObject();
+        const char *uuid_str = cJSON_GetObjectItem(input, "_id")->valuestring;
+
+        for (int i = 0; i < (sizeof(json_type_copy_fields) / sizeof(json_type_copy_fields[0])); i++) {
+            cJSON *value = cJSON_GetObjectItem(input, json_type_copy_fields[i]);
+            if (value != NULL) {
+                cJSON_AddItemReferenceToObject(document, json_type_copy_fields[i], value);
+            }
+        }
+
+        for (int i = 0; i < (sizeof(json_type_array_fields) / sizeof(json_type_array_fields[0])); i++) {
+            cJSON *arr = cJSON_GetObjectItem(input, json_type_array_fields[i]);
+            if (arr != NULL) {
+                cJSON *obj;
+                cJSON_ArrayForEach(obj, arr) {
+                    char key[1024];
+                    cJSON *k = cJSON_GetObjectItem(obj, "k");
+                    cJSON *v = cJSON_GetObjectItem(obj, "v");
+                    if (k == NULL || v == NULL || !cJSON_IsString(k) || !cJSON_IsString(v)) {
+                        char *str = cJSON_Print(obj);
+                        LOG_FATALF("serialize.c", "Invalid %s member: must contain .k and .v string fields: \n%s",
+                                   json_type_array_fields[i], str)
+                    }
+                    snprintf(key, sizeof(key), "%s.%s", json_type_array_fields[i], k->valuestring);
+                    cJSON_AddStringToObject(document, key, v->valuestring);
+                }
+            }
+        }
+
+        func(document, uuid_str);
+        cJSON_Delete(document);
+        cJSON_Delete(input);
+
     }
     fclose(file);
 }
 
+void read_index(const char *path, const char index_id[UUID_STR_LEN], const char *type, index_func func) {
+
+    if (strcmp(type, INDEX_TYPE_BIN) == 0) {
+        read_index_bin(path, index_id, func);
+    } else if (strcmp(type, INDEX_TYPE_JSON) == 0) {
+        read_index_json(path, index_id, func);
+    }
+}
+
 void incremental_read(GHashTable *table, const char *filepath) {
     FILE *file = fopen(filepath, "rb");
     line_t line;

src/io/serialize.h

@@ -11,14 +11,14 @@ void incremental_copy(store_t *store, store_t *dst_store, const char *filepath,
 
 void write_document(document_t *doc);
 
-void read_index(const char *path, const char[UUID_STR_LEN], index_func);
+void read_index(const char *path, const char[UUID_STR_LEN], const char *type, index_func);
 
 void incremental_read(GHashTable *table, const char *filepath);
 
 /**
  * Must be called after write_document
  */
-void serializer_cleanup();
+void thread_cleanup();
 
 void write_index_descriptor(char *path, index_descriptor_t *desc);
 

src/io/store.c

@@ -9,13 +9,13 @@ store_t *store_create(char *path) {
     mdb_env_create(&store->env);
 
     int open_ret = mdb_env_open(store->env,
-                 path,
-                 MDB_WRITEMAP | MDB_MAPASYNC,
-                 S_IRUSR | S_IWUSR
+                                path,
+                                MDB_WRITEMAP | MDB_MAPASYNC,
+                                S_IRUSR | S_IWUSR
     );
 
     if (open_ret != 0) {
-        fprintf(stderr, "Error while opening store: %s", mdb_strerror(open_ret));
+        fprintf(stderr, "Error while opening store: %s (%s)\n", mdb_strerror(open_ret), path);
         exit(1);
     }
 
@@ -42,6 +42,12 @@ void store_destroy(store_t *store) {
 
 void store_write(store_t *store, char *key, size_t key_len, char *buf, size_t buf_len) {
 
+    if (LogCtx.very_verbose) {
+        char uuid_str[UUID_STR_LEN];
+        uuid_unparse((unsigned char *) key, uuid_str);
+        LOG_DEBUGF("store.c", "Store write {%s} %lu bytes", uuid_str, buf_len)
+    }
+
     MDB_val mdb_key;
     mdb_key.mv_data = key;
     mdb_key.mv_size = key_len;
@@ -68,6 +74,8 @@ void store_write(store_t *store, char *key, size_t key_len, char *buf, size_t bu
         mdb_env_set_mapsize(store->env, store->size);
         mdb_txn_begin(store->env, NULL, 0, &txn);
         put_ret = mdb_put(txn, store->dbi, &mdb_key, &mdb_value, 0);
+
+        LOG_INFOF("store.c", "Updated mdb mapsize to %lu bytes", store->size)
     }
 
     mdb_txn_commit(txn);

src/io/walk.c

@@ -1,7 +1,8 @@
 #include "walk.h"
 #include "src/ctx.h"
 
-parse_job_t *create_parse_job(const char *filepath, const struct stat *info, int base) {
+__always_inline
+parse_job_t *create_fs_parse_job(const char *filepath, const struct stat *info, int base) {
     int len = (int) strlen(filepath);
     parse_job_t *job = malloc(sizeof(parse_job_t) + len);
 
@@ -14,14 +15,22 @@ parse_job_t *create_parse_job(const char *filepath, const struct stat *info, int
         job->ext = len;
     }
 
-    memcpy(&(job->info), info, sizeof(struct stat));
+    job->info = *info;
+
+    memset(job->parent, 0, 16);
+
+    job->vfile.filepath = job->filepath;
+    job->vfile.read = fs_read;
+    job->vfile.close = fs_close;
+    job->vfile.fd = -1;
+    job->vfile.is_fs_file = TRUE;
 
     return job;
 }
 
 int handle_entry(const char *filepath, const struct stat *info, int typeflag, struct FTW *ftw) {
-    if (typeflag == FTW_F && S_ISREG(info->st_mode)) {
-        parse_job_t *job = create_parse_job(filepath, info, ftw->base);
+    if (ftw->level <= ScanCtx.depth && typeflag == FTW_F && S_ISREG(info->st_mode)) {
+        parse_job_t *job = create_fs_parse_job(filepath, info, ftw->base);
         tpool_add_work(ScanCtx.pool, parse, job);
     }
 

src/log.c

@@ -0,0 +1,99 @@
+#include "log.h"
+
+const char *log_colors[] = {
+        "\033[34m", "\033[01;34m", "\033[0m",
+        "\033[01;33m", "\033[31m", "\033[01;31m"
+};
+
+const char *log_levels[] = {
+        "DEBUG", "INFO", "WARNING", "ERROR", "FATAL"
+};
+
+void sist_logf(char *filepath, int level, char *format, ...) {
+
+    static int is_tty = -1;
+    if (is_tty == -1) {
+        is_tty = isatty(STDERR_FILENO);
+    }
+
+    char log_str[LOG_MAX_LENGTH];
+
+    unsigned long long pid = (unsigned long long) pthread_self();
+
+    char datetime[32];
+    time_t t;
+    struct tm result;
+    t = time(NULL);
+    localtime_r(&t, &result);
+    strftime(datetime, sizeof(datetime), "%Y-%m-%d %H:%M:%S", &result);
+
+    int log_len;
+    if (is_tty) {
+        log_len = snprintf(
+                log_str, sizeof(log_str),
+                "\033[%dm[%04X]%s [%s] [%s %s] ",
+                31 + ((unsigned int) (pid)) % 7, pid, log_colors[level],
+                datetime, log_levels[level], filepath
+        );
+    } else {
+        log_len = snprintf(
+                log_str, sizeof(log_str),
+                "[%04X] [%s] [%s %s] ",
+                pid, datetime, log_levels[level], filepath
+        );
+    }
+
+    va_list ap;
+    va_start(ap, format);
+    size_t maxsize = sizeof(log_str) - log_len;
+    log_len += vsnprintf(log_str + log_len, maxsize, format, ap);
+    va_end(ap);
+
+    if (is_tty) {
+        log_len += sprintf(log_str + log_len, "\033[0m\n");
+    } else {
+        *(log_str + log_len) = '\n';
+        log_len += 1;
+    }
+
+    write(STDERR_FILENO, log_str, log_len);
+}
+
+void sist_log(char *filepath, int level, char *str) {
+
+    static int is_tty = -1;
+    if (is_tty == -1) {
+        is_tty = isatty(STDERR_FILENO);
+    }
+
+    char log_str[LOG_MAX_LENGTH];
+
+    unsigned long long pid = (unsigned long long) pthread_self();
+
+    char datetime[32];
+    time_t t;
+    struct tm result;
+    t = time(NULL);
+    localtime_r(&t, &result);
+    strftime(datetime, sizeof(datetime), "%Y-%m-%d %H:%M:%S", &result);
+
+    int log_len;
+    if (is_tty) {
+        log_len = snprintf(
+                log_str, sizeof(log_str),
+                "\033[%dm[%04X]%s [%s] [%s %s] %s \033[0m\n",
+                31 + ((unsigned int) (pid)) % 7, pid, log_colors[level],
+                datetime, log_levels[level], filepath,
+                str
+        );
+    } else {
+        log_len = snprintf(
+                log_str, sizeof(log_str),
+                "[%04X] [%s] [%s %s] %s \n",
+                pid, datetime, log_levels[level], filepath,
+                str
+        );
+    }
+
+    write(STDERR_FILENO, log_str, log_len);
+}

src/log.h

@@ -0,0 +1,45 @@
+#ifndef SIST2_LOG_H
+#define SIST2_LOG_H
+
+#define LOG_MAX_LENGTH 8192
+
+#define SIST_DEBUG 0
+#define SIST_INFO 1
+#define SIST_WARNING 2
+#define SIST_ERROR 3
+#define SIST_FATAL 4
+
+#define LOG_DEBUGF(filepath, fmt, ...) \
+    if (LogCtx.very_verbose) {sist_logf(filepath, SIST_DEBUG, fmt, __VA_ARGS__);}
+#define LOG_DEBUG(filepath, str) \
+    if (LogCtx.very_verbose) {sist_log(filepath, SIST_DEBUG, str);}
+
+#define LOG_INFOF(filepath, fmt, ...) \
+    if (LogCtx.verbose) {sist_logf(filepath, SIST_INFO, fmt, __VA_ARGS__);}
+#define LOG_INFO(filepath, str) \
+    if (LogCtx.verbose) {sist_log(filepath, SIST_INFO, str);}
+
+#define LOG_WARNINGF(filepath, fmt, ...) \
+    if (LogCtx.verbose) {sist_logf(filepath, SIST_WARNING, fmt, __VA_ARGS__);}
+#define LOG_WARNING(filepath, str) \
+    if (LogCtx.verbose) {sist_log(filepath, SIST_WARNING, str);}
+
+#define LOG_ERRORF(filepath, fmt, ...) \
+    if (LogCtx.verbose) {sist_logf(filepath, SIST_ERROR, fmt, __VA_ARGS__);}
+#define LOG_ERROR(filepath, str) \
+    if (LogCtx.verbose) {sist_log(filepath, SIST_ERROR, str);}
+
+#define LOG_FATALF(filepath, fmt, ...) \
+    sist_logf(filepath, SIST_FATAL, fmt, __VA_ARGS__);\
+    exit(-1);
+#define LOG_FATAL(filepath, str) \
+    sist_log(filepath, SIST_FATAL, str);\
+    exit(-1);
+
+#include "src/sist.h"
+
+void sist_logf(char *filepath, int level, char *format, ...);
+
+void sist_log(char *filepath, int level, char *str);
+
+#endif

src/main.c

@@ -1,16 +1,12 @@
 #include "sist.h"
 #include "ctx.h"
 
-#ifndef SIST_SCAN_ONLY
 #define DESCRIPTION "Lightning-fast file system indexer and search tool."
-#else
-#define DESCRIPTION "Lightning-fast file system indexer and search tool. (SCAN ONLY)"
-#endif
 
 #define EPILOG "Made by simon987 <me@simon987.net>. Released under GPL-3.0"
 
 
-static const char *const Version = "1.1.3";
+static const char *const Version = "1.2.4";
 static const char *const usage[] = {
         "sist2 scan [OPTION]... PATH",
         "sist2 index [OPTION]... INDEX",
@@ -19,10 +15,9 @@ static const char *const usage[] = {
 };
 
 void global_init() {
-    #ifndef SIST_SCAN_ONLY
     curl_global_init(CURL_GLOBAL_NOTHING);
-    #endif
     av_log_set_level(AV_LOG_QUIET);
+    opcInitLibrary();
 }
 
 void init_dir(const char *dirpath) {
@@ -34,17 +29,13 @@ void init_dir(const char *dirpath) {
     uuid_unparse(uuid, ScanCtx.index.desc.uuid);
     time(&ScanCtx.index.desc.timestamp);
     strcpy(ScanCtx.index.desc.version, Version);
+    strcpy(ScanCtx.index.desc.type, INDEX_TYPE_BIN);
 
     write_index_descriptor(path, &ScanCtx.index.desc);
 }
 
 void scan_print_header() {
-    printf("sist2 V%s\n", Version);
-    printf("---------------------\n");
-    printf("threads\t\t%d\n", ScanCtx.threads);
-    printf("tn_qscale\t%.1f/31.0\n", ScanCtx.tn_qscale);
-    printf("tn_size\t\t%dpx\n", ScanCtx.tn_size);
-    printf("output\t\t%s\n", ScanCtx.index.path);
+    LOG_INFOF("main.c", "sist2 v%s", Version)
 }
 
 void sist2_scan(scan_args_t *args) {
@@ -53,10 +44,14 @@ void sist2_scan(scan_args_t *args) {
     ScanCtx.tn_size = args->size;
     ScanCtx.content_size = args->content_size;
     ScanCtx.threads = args->threads;
+    ScanCtx.depth = args->depth;
+    ScanCtx.archive_mode = args->archive_mode;
     strncpy(ScanCtx.index.path, args->output, sizeof(ScanCtx.index.path));
     strncpy(ScanCtx.index.desc.name, args->name, sizeof(ScanCtx.index.desc.name));
     strncpy(ScanCtx.index.desc.root, args->path, sizeof(ScanCtx.index.desc.root));
     ScanCtx.index.desc.root_len = (short) strlen(ScanCtx.index.desc.root);
+    ScanCtx.tesseract_lang = args->tesseract_lang;
+    ScanCtx.tesseract_path = args->tesseract_path;
 
     init_dir(ScanCtx.index.path);
 
@@ -92,7 +87,7 @@ void sist2_scan(scan_args_t *args) {
         printf("Loaded %d items in to mtime table.", g_hash_table_size(ScanCtx.original_table));
     }
 
-    ScanCtx.pool = tpool_create(args->threads, serializer_cleanup);
+    ScanCtx.pool = tpool_create(args->threads, thread_cleanup);
     tpool_start(ScanCtx.pool);
     walk_directory_tree(ScanCtx.index.desc.root);
     tpool_wait(ScanCtx.pool);
@@ -124,10 +119,10 @@ void sist2_scan(scan_args_t *args) {
     store_destroy(ScanCtx.index.store);
 }
 
-#ifndef SIST_SCAN_ONLY
 void sist2_index(index_args_t *args) {
 
     IndexCtx.es_url = args->es_url;
+    IndexCtx.batch_size = args->batch_size;
 
     if (!args->print) {
         elastic_init(args->force_reset);
@@ -137,8 +132,12 @@ void sist2_index(index_args_t *args) {
     snprintf(descriptor_path, PATH_MAX, "%s/descriptor.json", args->index_path);
 
     index_descriptor_t desc = read_index_descriptor(descriptor_path);
-    if (strcmp(desc.version, Version) != 0) {
-        fprintf(stderr, "Version mismatch! Index is v%s but executable is v%s\n", desc.version, Version);
+
+    LOG_DEBUGF("main.c", "descriptor version %s (%s)", desc.version, desc.type)
+
+    if (strcmp(desc.version, Version) != 0 && strcmp(desc.version, INDEX_VERSION_EXTERNAL) != 0) {
+        fprintf(stderr, "Version mismatch! Index is %s but executable is %s/%s\n",
+                desc.version, Version, INDEX_VERSION_EXTERNAL);
         return;
     }
 
@@ -160,13 +159,14 @@ void sist2_index(index_args_t *args) {
         if (strncmp(de->d_name, "_index_", sizeof("_index_") - 1) == 0) {
             char file_path[PATH_MAX];
             snprintf(file_path, PATH_MAX, "%s/%s", args->index_path, de->d_name);
-            read_index(file_path, desc.uuid, f);
+            read_index(file_path, desc.uuid, desc.type, f);
         }
     }
+    closedir(dir);
 
     if (!args->print) {
         elastic_flush();
-        destroy_indexer();
+        destroy_indexer(args->script, desc.uuid);
     }
 }
 
@@ -174,6 +174,7 @@ void sist2_web(web_args_t *args) {
 
     WebCtx.es_url = args->es_url;
     WebCtx.index_count = args->index_count;
+    WebCtx.b64credentials = args->b64credentials;
 
     for (int i = 0; i < args->index_count; i++) {
         char *abs_path = abspath(args->indices[i]);
@@ -195,7 +196,6 @@ void sist2_web(web_args_t *args) {
 
     serve(args->bind, args->port);
 }
-#endif
 
 
 int main(int argc, const char *argv[]) {
@@ -203,41 +203,54 @@ int main(int argc, const char *argv[]) {
     global_init();
 
     scan_args_t *scan_args = scan_args_create();
-    #ifndef SIST_SCAN_ONLY
     index_args_t *index_args = index_args_create();
     web_args_t *web_args = web_args_create();
-    #endif
 
-    char * common_es_url = NULL;
+    int arg_version = 0;
+
+    char *common_es_url = NULL;
 
     struct argparse_option options[] = {
             OPT_HELP(),
 
+            OPT_BOOLEAN('v', "version", &arg_version, "Show version and exit"),
+            OPT_BOOLEAN(0, "verbose", &LogCtx.verbose, "Turn on logging"),
+            OPT_BOOLEAN(0, "very-verbose", &LogCtx.very_verbose, "Turn on debug messages"),
+
             OPT_GROUP("Scan options"),
             OPT_INTEGER('t', "threads", &scan_args->threads, "Number of threads. DEFAULT=1"),
             OPT_FLOAT('q', "quality", &scan_args->quality,
-                      "Thumbnail quality, on a scale of 1.0 to 31.0, 1.0 being the best. DEFAULT=15"),
-            OPT_INTEGER(0, "size", &scan_args->size, "Thumbnail size, in pixels. DEFAULT=200"),
+                      "Thumbnail quality, on a scale of 1.0 to 31.0, 1.0 being the best. DEFAULT=5"),
+            OPT_INTEGER(0, "size", &scan_args->size,
+                        "Thumbnail size, in pixels. Use negative value to disable. DEFAULT=500"),
             OPT_INTEGER(0, "content-size", &scan_args->content_size,
-                        "Number of bytes to be extracted from text documents. DEFAULT=4096"),
+                        "Number of bytes to be extracted from text documents. Use negative value to disable. DEFAULT=32768"),
             OPT_STRING(0, "incremental", &scan_args->incremental,
                        "Reuse an existing index and only scan modified files."),
             OPT_STRING('o', "output", &scan_args->output, "Output directory. DEFAULT=index.sist2/"),
             OPT_STRING(0, "rewrite-url", &scan_args->rewrite_url, "Serve files from this url instead of from disk."),
             OPT_STRING(0, "name", &scan_args->name, "Index display name. DEFAULT: (name of the directory)"),
+            OPT_INTEGER(0, "depth", &scan_args->depth, "Scan up to DEPTH subdirectories deep. "
+                                                       "Use 0 to only scan files in PATH. DEFAULT: -1"),
+            OPT_STRING(0, "archive", &scan_args->archive, "Archive file mode (skip|list|shallow|recurse). "
+                                                          "skip: Don't parse, list: only get file names as text, "
+                                                          "shallow: Don't parse archives inside archives. DEFAULT: recurse"),
+            OPT_STRING(0, "ocr", &scan_args->tesseract_lang, "Tesseract language (use tesseract --list-langs to see "
+                                                             "which are installed on your machine)"),
 
-            #ifndef SIST_SCAN_ONLY
             OPT_GROUP("Index options"),
-            OPT_STRING(0, "es-url", &common_es_url, "Elasticsearch url. DEFAULT=http://localhost:9200"),
+            OPT_STRING(0, "es-url", &common_es_url, "Elasticsearch url with port. DEFAULT=http://localhost:9200"),
             OPT_BOOLEAN('p', "print", &index_args->print, "Just print JSON documents to stdout."),
+            OPT_STRING(0, "script-file", &index_args->script_path, "Path to user script."),
+            OPT_INTEGER(0, "batch-size", &index_args->batch_size, "Index batch size. DEFAULT: 100"),
             OPT_BOOLEAN('f', "force-reset", &index_args->force_reset, "Reset Elasticsearch mappings and settings. "
-                                                              "(You must use this option the first time you use the index command)"),
+                                                                      "(You must use this option the first time you use the index command)"),
 
             OPT_GROUP("Web options"),
             OPT_STRING(0, "es-url", &common_es_url, "Elasticsearch url. DEFAULT=http://localhost:9200"),
             OPT_STRING(0, "bind", &web_args->bind, "Listen on this address. DEFAULT=localhost"),
             OPT_STRING(0, "port", &web_args->port, "Listen on this port. DEFAULT=4090"),
-            #endif
+            OPT_STRING(0, "auth", &web_args->credentials, "Basic auth in user:password format"),
 
             OPT_END(),
     };
@@ -247,10 +260,17 @@ int main(int argc, const char *argv[]) {
     argparse_describe(&argparse, DESCRIPTION, EPILOG);
     argc = argparse_parse(&argparse, argc, argv);
 
-    #ifndef SIST_SCAN_ONLY
+    if (arg_version) {
+        printf(Version);
+        exit(0);
+    }
+
+    if (LogCtx.very_verbose != 0) {
+        LogCtx.verbose = 1;
+    }
+
     web_args->es_url = common_es_url;
     index_args->es_url = common_es_url;
-    #endif
 
     if (argc == 0) {
         argparse_usage(&argparse);
@@ -265,7 +285,6 @@ int main(int argc, const char *argv[]) {
 
     }
 
-    #ifndef SIST_SCAN_ONLY
     else if (strcmp(argv[0], "index") == 0) {
 
         int err = index_args_validate(index_args, argc, argv);
@@ -283,12 +302,17 @@ int main(int argc, const char *argv[]) {
         sist2_web(web_args);
 
     }
-    #endif
     else {
         fprintf(stderr, "Invalid command: '%s'\n", argv[0]);
         argparse_usage(&argparse);
         return 1;
     }
     printf("\n");
+
+    scan_args_destroy(scan_args);
+
+    index_args_destroy(index_args);
+    web_args_destroy(web_args);
+
     return 0;
 }

src/parsing/arc.c

@@ -0,0 +1,157 @@
+#include "arc.h"
+#include "src/ctx.h"
+
+#define ARC_BUF_SIZE 8192
+
+int should_parse_filtered_file(const char *filepath, int ext) {
+    char tmp[PATH_MAX * 2];
+
+    if (ext == 0) {
+        return FALSE;
+    }
+
+    memcpy(tmp, filepath, ext - 1);
+    *(tmp + ext - 1) = '\0';
+
+    char *idx = strrchr(tmp, '.');
+
+    if (idx == NULL) {
+        return FALSE;
+    }
+
+    if (strcmp(idx, ".tar") == 0) {
+        return TRUE;
+    }
+
+    return FALSE;
+}
+
+int arc_read(struct vfile *f, void *buf, size_t size) {
+    return archive_read_data(f->arc, buf, size);
+}
+
+typedef struct arc_data {
+    vfile_t *f;
+    char buf[ARC_BUF_SIZE];
+} arc_data_f;
+
+int vfile_open_callback(struct archive *a, void *user_data) {
+    arc_data_f *data = user_data;
+
+    if (data->f->is_fs_file && data->f->fd == -1) {
+        data->f->fd = open(data->f->filepath, O_RDONLY);
+    }
+
+    return ARCHIVE_OK;
+}
+
+long vfile_read_callback(struct archive *a, void *user_data, const void **buf) {
+    arc_data_f *data = user_data;
+
+    *buf = data->buf;
+    return data->f->read(data->f, data->buf, ARC_BUF_SIZE);
+}
+
+int vfile_close_callback(struct archive *a, void *user_data) {
+    arc_data_f *data = user_data;
+
+    if (data->f->close != NULL) {
+        data->f->close(data->f);
+    }
+
+    return ARCHIVE_OK;
+}
+
+void parse_archive(vfile_t *f, document_t *doc) {
+
+    struct archive *a;
+    struct archive_entry *entry;
+
+
+    arc_data_f data;
+    data.f = f;
+
+    int ret = 0;
+    if (data.f->is_fs_file) {
+
+        a = archive_read_new();
+        archive_read_support_filter_all(a);
+        archive_read_support_format_all(a);
+
+        ret = archive_read_open_filename(a, doc->filepath, ARC_BUF_SIZE);
+    } else if (ScanCtx.archive_mode == ARC_MODE_RECURSE) {
+
+        a = archive_read_new();
+        archive_read_support_filter_all(a);
+        archive_read_support_format_all(a);
+
+        ret = archive_read_open(
+                a, &data,
+                vfile_open_callback,
+                vfile_read_callback,
+                vfile_close_callback
+        );
+    } else {
+        return;
+    }
+
+    if (ret != ARCHIVE_OK) {
+        LOG_ERRORF(doc->filepath, "(arc.c) [%d] %s", ret, archive_error_string(a))
+        archive_read_free(a);
+        return;
+    }
+
+    if (ScanCtx.archive_mode == ARC_MODE_LIST) {
+
+        dyn_buffer_t buf = dyn_buffer_create();
+
+        while (archive_read_next_header(a, &entry) == ARCHIVE_OK) {
+            if (S_ISREG(archive_entry_stat(entry)->st_mode)) {
+
+                char *path = (char *) archive_entry_pathname(entry);
+
+                dyn_buffer_append_string(&buf, path);
+                dyn_buffer_write_char(&buf, '\n');
+            }
+        }
+        dyn_buffer_write_char(&buf, '\0');
+
+        meta_line_t *meta_list = malloc(sizeof(meta_line_t) + buf.cur);
+        meta_list->key = MetaContent;
+        strcpy(meta_list->strval, buf.buf);
+        APPEND_META(doc, meta_list);
+        dyn_buffer_destroy(&buf);
+
+    } else {
+
+        parse_job_t *sub_job = malloc(sizeof(parse_job_t) + PATH_MAX * 2);
+
+        sub_job->vfile.close = NULL;
+        sub_job->vfile.read = arc_read;
+        sub_job->vfile.arc = a;
+        sub_job->vfile.filepath = sub_job->filepath;
+        sub_job->vfile.is_fs_file = FALSE;
+        memcpy(sub_job->parent, doc->uuid, sizeof(uuid_t));
+
+        while (archive_read_next_header(a, &entry) == ARCHIVE_OK) {
+            sub_job->info = *archive_entry_stat(entry);
+            if (S_ISREG(sub_job->info.st_mode)) {
+                sprintf(sub_job->filepath, "%s#/%s", f->filepath, archive_entry_pathname(entry));
+                sub_job->base = (int) (strrchr(sub_job->filepath, '/') - sub_job->filepath) + 1;
+
+                char *p = strrchr(sub_job->filepath, '.');
+                if (p != NULL) {
+                    sub_job->ext = (int) (p - sub_job->filepath + 1);
+                } else {
+                    sub_job->ext = (int) strlen(sub_job->filepath);
+                }
+
+                parse(sub_job);
+            }
+        }
+
+        free(sub_job);
+    }
+
+    archive_read_free(a);
+}

src/parsing/arc.h

@@ -0,0 +1,12 @@
+#ifndef SIST2_ARC_H
+#define SIST2_ARC_H
+
+#include "src/sist.h"
+
+int should_parse_filtered_file(const char *filepath, int ext);
+
+void parse_archive(vfile_t *f, document_t *doc);
+
+int arc_read(struct vfile * f, void *buf, size_t size);
+
+#endif

src/parsing/doc.c

@@ -0,0 +1,114 @@
+#include "doc.h"
+#include "src/ctx.h"
+
+void dump_text(mceTextReader_t *reader, dyn_buffer_t *buf) {
+
+    mce_skip_attributes(reader);
+
+    mce_start_children(reader) {
+        mce_start_element(reader, NULL, _X("t")) {
+            mce_skip_attributes(reader);
+            mce_start_children(reader) {
+                mce_start_text(reader) {
+                    char *str = (char *) xmlTextReaderConstValue(reader->reader);
+                    dyn_buffer_append_string(buf, str);
+                    dyn_buffer_write_char(buf, ' ');
+                } mce_end_text(reader);
+            } mce_end_children(reader);
+        } mce_end_element(reader);
+
+        mce_start_element(reader, NULL, NULL) {
+            dump_text(reader, buf);
+        } mce_end_element(reader);
+
+    } mce_end_children(reader)
+}
+
+__always_inline
+int should_read_part(opcPart part) {
+
+    char *part_name = (char *) part;
+
+    if (part == NULL) {
+        return FALSE;
+    }
+
+    if (    // Word
+            strcmp(part_name, "word/document.xml") == 0
+            || strncmp(part_name, "word/footer", sizeof("word/footer") - 1) == 0
+            || strncmp(part_name, "word/header", sizeof("word/header") - 1) == 0
+            // PowerPoint
+            || strncmp(part_name, "ppt/slides/slide", sizeof("ppt/slides/slide") - 1) == 0
+            || strncmp(part_name, "ppt/notesSlides/notesSlide", sizeof("ppt/notesSlides/notesSlide") - 1) == 0
+            // Excel
+            || strncmp(part_name, "xl/worksheets/sheet", sizeof("xl/worksheets/sheet") - 1) == 0
+            || strcmp(part_name, "xl/sharedStrings.xml") == 0
+            || strcmp(part_name, "xl/workbook.xml") == 0
+            ) {
+        return TRUE;
+    }
+
+    return FALSE;
+}
+
+__always_inline
+void read_part(opcContainer *c, dyn_buffer_t *buf, opcPart part, document_t *doc) {
+
+    mceTextReader_t reader;
+    int options;
+    if (LogCtx.very_verbose) {
+        options = XML_PARSE_NONET;
+    } else {
+        options = XML_PARSE_NOWARNING | XML_PARSE_NOERROR | XML_PARSE_NONET;
+    }
+
+    int ret = opcXmlReaderOpen(c, &reader, part, NULL, "UTF-8", options);
+
+    if (ret != OPC_ERROR_NONE) {
+        LOG_ERRORF(doc->filepath, "(doc.c) opcXmlReaderOpen() returned error code %d", ret);
+        return;
+    }
+
+    mce_start_document(&reader) {
+        mce_start_element(&reader, NULL, NULL) {
+                dump_text(&reader, buf);
+        } mce_end_element(&reader);
+    } mce_end_document(&reader);
+
+    mceTextReaderCleanup(&reader);
+}
+
+void parse_doc(void *mem, size_t mem_len, document_t *doc) {
+
+    if (mem == NULL) {
+        return;
+    }
+
+    opcContainer *c = opcContainerOpenMem(mem, mem_len, OPC_OPEN_READ_ONLY, NULL);
+    if (c == NULL) {
+        LOG_ERROR(doc->filepath, "(doc.c) Couldn't open document with opcContainerOpenMem()");
+        return;
+    }
+
+    dyn_buffer_t buf = dyn_buffer_create();
+
+    opcPart part = opcPartGetFirst(c);
+    do {
+        if (should_read_part(part)) {
+            read_part(c, &buf, part, doc);
+        }
+    } while ((part = opcPartGetNext(c, part)));
+
+    opcContainerClose(c, OPC_CLOSE_NOW);
+
+    if (buf.cur > 0) {
+        dyn_buffer_write_char(&buf, '\0');
+
+        meta_line_t *meta = malloc(sizeof(meta_line_t) + buf.cur);
+        meta->key = MetaContent;
+        strcpy(meta->strval, buf.buf);
+        APPEND_META(doc, meta)
+    }
+
+    dyn_buffer_destroy(&buf);
+}

src/parsing/doc.h

@@ -0,0 +1,8 @@
+#ifndef SIST2_DOC_H
+#define SIST2_DOC_H
+
+#include "src/sist.h"
+
+void parse_doc(void *buf, size_t buf_len, document_t *doc);
+
+#endif

src/parsing/font.c

@@ -1,11 +1,9 @@
 #include "font.h"
 
-#include "ft2build.h"
-#include "freetype/freetype.h"
 
 #include "src/ctx.h"
 
-__thread FT_Library library = NULL;
+__thread FT_Library ft_lib = NULL;
 
 
 typedef struct text_dimensions {
@@ -139,16 +137,17 @@ void bmp_format(dyn_buffer_t *buf, text_dimensions_t dimensions, const unsigned
 }
 
 void parse_font(const char *buf, size_t buf_len, document_t *doc) {
-    if (library == NULL) {
-        FT_Init_FreeType(&library);
+    if (ft_lib == NULL) {
+        FT_Init_FreeType(&ft_lib);
     }
     if (buf == NULL) {
         return;
     }
 
     FT_Face face;
-    FT_Error err = FT_New_Memory_Face(library, (unsigned char *) buf, buf_len, 0, &face);
+    FT_Error err = FT_New_Memory_Face(ft_lib, (unsigned char *) buf, buf_len, 0, &face);
     if (err != 0) {
+        LOG_ERRORF(doc->filepath, "(font.c) FT_New_Memory_Face() returned error code [%d] %s", err, ft_error_string(err));
         return;
     }
 
@@ -169,11 +168,16 @@ void parse_font(const char *buf, size_t buf_len, document_t *doc) {
     strcpy(meta_name->strval, font_name);
     APPEND_META(doc, meta_name)
 
+    if (ScanCtx.tn_size <= 0) {
+        return;
+    }
+
     int pixel = 64;
     int num_chars = (int) strlen(font_name);
 
     err = FT_Set_Pixel_Sizes(face, 0, pixel);
     if (err != 0) {
+        LOG_WARNINGF(doc->filepath, "(font.c) FT_Set_Pixel_Sizes() returned error code [%d] %s", err, ft_error_string(err))
         return;
     }
 
@@ -192,6 +196,7 @@ void parse_font(const char *buf, size_t buf_len, document_t *doc) {
             c = c >= 'a' && c <= 'z' ? c - 32 : c + 32;
             err = FT_Load_Char(face, c, FT_LOAD_NO_HINTING | FT_LOAD_RENDER);
             if (err != 0) {
+                LOG_WARNINGF(doc->filepath, "(font.c) FT_Load_Char() returned error code [%d] %s", err, ft_error_string(err));
                 continue;
             }
         }

src/parsing/media.c

@@ -2,6 +2,7 @@
 #include "src/ctx.h"
 
 #define MIN_SIZE 32
+#define AVIO_BUF_SIZE 8192
 
 __always_inline
 AVCodecContext *alloc_jpeg_encoder(int dstW, int dstH, float qscale) {
@@ -76,7 +77,8 @@ AVFrame *scale_frame(const AVCodecContext *decoder, const AVFrame *frame, int si
     return scaled_frame;
 }
 
-AVFrame *read_frame(AVFormatContext *pFormatCtx, AVCodecContext *decoder, int stream_idx) {
+__always_inline
+AVFrame *read_frame(AVFormatContext *pFormatCtx, AVCodecContext *decoder, int stream_idx, document_t *doc) {
     AVFrame *frame = av_frame_alloc();
 
     AVPacket avPacket;
@@ -90,7 +92,10 @@ AVFrame *read_frame(AVFormatContext *pFormatCtx, AVCodecContext *decoder, int st
 
             if (read_frame_ret != 0) {
                 if (read_frame_ret != AVERROR_EOF) {
-                    fprintf(stderr, "Error reading frame: %d\n", read_frame_ret);
+                    LOG_WARNINGF(doc->filepath,
+                                 "(media.c) avcodec_read_frame() returned error code [%d] %s",
+                                 read_frame_ret, av_err2str(read_frame_ret)
+                    )
                 }
                 av_frame_free(&frame);
                 av_packet_unref(&avPacket);
@@ -108,7 +113,10 @@ AVFrame *read_frame(AVFormatContext *pFormatCtx, AVCodecContext *decoder, int st
         // Feed it to decoder
         int decode_ret = avcodec_send_packet(decoder, &avPacket);
         if (decode_ret != 0) {
-            printf("Error decoding frame: %s\n", av_err2str(decode_ret));
+            LOG_WARNINGF(doc->filepath,
+                         "(media.c) avcodec_send_packet() returned error code [%d] %s",
+                         decode_ret, av_err2str(decode_ret)
+            )
         }
         av_packet_unref(&avPacket);
         receive_ret = avcodec_receive_frame(decoder, frame);
@@ -119,6 +127,7 @@ AVFrame *read_frame(AVFormatContext *pFormatCtx, AVCodecContext *decoder, int st
 #define APPEND_TAG_META(doc, tag_, keyname) \
     text_buffer_t tex = text_buffer_create(-1); \
     text_buffer_append_string0(&tex, tag_->value); \
+    text_buffer_terminate_string(&tex); \
     meta_line_t *meta_tag = malloc(sizeof(meta_line_t) + tex.dyn_buffer.cur); \
     meta_tag->key = keyname; \
     strcpy(meta_tag->strval, tex.dyn_buffer.buf); \
@@ -151,7 +160,8 @@ void append_audio_meta(AVFormatContext *pFormatCtx, document_t *doc) {
 }
 
 __always_inline
-void append_video_meta(AVFormatContext *pFormatCtx, AVFrame *frame, document_t *doc, int include_audio_tags, int is_video) {
+void
+append_video_meta(AVFormatContext *pFormatCtx, AVFrame *frame, document_t *doc, int include_audio_tags, int is_video) {
 
     if (is_video) {
         meta_line_t *meta_duration = malloc(sizeof(meta_line_t));
@@ -188,22 +198,11 @@ void append_video_meta(AVFormatContext *pFormatCtx, AVFrame *frame, document_t *
     }
 }
 
-void parse_media(const char *filepath, document_t *doc) {
+void parse_media(AVFormatContext *pFormatCtx, document_t *doc) {
 
     int video_stream = -1;
     int audio_stream = -1;
 
-    AVFormatContext *pFormatCtx = avformat_alloc_context();
-    if (pFormatCtx == NULL) {
-        fprintf(stderr, "Could not allocate AVFormatContext! %s \n", filepath);
-        return;
-    }
-    int res = avformat_open_input(&pFormatCtx, filepath, NULL, NULL);
-    if (res < 0) {
-        fprintf(stderr, "media error: %s %s\n", filepath, av_err2str(res));
-        return;
-    }
-
     avformat_find_stream_info(pFormatCtx, NULL);
 
     for (int i = (int) pFormatCtx->nb_streams - 1; i >= 0; i--) {
@@ -242,7 +241,7 @@ void parse_media(const char *filepath, document_t *doc) {
         }
     }
 
-    if (video_stream != -1) {
+    if (video_stream != -1 && ScanCtx.tn_size > 0) {
         AVStream *stream = pFormatCtx->streams[video_stream];
 
         if (stream->codecpar->width <= MIN_SIZE || stream->codecpar->height <= MIN_SIZE) {
@@ -269,7 +268,7 @@ void parse_media(const char *filepath, document_t *doc) {
             }
         }
 
-        AVFrame *frame = read_frame(pFormatCtx, decoder, video_stream);
+        AVFrame *frame = read_frame(pFormatCtx, decoder, video_stream, doc);
         if (frame == NULL) {
             avcodec_free_context(&decoder);
             avformat_close_input(&pFormatCtx);
@@ -314,3 +313,69 @@ void parse_media(const char *filepath, document_t *doc) {
     avformat_free_context(pFormatCtx);
 }
 
+void parse_media_filename(const char *filepath, document_t *doc) {
+
+    AVFormatContext *pFormatCtx = avformat_alloc_context();
+    if (pFormatCtx == NULL) {
+        LOG_ERROR(doc->filepath, "(media.c) Could not allocate context with avformat_alloc_context()")
+        return;
+    }
+    int res = avformat_open_input(&pFormatCtx, filepath, NULL, NULL);
+    if (res < 0) {
+        LOG_ERRORF(doc->filepath, "(media.c) avformat_open_input() returned [%d] %s", res, av_err2str(res))
+        avformat_close_input(&pFormatCtx);
+        avformat_free_context(pFormatCtx);
+        return;
+    }
+
+    parse_media(pFormatCtx, doc);
+}
+
+
+int vfile_read(void *ptr, uint8_t *buf, int buf_size) {
+    struct vfile *f = ptr;
+
+    int ret = f->read(f, buf, buf_size);
+
+    if (ret == 0) {
+        return AVERROR_EOF;
+    }
+    return ret;
+}
+
+void parse_media_vfile(struct vfile *f, document_t *doc) {
+
+    AVFormatContext *pFormatCtx = avformat_alloc_context();
+    if (pFormatCtx == NULL) {
+        LOG_ERROR(doc->filepath, "(media.c) Could not allocate context with avformat_alloc_context()")
+        return;
+    }
+
+    unsigned char *buffer = (unsigned char *) av_malloc(AVIO_BUF_SIZE);
+    AVIOContext *io_ctx = avio_alloc_context(buffer, AVIO_BUF_SIZE, 0, f, vfile_read, NULL, NULL);
+
+    pFormatCtx->pb = io_ctx;
+    pFormatCtx->flags |= AVFMT_FLAG_CUSTOM_IO;
+
+    int res = avformat_open_input(&pFormatCtx, "", NULL, NULL);
+    if (res == -5) {
+        // Tried to parse media that requires seek
+        av_free(io_ctx->buffer);
+        avio_context_free(&io_ctx);
+        avformat_close_input(&pFormatCtx);
+        avformat_free_context(pFormatCtx);
+        return;
+    } else if (res < 0) {
+        LOG_ERRORF(doc->filepath, "(media.c) avformat_open_input() returned [%d] %s", res, av_err2str(res))
+        av_free(io_ctx->buffer);
+        avio_context_free(&io_ctx);
+        avformat_close_input(&pFormatCtx);
+        avformat_free_context(pFormatCtx);
+        return;
+    }
+
+    parse_media(pFormatCtx, doc);
+    av_free(io_ctx->buffer);
+    avio_context_free(&io_ctx);
+}
+

src/parsing/media.h

@@ -7,6 +7,8 @@
 #define MIN_VIDEO_SIZE 1024 * 64
 #define MIN_IMAGE_SIZE 1024 * 2
 
-void parse_media(const char * filepath, document_t *doc);
+void parse_media_filename(const char * filepath, document_t *doc);
+
+void parse_media_vfile(struct vfile *f, document_t *doc);
 
 #endif

src/parsing/mime.h

@@ -8,7 +8,7 @@
 #define MIME_EMPTY 1
 
 #define DONT_PARSE 0x80000000
-#define SHOULD_PARSE(mime_id) (mime_id & DONT_PARSE) != DONT_PARSE
+#define SHOULD_PARSE(mime_id) (mime_id & DONT_PARSE) != DONT_PARSE && mime_id != 0
 
 #define PDF_MASK 0x40000000
 #define IS_PDF(mime_id) (mime_id & PDF_MASK) == PDF_MASK
@@ -16,6 +16,15 @@
 #define FONT_MASK 0x20000000
 #define IS_FONT(mime_id) (mime_id & FONT_MASK) == FONT_MASK
 
+#define ARC_MASK 0x10000000
+#define IS_ARC(mime_id) (mime_id & ARC_MASK) == ARC_MASK
+
+#define ARC_FILTER_MASK 0x08000000
+#define IS_ARC_FILTER(mime_id) (mime_id & ARC_FILTER_MASK) == ARC_FILTER_MASK
+
+#define DOC_MASK 0x04000000
+#define IS_DOC(mime_id) (mime_id & DOC_MASK) == DOC_MASK
+
 enum major_mime {
     MimeInvalid = 0,
     MimeModel = 1,

src/parsing/mime_generated.c

@@ -20,7 +20,7 @@ enum mime {
     application_freeloader=655372,
     application_futuresplash=655373,
     application_groupwise=655374,
-    application_gzip=655375,
+    application_gzip=655375 | 0x08000000,
     application_hta=655376,
     application_i_deas=655377,
     application_iges=655378,
@@ -72,9 +72,9 @@ enum mime {
     application_vnd_oasis_opendocument_presentation=655424,
     application_vnd_oasis_opendocument_spreadsheet=655425,
     application_vnd_oasis_opendocument_text=655426,
-    application_vnd_openxmlformats_officedocument_presentationml_presentation=655427,
-    application_vnd_openxmlformats_officedocument_spreadsheetml_sheet=655428,
-    application_vnd_openxmlformats_officedocument_wordprocessingml_document=655429,
+    application_vnd_openxmlformats_officedocument_presentationml_presentation=655427 | 0x04000000,
+    application_vnd_openxmlformats_officedocument_spreadsheetml_sheet=655428 | 0x04000000,
+    application_vnd_openxmlformats_officedocument_wordprocessingml_document=655429 | 0x04000000,
     application_vnd_symbian_install=655430,
     application_vnd_tcpdump_pcap=655431,
     application_vnd_wap_wmlc=655432,
@@ -82,342 +82,346 @@ enum mime {
     application_vnd_xara=655434,
     application_vocaltec_media_desc=655435,
     application_vocaltec_media_file=655436,
-    application_winhelp=655437,
-    application_wordperfect=655438,
-    application_wordperfect6_0=655439,
-    application_wordperfect6_1=655440,
-    application_x_123=655441,
-    application_x_7z_compressed=655442,
-    application_x_aim=655443,
-    application_x_apple_diskimage=655444,
-    application_x_arc=655445,
-    application_x_archive=655446,
-    application_x_atari_7800_rom=655447,
-    application_x_authorware_bin=655448,
-    application_x_authorware_map=655449,
-    application_x_authorware_seg=655450,
-    application_x_avira_qua=655451,
-    application_x_bcpio=655452,
-    application_x_bittorrent=655453,
-    application_x_bsh=655454,
-    application_x_bytecode_python=655455,
-    application_x_bzip=655456,
-    application_x_bzip2=655457,
-    application_x_cbr=655458,
-    application_x_cbz=655459 | 0x40000000,
-    application_x_cdlink=655460,
-    application_x_chat=655461,
-    application_x_chrome_extension=655462,
-    application_x_cocoa=655463,
-    application_x_conference=655464,
-    application_x_coredump=655465,
-    application_x_cpio=655466,
-    application_x_dbf=655467,
-    application_x_dbt=655468,
-    application_x_debian_package=655469,
-    application_x_deepv=655470,
-    application_x_director=655471,
-    application_x_dmp=655472,
-    application_x_dosdriver=655473,
-    application_x_dosexec=655474,
-    application_x_dvi=655475,
-    application_x_elc=655476,
+    application_warc=655437,
+    application_winhelp=655438,
+    application_wordperfect=655439,
+    application_wordperfect6_0=655440,
+    application_wordperfect6_1=655441,
+    application_x_123=655442,
+    application_x_7z_compressed=655443 | 0x10000000,
+    application_x_aim=655444,
+    application_x_apple_diskimage=655445,
+    application_x_arc=655446 | 0x10000000,
+    application_x_archive=655447,
+    application_x_atari_7800_rom=655448,
+    application_x_authorware_bin=655449,
+    application_x_authorware_map=655450,
+    application_x_authorware_seg=655451,
+    application_x_avira_qua=655452,
+    application_x_bcpio=655453,
+    application_x_bittorrent=655454,
+    application_x_bsh=655455,
+    application_x_bytecode_python=655456,
+    application_x_bzip=655457,
+    application_x_bzip2=655458 | 0x08000000,
+    application_x_cbr=655459,
+    application_x_cbz=655460 | 0x40000000,
+    application_x_cdlink=655461,
+    application_x_chat=655462,
+    application_x_chrome_extension=655463,
+    application_x_cocoa=655464,
+    application_x_conference=655465,
+    application_x_coredump=655466,
+    application_x_cpio=655467,
+    application_x_dbf=655468,
+    application_x_dbt=655469,
+    application_x_debian_package=655470,
+    application_x_deepv=655471,
+    application_x_director=655472,
+    application_x_dmp=655473,
+    application_x_dosdriver=655474,
+    application_x_dosexec=655475,
+    application_x_dvi=655476,
+    application_x_elc=655477,
     application_x_empty=1,
-    application_x_envoy=655478,
-    application_x_esrehber=655479,
-    application_x_excel=655480,
-    application_x_executable=655481,
-    application_x_font_gdos=655482,
-    application_x_font_pf2=655483,
-    application_x_font_pfm=655484,
-    application_x_font_sfn=655485,
-    application_x_font_ttf=655486 | 0x20000000,
-    application_x_freelance=655487,
-    application_x_gamecube_rom=655488,
-    application_x_gdbm=655489,
-    application_x_gettext_translation=655490,
-    application_x_git=655491,
-    application_x_gsp=655492,
-    application_x_gss=655493,
-    application_x_gtar=655494,
-    application_x_gzip=655495,
-    application_x_hdf=655496,
-    application_x_helpfile=655497,
-    application_x_httpd_imap=655498,
-    application_x_ima=655499,
-    application_x_innosetup=655500,
-    application_x_internett_signup=655501,
-    application_x_inventor=655502,
-    application_x_ip2=655503,
-    application_x_java_applet=655504,
-    application_x_java_commerce=655505,
-    application_x_java_image=655506,
-    application_x_java_jmod=655507,
-    application_x_java_keystore=655508,
-    application_x_kdelnk=655509,
-    application_x_koan=655510,
-    application_x_latex=655511,
-    application_x_livescreen=655512,
-    application_x_lotus=655513,
-    application_x_lz4=655514,
-    application_x_lz4_json=655515,
-    application_x_lzh=655516,
-    application_x_lzh_compressed=655517,
-    application_x_lzx=655518,
-    application_x_mach_binary=655519,
-    application_x_mach_executable=655520,
-    application_x_magic_cap_package_1_0=655521,
-    application_x_mathcad=655522,
-    application_x_maxis_dbpf=655523,
-    application_x_meme=655524,
-    application_x_midi=655525,
-    application_x_mif=655526,
-    application_x_mix_transfer=655527,
-    application_x_mobipocket_ebook=655528,
-    application_x_ms_compress_szdd=655529,
-    application_x_ms_pdb=655530,
-    application_x_ms_reader=655531,
-    application_x_msaccess=655532,
-    application_x_navi_animation=655533,
-    application_x_navidoc=655534,
-    application_x_navimap=655535,
-    application_x_navistyle=655536,
-    application_x_nes_rom=655537,
-    application_x_netcdf=655538,
-    application_x_newton_compatible_pkg=655539,
-    application_x_nintendo_ds_rom=655540,
-    application_x_object=655541,
-    application_x_omc=655542,
-    application_x_omcdatamaker=655543,
-    application_x_omcregerator=655544,
-    application_x_pagemaker=655545,
-    application_x_pcl=655546,
-    application_x_pgp_keyring=655547,
-    application_x_pixclscript=655548,
-    application_x_pkcs7_certreqresp=655549,
-    application_x_pkcs7_signature=655550,
-    application_x_project=655551,
-    application_x_qpro=655552,
-    application_x_rar=655553,
-    application_x_rpm=655554,
-    application_x_sdp=655555,
-    application_x_sea=655556,
-    application_x_seelogo=655557,
-    application_x_setupscript=655558,
-    application_x_shar=655559,
-    application_x_sharedlib=655560,
-    application_x_shockwave_flash=655561,
-    application_x_snappy_framed=655562,
-    application_x_sprite=655563,
-    application_x_sqlite3=655564,
-    application_x_sv4cpio=655565,
-    application_x_sv4crc=655566,
-    application_x_tar=655567,
-    application_x_tbook=655568,
-    application_x_terminfo=655569,
-    application_x_terminfo2=655570,
-    application_x_tex_tfm=655571,
-    application_x_texinfo=655572,
-    application_x_ustar=655573,
-    application_x_visio=655574,
-    application_x_vnd_audioexplosion_mzz=655575,
-    application_x_vnd_ls_xpix=655576,
-    application_x_vrml=655577,
-    application_x_wais_source=655578,
-    application_x_wine_extension_ini=655579,
-    application_x_wintalk=655580,
-    application_x_world=655581,
-    application_x_wri=655582,
-    application_x_x509_ca_cert=655583,
-    application_x_xz=655584,
-    application_x_zip=655585,
-    application_x_zstd=655586,
-    application_xml=655587,
-    application_zip=655588,
-    application_zlib=655589,
-    audio_it=458982,
-    audio_make=458983,
-    audio_mid=458984,
-    audio_midi=458985,
-    audio_mp4=458986,
-    audio_mpeg=458987,
-    audio_ogg=458988,
-    audio_s3m=458989,
-    audio_tsp_audio=458990,
-    audio_tsplayer=458991,
-    audio_vnd_qcelp=458992,
-    audio_voxware=458993,
-    audio_x_aiff=458994,
-    audio_x_flac=458995,
-    audio_x_gsm=458996,
-    audio_x_hx_aac_adts=458997,
-    audio_x_jam=458998,
-    audio_x_liveaudio=458999,
-    audio_x_m4a=459000,
-    audio_x_midi=459001,
-    audio_x_mod=459002,
-    audio_x_mp4a_latm=459003,
-    audio_x_mpeg_3=459004,
-    audio_x_mpequrl=459005,
-    audio_x_nspaudio=459006,
-    audio_x_pn_realaudio=459007,
-    audio_x_psid=459008,
-    audio_x_realaudio=459009,
-    audio_x_twinvq=459010,
-    audio_x_twinvq_plugin=459011,
-    audio_x_voc=459012,
-    audio_x_wav=459013,
-    audio_xm=459014,
-    font_otf=327943 | 0x20000000,
-    font_sfnt=327944 | 0x20000000,
-    font_woff=327945 | 0x20000000,
-    font_woff2=327946 | 0x20000000,
-    image_cmu_raster=524555,
-    image_fif=524556,
-    image_florian=524557,
-    image_g3fax=524558,
-    image_gif=524559,
-    image_heic=524560,
-    image_ief=524561,
-    image_jpeg=524562,
-    image_jutvision=524563,
-    image_naplps=524564,
-    image_pict=524565,
-    image_png=524566,
-    image_svg=524567 | 0x80000000,
-    image_svg_xml=524568 | 0x80000000,
-    image_tiff=524569,
-    image_vnd_adobe_photoshop=524570 | 0x80000000,
-    image_vnd_djvu=524571 | 0x80000000,
-    image_vnd_fpx=524572,
-    image_vnd_microsoft_icon=524573,
-    image_vnd_rn_realflash=524574,
-    image_vnd_rn_realpix=524575,
-    image_vnd_wap_wbmp=524576,
-    image_vnd_xiff=524577,
-    image_webp=524578,
-    image_wmf=524579,
-    image_x_3ds=524580,
-    image_x_cmu_raster=524581,
-    image_x_cur=524582,
-    image_x_dwg=524583,
-    image_x_eps=524584,
-    image_x_exr=524585,
-    image_x_gem=524586,
-    image_x_icns=524587,
-    image_x_icon=524588 | 0x80000000,
-    image_x_jg=524589,
-    image_x_jps=524590,
-    image_x_ms_bmp=524591,
-    image_x_niff=524592,
-    image_x_pcx=524593,
-    image_x_pict=524594,
-    image_x_portable_bitmap=524595,
-    image_x_portable_graymap=524596,
-    image_x_portable_pixmap=524597,
-    image_x_quicktime=524598,
-    image_x_rgb=524599,
-    image_x_tga=524600,
-    image_x_tiff=524601,
-    image_x_win_bitmap=524602,
-    image_x_xcf=524603 | 0x80000000,
-    image_x_xpixmap=524604 | 0x80000000,
-    image_x_xwindowdump=524605,
-    message_news=196926,
-    message_rfc822=196927,
-    model_vnd_dwf=65856,
-    model_vnd_gdl=65857,
-    model_vnd_gs_gdl=65858,
-    model_vrml=65859,
-    model_x_pov=65860,
-    text_PGP=590149,
-    text_asp=590150,
-    text_css=590151,
-    text_html=590152,
-    text_javascript=590153,
-    text_mcf=590154,
-    text_pascal=590155,
-    text_plain=590156,
-    text_richtext=590157,
-    text_rtf=590158,
-    text_scriplet=590159,
-    text_tab_separated_values=590160,
-    text_troff=590161,
-    text_uri_list=590162,
-    text_vnd_abc=590163,
-    text_vnd_fmi_flexstor=590164,
-    text_vnd_wap_wml=590165,
-    text_vnd_wap_wmlscript=590166,
-    text_webviewhtml=590167,
-    text_x_Algol68=590168,
-    text_x_asm=590169,
-    text_x_audiosoft_intra=590170,
-    text_x_awk=590171,
-    text_x_bcpl=590172,
-    text_x_c=590173,
-    text_x_c__=590174,
-    text_x_component=590175,
-    text_x_diff=590176,
-    text_x_fortran=590177,
-    text_x_java=590178,
-    text_x_la_asf=590179,
-    text_x_lisp=590180,
-    text_x_m=590181,
-    text_x_m4=590182,
-    text_x_makefile=590183,
-    text_x_ms_regedit=590184,
-    text_x_msdos_batch=590185,
-    text_x_objective_c=590186,
-    text_x_pascal=590187,
-    text_x_perl=590188,
-    text_x_php=590189,
-    text_x_po=590190,
-    text_x_python=590191,
-    text_x_ruby=590192,
-    text_x_sass=590193,
-    text_x_scss=590194,
-    text_x_server_parsed_html=590195,
-    text_x_setext=590196,
-    text_x_sgml=590197,
-    text_x_shellscript=590198,
-    text_x_speech=590199,
-    text_x_tcl=590200,
-    text_x_tex=590201,
-    text_x_uil=590202,
-    text_x_uuencode=590203,
-    text_x_vcalendar=590204,
-    text_x_vcard=590205,
-    text_xml=590206,
-    video_MP2T=393599,
-    video_animaflex=393600,
-    video_avi=393601,
-    video_avs_video=393602,
-    video_mp4=393603,
-    video_mpeg=393604,
-    video_quicktime=393605,
-    video_vdo=393606,
-    video_vivo=393607,
-    video_vnd_rn_realvideo=393608,
-    video_vosaic=393609,
-    video_webm=393610,
-    video_x_amt_demorun=393611,
-    video_x_amt_showrun=393612,
-    video_x_atomic3d_feature=393613,
-    video_x_dl=393614,
-    video_x_dv=393615,
-    video_x_fli=393616,
-    video_x_flv=393617,
-    video_x_isvideo=393618,
-    video_x_jng=393619 | 0x80000000,
-    video_x_m4v=393620,
-    video_x_matroska=393621,
-    video_x_mng=393622,
-    video_x_motion_jpeg=393623,
-    video_x_ms_asf=393624,
-    video_x_msvideo=393625,
-    video_x_qtc=393626,
-    video_x_sgi_movie=393627,
-    x_epoc_x_sisx_app=721308,
+    application_x_envoy=655479,
+    application_x_esrehber=655480,
+    application_x_excel=655481,
+    application_x_executable=655482,
+    application_x_font_gdos=655483,
+    application_x_font_pf2=655484,
+    application_x_font_pfm=655485,
+    application_x_font_sfn=655486,
+    application_x_font_ttf=655487 | 0x20000000,
+    application_x_freelance=655488,
+    application_x_gamecube_rom=655489,
+    application_x_gdbm=655490,
+    application_x_gettext_translation=655491,
+    application_x_git=655492,
+    application_x_gsp=655493,
+    application_x_gss=655494,
+    application_x_gtar=655495,
+    application_x_gzip=655496,
+    application_x_hdf=655497,
+    application_x_helpfile=655498,
+    application_x_httpd_imap=655499,
+    application_x_ima=655500,
+    application_x_innosetup=655501,
+    application_x_internett_signup=655502,
+    application_x_inventor=655503,
+    application_x_ip2=655504,
+    application_x_java_applet=655505,
+    application_x_java_commerce=655506,
+    application_x_java_image=655507,
+    application_x_java_jmod=655508,
+    application_x_java_keystore=655509,
+    application_x_kdelnk=655510,
+    application_x_koan=655511,
+    application_x_latex=655512,
+    application_x_livescreen=655513,
+    application_x_lotus=655514,
+    application_x_lz4=655515 | 0x08000000,
+    application_x_lz4_json=655516,
+    application_x_lzh=655517,
+    application_x_lzh_compressed=655518,
+    application_x_lzip=655519 | 0x08000000,
+    application_x_lzma=655520 | 0x08000000,
+    application_x_lzop=655521 | 0x08000000,
+    application_x_lzx=655522,
+    application_x_mach_binary=655523,
+    application_x_mach_executable=655524,
+    application_x_magic_cap_package_1_0=655525,
+    application_x_mathcad=655526,
+    application_x_maxis_dbpf=655527,
+    application_x_meme=655528,
+    application_x_midi=655529,
+    application_x_mif=655530,
+    application_x_mix_transfer=655531,
+    application_x_mobipocket_ebook=655532,
+    application_x_ms_compress_szdd=655533,
+    application_x_ms_pdb=655534,
+    application_x_ms_reader=655535,
+    application_x_msaccess=655536,
+    application_x_navi_animation=655537,
+    application_x_navidoc=655538,
+    application_x_navimap=655539,
+    application_x_navistyle=655540,
+    application_x_nes_rom=655541,
+    application_x_netcdf=655542,
+    application_x_newton_compatible_pkg=655543,
+    application_x_nintendo_ds_rom=655544,
+    application_x_object=655545,
+    application_x_omc=655546,
+    application_x_omcdatamaker=655547,
+    application_x_omcregerator=655548,
+    application_x_pagemaker=655549,
+    application_x_pcl=655550,
+    application_x_pgp_keyring=655551,
+    application_x_pixclscript=655552,
+    application_x_pkcs7_certreqresp=655553,
+    application_x_pkcs7_signature=655554,
+    application_x_project=655555,
+    application_x_qpro=655556,
+    application_x_rar=655557 | 0x10000000,
+    application_x_rpm=655558,
+    application_x_sdp=655559,
+    application_x_sea=655560,
+    application_x_seelogo=655561,
+    application_x_setupscript=655562,
+    application_x_shar=655563,
+    application_x_sharedlib=655564,
+    application_x_shockwave_flash=655565,
+    application_x_snappy_framed=655566,
+    application_x_sprite=655567,
+    application_x_sqlite3=655568,
+    application_x_sv4cpio=655569,
+    application_x_sv4crc=655570,
+    application_x_tar=655571 | 0x10000000,
+    application_x_tbook=655572,
+    application_x_terminfo=655573,
+    application_x_terminfo2=655574,
+    application_x_tex_tfm=655575,
+    application_x_texinfo=655576,
+    application_x_ustar=655577,
+    application_x_visio=655578,
+    application_x_vnd_audioexplosion_mzz=655579,
+    application_x_vnd_ls_xpix=655580,
+    application_x_vrml=655581,
+    application_x_wais_source=655582,
+    application_x_wine_extension_ini=655583,
+    application_x_wintalk=655584,
+    application_x_world=655585,
+    application_x_wri=655586,
+    application_x_x509_ca_cert=655587,
+    application_x_xz=655588 | 0x08000000,
+    application_x_zip=655589,
+    application_x_zstd=655590 | 0x08000000,
+    application_xml=655591,
+    application_zip=655592 | 0x10000000,
+    application_zlib=655593,
+    audio_it=458986,
+    audio_make=458987,
+    audio_mid=458988,
+    audio_midi=458989,
+    audio_mp4=458990,
+    audio_mpeg=458991,
+    audio_ogg=458992,
+    audio_s3m=458993,
+    audio_tsp_audio=458994,
+    audio_tsplayer=458995,
+    audio_vnd_qcelp=458996,
+    audio_voxware=458997,
+    audio_x_aiff=458998,
+    audio_x_flac=458999,
+    audio_x_gsm=459000,
+    audio_x_hx_aac_adts=459001,
+    audio_x_jam=459002,
+    audio_x_liveaudio=459003,
+    audio_x_m4a=459004,
+    audio_x_midi=459005,
+    audio_x_mod=459006,
+    audio_x_mp4a_latm=459007,
+    audio_x_mpeg_3=459008,
+    audio_x_mpequrl=459009,
+    audio_x_nspaudio=459010,
+    audio_x_pn_realaudio=459011,
+    audio_x_psid=459012,
+    audio_x_realaudio=459013,
+    audio_x_twinvq=459014,
+    audio_x_twinvq_plugin=459015,
+    audio_x_voc=459016,
+    audio_x_wav=459017,
+    audio_xm=459018,
+    font_otf=327947 | 0x20000000,
+    font_sfnt=327948 | 0x20000000,
+    font_woff=327949 | 0x20000000,
+    font_woff2=327950 | 0x20000000,
+    image_cmu_raster=524559,
+    image_fif=524560,
+    image_florian=524561,
+    image_g3fax=524562,
+    image_gif=524563,
+    image_heic=524564,
+    image_ief=524565,
+    image_jpeg=524566,
+    image_jutvision=524567,
+    image_naplps=524568,
+    image_pict=524569,
+    image_png=524570,
+    image_svg=524571 | 0x80000000,
+    image_svg_xml=524572 | 0x80000000,
+    image_tiff=524573,
+    image_vnd_adobe_photoshop=524574 | 0x80000000,
+    image_vnd_djvu=524575 | 0x80000000,
+    image_vnd_fpx=524576,
+    image_vnd_microsoft_icon=524577,
+    image_vnd_rn_realflash=524578,
+    image_vnd_rn_realpix=524579,
+    image_vnd_wap_wbmp=524580,
+    image_vnd_xiff=524581,
+    image_webp=524582,
+    image_wmf=524583,
+    image_x_3ds=524584,
+    image_x_cmu_raster=524585,
+    image_x_cur=524586,
+    image_x_dwg=524587,
+    image_x_eps=524588,
+    image_x_exr=524589,
+    image_x_gem=524590,
+    image_x_icns=524591,
+    image_x_icon=524592 | 0x80000000,
+    image_x_jg=524593,
+    image_x_jps=524594,
+    image_x_ms_bmp=524595,
+    image_x_niff=524596,
+    image_x_pcx=524597,
+    image_x_pict=524598,
+    image_x_portable_bitmap=524599,
+    image_x_portable_graymap=524600,
+    image_x_portable_pixmap=524601,
+    image_x_quicktime=524602,
+    image_x_rgb=524603,
+    image_x_tga=524604,
+    image_x_tiff=524605,
+    image_x_win_bitmap=524606,
+    image_x_xcf=524607 | 0x80000000,
+    image_x_xpixmap=524608 | 0x80000000,
+    image_x_xwindowdump=524609,
+    message_news=196930,
+    message_rfc822=196931,
+    model_vnd_dwf=65860,
+    model_vnd_gdl=65861,
+    model_vnd_gs_gdl=65862,
+    model_vrml=65863,
+    model_x_pov=65864,
+    text_PGP=590153,
+    text_asp=590154,
+    text_css=590155,
+    text_html=590156,
+    text_javascript=590157,
+    text_mcf=590158,
+    text_pascal=590159,
+    text_plain=590160,
+    text_richtext=590161,
+    text_rtf=590162,
+    text_scriplet=590163,
+    text_tab_separated_values=590164,
+    text_troff=590165,
+    text_uri_list=590166,
+    text_vnd_abc=590167,
+    text_vnd_fmi_flexstor=590168,
+    text_vnd_wap_wml=590169,
+    text_vnd_wap_wmlscript=590170,
+    text_webviewhtml=590171,
+    text_x_Algol68=590172,
+    text_x_asm=590173,
+    text_x_audiosoft_intra=590174,
+    text_x_awk=590175,
+    text_x_bcpl=590176,
+    text_x_c=590177,
+    text_x_c__=590178,
+    text_x_component=590179,
+    text_x_diff=590180,
+    text_x_fortran=590181,
+    text_x_java=590182,
+    text_x_la_asf=590183,
+    text_x_lisp=590184,
+    text_x_m=590185,
+    text_x_m4=590186,
+    text_x_makefile=590187,
+    text_x_ms_regedit=590188,
+    text_x_msdos_batch=590189,
+    text_x_objective_c=590190,
+    text_x_pascal=590191,
+    text_x_perl=590192,
+    text_x_php=590193,
+    text_x_po=590194,
+    text_x_python=590195,
+    text_x_ruby=590196,
+    text_x_sass=590197,
+    text_x_scss=590198,
+    text_x_server_parsed_html=590199,
+    text_x_setext=590200,
+    text_x_sgml=590201,
+    text_x_shellscript=590202,
+    text_x_speech=590203,
+    text_x_tcl=590204,
+    text_x_tex=590205,
+    text_x_uil=590206,
+    text_x_uuencode=590207,
+    text_x_vcalendar=590208,
+    text_x_vcard=590209,
+    text_xml=590210,
+    video_MP2T=393603,
+    video_animaflex=393604,
+    video_avi=393605,
+    video_avs_video=393606,
+    video_mp4=393607,
+    video_mpeg=393608,
+    video_quicktime=393609,
+    video_vdo=393610,
+    video_vivo=393611,
+    video_vnd_rn_realvideo=393612,
+    video_vosaic=393613,
+    video_webm=393614,
+    video_x_amt_demorun=393615,
+    video_x_amt_showrun=393616,
+    video_x_atomic3d_feature=393617,
+    video_x_dl=393618,
+    video_x_dv=393619,
+    video_x_fli=393620,
+    video_x_flv=393621,
+    video_x_isvideo=393622,
+    video_x_jng=393623 | 0x80000000,
+    video_x_m4v=393624,
+    video_x_matroska=393625,
+    video_x_mng=393626,
+    video_x_motion_jpeg=393627,
+    video_x_ms_asf=393628,
+    video_x_msvideo=393629,
+    video_x_qtc=393630,
+    video_x_sgi_movie=393631,
+    x_epoc_x_sisx_app=721312,
 };
 char *mime_get_mime_text(unsigned int mime_id) {switch (mime_id) {
 case application_arj: return "application/arj";
@@ -832,6 +836,10 @@ case audio_x_hx_aac_adts: return "audio/x-hx-aac-adts";
 case application_x_chrome_extension: return "application/x-chrome-extension";
 case image_heic: return "image/heic";
 case image_x_gem: return "image/x-gem";
+case application_x_lzma: return "application/x-lzma";
+case application_warc: return "application/warc";
+case application_x_lzip: return "application/x-lzip";
+case application_x_lzop: return "application/x-lzop";
 default: return NULL;}}
 GHashTable *mime_get_ext_table() {GHashTable *ext_table = g_hash_table_new(g_str_hash, g_str_equal);
 g_hash_table_insert(ext_table, "arj", (gpointer)application_arj);
@@ -1182,6 +1190,11 @@ g_hash_table_insert(ext_table, "d", (gpointer)text_plain);
 g_hash_table_insert(ext_table, "cs", (gpointer)text_plain);
 g_hash_table_insert(ext_table, "hpp", (gpointer)text_plain);
 g_hash_table_insert(ext_table, "srt", (gpointer)text_plain);
+g_hash_table_insert(ext_table, "nfo", (gpointer)text_plain);
+g_hash_table_insert(ext_table, "sfv", (gpointer)text_plain);
+g_hash_table_insert(ext_table, "m3u", (gpointer)text_plain);
+g_hash_table_insert(ext_table, "csv", (gpointer)text_plain);
+g_hash_table_insert(ext_table, "eml", (gpointer)text_plain);
 g_hash_table_insert(ext_table, "rt", (gpointer)text_richtext);
 g_hash_table_insert(ext_table, "rtf", (gpointer)text_richtext);
 g_hash_table_insert(ext_table, "rtx", (gpointer)text_richtext);
@@ -1334,6 +1347,10 @@ g_hash_table_insert(ext_table, "z", (gpointer)application_zlib);
 g_hash_table_insert(ext_table, "pf2", (gpointer)application_x_font_pf2);
 g_hash_table_insert(ext_table, "jmod", (gpointer)application_x_java_jmod);
 g_hash_table_insert(ext_table, "heic", (gpointer)image_heic);
+g_hash_table_insert(ext_table, "lzma", (gpointer)application_x_lzma);
+g_hash_table_insert(ext_table, "warc", (gpointer)application_warc);
+g_hash_table_insert(ext_table, "lz", (gpointer)application_x_lzip);
+g_hash_table_insert(ext_table, "lzo", (gpointer)application_x_lzop);
 return ext_table;}
 GHashTable *mime_get_mime_table() {GHashTable *mime_table = g_hash_table_new(g_str_hash, g_str_equal);
 g_hash_table_insert(mime_table, "application/arj", (gpointer)application_arj);
@@ -1748,5 +1765,9 @@ g_hash_table_insert(mime_table, "audio/x-hx-aac-adts", (gpointer)audio_x_hx_aac_
 g_hash_table_insert(mime_table, "application/x-chrome-extension", (gpointer)application_x_chrome_extension);
 g_hash_table_insert(mime_table, "image/heic", (gpointer)image_heic);
 g_hash_table_insert(mime_table, "image/x-gem", (gpointer)image_x_gem);
+g_hash_table_insert(mime_table, "application/x-lzma", (gpointer)application_x_lzma);
+g_hash_table_insert(mime_table, "application/warc", (gpointer)application_warc);
+g_hash_table_insert(mime_table, "application/x-lzip", (gpointer)application_x_lzip);
+g_hash_table_insert(mime_table, "application/x-lzop", (gpointer)application_x_lzop);
 return mime_table;}
 #endif

src/parsing/parse.c

@@ -3,7 +3,28 @@
 
 __thread magic_t Magic = NULL;
 
-void *read_all(parse_job_t *job, const char *buf, int bytes_read, int *fd) {
+int fs_read(struct vfile *f, void *buf, size_t size) {
+
+    if (f->fd == -1) {
+        f->fd = open(f->filepath, O_RDONLY);
+        if (f->fd == -1) {
+            LOG_ERRORF(f->filepath, "open(): [%d] %s", errno, strerror(errno))
+            return -1;
+        }
+    }
+
+    return read(f->fd, buf, size);
+}
+
+#define CLOSE_FILE(f) if (f.close != NULL) {f.close(&f);};
+
+void fs_close(struct vfile *f) {
+    if (f->fd != -1) {
+        close(f->fd);
+    }
+}
+
+void *read_all(parse_job_t *job, const char *buf, int bytes_read) {
 
     void *full_buf;
 
@@ -11,19 +32,12 @@ void *read_all(parse_job_t *job, const char *buf, int bytes_read, int *fd) {
         full_buf = malloc(job->info.st_size);
         memcpy(full_buf, buf, job->info.st_size);
     } else {
-        if (*fd == -1) {
-            *fd = open(job->filepath, O_RDONLY);
-            if (*fd == -1) {
-                perror("open");
-                printf("%s\n", job->filepath);
-                return NULL;
-            }
-        }
         full_buf = malloc(job->info.st_size);
         memcpy(full_buf, buf, bytes_read);
-        int ret = read(*fd, full_buf + bytes_read, job->info.st_size - bytes_read);
+
+        int ret = job->vfile.read(&job->vfile, full_buf + bytes_read, job->info.st_size - bytes_read);
         if (ret == -1) {
-            perror("read");
+            LOG_ERRORF(job->filepath, "read(): [%d] %s", errno, strerror(errno))
             return NULL;
         }
     }
@@ -36,15 +50,14 @@ void parse(void *arg) {
     parse_job_t *job = arg;
     document_t doc;
 
-    if (incremental_get(ScanCtx.original_table, job->info.st_ino) == job->info.st_mtim.tv_sec) {
+    int inc_ts = incremental_get(ScanCtx.original_table, job->info.st_ino);
+    if (inc_ts != 0 && inc_ts == job->info.st_mtim.tv_sec) {
         incremental_mark_file_for_copy(ScanCtx.copy_table, job->info.st_ino);
-        free(job);
         return;
     }
 
     if (Magic == NULL) {
         Magic = magic_open(MAGIC_MIME_TYPE);
-        magic_load(Magic, NULL);
     }
 
     doc.filepath = job->filepath;
@@ -60,30 +73,26 @@ void parse(void *arg) {
     uuid_generate(doc.uuid);
     char *buf[PARSE_BUF_SIZE];
 
+    if (LogCtx.very_verbose) {
+        char uuid_str[UUID_STR_LEN];
+        uuid_unparse(doc.uuid, uuid_str);
+        LOG_DEBUGF(job->filepath, "Starting parse job {%s}", uuid_str)
+    }
+
     if (job->info.st_size == 0) {
         doc.mime = MIME_EMPTY;
     } else if (*(job->filepath + job->ext) != '\0' && (job->ext - job->base != 1)) {
         doc.mime = mime_get_mime_by_ext(ScanCtx.ext_table, job->filepath + job->ext);
     }
 
-    int fd = -1;
     int bytes_read = 0;
 
     if (doc.mime == 0) {
         // Get mime type with libmagic
-        fd = open(job->filepath, O_RDONLY);
-        if (fd == -1) {
-            perror("open");
-            free(job);
-            return;
-        }
-
-        bytes_read = read(fd, buf, PARSE_BUF_SIZE);
-
+        bytes_read = job->vfile.read(&job->vfile, buf, PARSE_BUF_SIZE);
         if (bytes_read == -1) {
-            perror("read");
-            close(fd);
-            free(job);
+            LOG_WARNINGF(job->filepath, "read() Error: %s", strerror(errno))
+            CLOSE_FILE(job->vfile)
             return;
         }
 
@@ -91,7 +100,7 @@ void parse(void *arg) {
         if (magic_mime_str != NULL) {
             doc.mime = mime_get_mime_by_string(ScanCtx.mime_table, magic_mime_str);
             if (doc.mime == 0) {
-                fprintf(stderr, "Couldn't find mime %s, %s\n", magic_mime_str, job->filepath + job->base);
+                LOG_WARNINGF(job->filepath, "Couldn't find mime %s", magic_mime_str);
             }
         }
     }
@@ -101,11 +110,16 @@ void parse(void *arg) {
     if (!(SHOULD_PARSE(doc.mime))) {
 
     } else if ((mmime == MimeVideo && doc.size >= MIN_VIDEO_SIZE) ||
-    (mmime == MimeImage && doc.size >= MIN_IMAGE_SIZE) || mmime == MimeAudio) {
-        parse_media(job->filepath, &doc);
+               (mmime == MimeImage && doc.size >= MIN_IMAGE_SIZE) || mmime == MimeAudio) {
+
+        if (job->vfile.is_fs_file) {
+            parse_media_filename(job->filepath, &doc);
+        } else {
+            parse_media_vfile(&job->vfile, &doc);
+        }
 
     } else if (IS_PDF(doc.mime)) {
-        void *pdf_buf = read_all(job, (char *) buf, bytes_read, &fd);
+        void *pdf_buf = read_all(job, (char *) buf, bytes_read);
         parse_pdf(pdf_buf, doc.size, &doc);
 
         if (pdf_buf != buf && pdf_buf != NULL) {
@@ -113,22 +127,42 @@ void parse(void *arg) {
         }
 
     } else if (mmime == MimeText && ScanCtx.content_size > 0) {
-        parse_text(bytes_read, &fd, (char *) buf, &doc);
+        parse_text(bytes_read, &job->vfile, (char *) buf, &doc);
 
     } else if (IS_FONT(doc.mime)) {
-        void *font_buf = read_all(job, (char *) buf, bytes_read, &fd);
+        void *font_buf = read_all(job, (char *) buf, bytes_read);
         parse_font(font_buf, doc.size, &doc);
 
         if (font_buf != buf && font_buf != NULL) {
             free(font_buf);
         }
+    } else if (
+            ScanCtx.archive_mode != ARC_MODE_SKIP && (
+                    IS_ARC(doc.mime) ||
+                    (IS_ARC_FILTER(doc.mime) && should_parse_filtered_file(doc.filepath, doc.ext))
+            )) {
+        parse_archive(&job->vfile, &doc);
+    } else if (ScanCtx.content_size > 0 && IS_DOC(doc.mime)) {
+        void *doc_buf = read_all(job, (char *) buf, bytes_read);
+        parse_doc(doc_buf, doc.size, &doc);
+
+        if (doc_buf != buf && doc_buf != NULL) {
+            free(doc_buf);
+        }
+    }
+
+    //Parent meta
+    if (!uuid_is_null(job->parent)) {
+        char tmp[UUID_STR_LEN];
+        uuid_unparse(job->parent, tmp);
+
+        meta_line_t *meta_parent = malloc(sizeof(meta_line_t) + UUID_STR_LEN + 1);
+        meta_parent->key = MetaParent;
+        strcpy(meta_parent->strval, tmp);
+        APPEND_META((&doc), meta_parent)
     }
 
     write_document(&doc);
 
-    if (fd != -1) {
-        close(fd);
-    }
-
-    free(job);
+    CLOSE_FILE(job->vfile)
 }

src/parsing/parse.h

@@ -5,6 +5,9 @@
 
 #define PARSE_BUF_SIZE 4096
 
+int fs_read(struct vfile *f, void *buf, size_t size);
+void fs_close(struct vfile *f);
+
 void parse(void *arg);
 
 #endif

src/parsing/pdf.c

@@ -1,6 +1,10 @@
 #include "pdf.h"
 #include "src/ctx.h"
 
+#define MIN_OCR_SIZE 128
+__thread text_buffer_t thread_buffer;
+
+
 fz_page *render_cover(fz_context *ctx, document_t *doc, fz_document *fzdoc) {
 
     int err = 0;
@@ -14,6 +18,7 @@ fz_page *render_cover(fz_context *ctx, document_t *doc, fz_document *fzdoc) {
 
     if (err != 0) {
         fz_drop_page(ctx, cover);
+        LOG_WARNINGF(doc->filepath, "fz_load_page() returned error code [%d] %s", err, ctx->error.message)
         return NULL;
     }
 
@@ -52,6 +57,7 @@ fz_page *render_cover(fz_context *ctx, document_t *doc, fz_document *fzdoc) {
         err = ctx->error.errcode;
 
     if (err != 0) {
+        LOG_WARNINGF(doc->filepath, "fz_run_page() returned error code [%d] %s", err, ctx->error.message)
         fz_drop_page(ctx, cover);
         fz_drop_pixmap(ctx, pixmap);
         return NULL;
@@ -76,6 +82,8 @@ fz_page *render_cover(fz_context *ctx, document_t *doc, fz_document *fzdoc) {
     fz_drop_pixmap(ctx, pixmap);
 
     if (err != 0) {
+        LOG_WARNINGF(doc->filepath, "fz_new_buffer_from_pixmap_as_png() returned error code [%d] %s", err,
+                     ctx->error.message)
         fz_drop_page(ctx, cover);
         return NULL;
     }
@@ -83,14 +91,22 @@ fz_page *render_cover(fz_context *ctx, document_t *doc, fz_document *fzdoc) {
     return cover;
 }
 
-void fz_noop_callback(__attribute__((unused)) void *user, __attribute__((unused)) const char *message) {}
+void fz_err_callback(void *user, UNUSED(const char *message)) {
+    if (LogCtx.verbose) {
+        document_t *doc = (document_t *) user;
+        LOG_WARNINGF(doc->filepath, "FZ: %s", message)
+    }
+}
 
-
-void init_ctx(fz_context *ctx) {
+__always_inline
+void init_ctx(fz_context *ctx, document_t *doc) {
     fz_disable_icc(ctx);
     fz_register_document_handlers(ctx);
-    ctx->warn.print = fz_noop_callback;
-    ctx->error.print = fz_noop_callback;
+
+    ctx->warn.print_user = doc;
+    ctx->warn.print = fz_err_callback;
+    ctx->error.print_user = doc;
+    ctx->error.print = fz_err_callback;
 }
 
 int read_stext_block(fz_stext_block *block, text_buffer_t *tex) {
@@ -112,6 +128,40 @@ int read_stext_block(fz_stext_block *block, text_buffer_t *tex) {
     return 0;
 }
 
+
+void fill_image(fz_context *ctx, UNUSED(fz_device *dev),
+                fz_image *img, UNUSED(fz_matrix ctm), UNUSED(float alpha),
+                UNUSED(fz_color_params color_params)) {
+
+    int l2factor = 0;
+
+    if (img->w > MIN_OCR_SIZE && img->h > MIN_OCR_SIZE) {
+
+        fz_pixmap *pix = img->get_pixmap(ctx, img, NULL, img->w, img->h, &l2factor);
+
+        if (pix->h > MIN_OCR_SIZE && img->h > MIN_OCR_SIZE && img->xres != 0) {
+            TessBaseAPI *api = TessBaseAPICreate();
+            TessBaseAPIInit3(api, ScanCtx.tesseract_path, ScanCtx.tesseract_lang);
+
+            TessBaseAPISetImage(api, pix->samples, pix->w, pix->h, pix->n, pix->stride);
+            TessBaseAPISetSourceResolution(api, pix->xres);
+
+            char *text = TessBaseAPIGetUTF8Text(api);
+            size_t len = strlen(text);
+            text_buffer_append_string(&thread_buffer, text, len - 1);
+            LOG_DEBUGF(
+                    "pdf.c",
+                    "(OCR) %dx%d got %dB from tesseract (%s), buffer:%dB",
+                    pix->w, pix->h, len, ScanCtx.tesseract_lang, thread_buffer.dyn_buffer.cur
+            )
+
+            TessBaseAPIEnd(api);
+            TessBaseAPIDelete(api);
+        }
+        fz_drop_pixmap(ctx, pix);
+    }
+}
+
 void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
 
     if (buf == NULL) {
@@ -125,7 +175,7 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
     }
     fz_context *ctx = fz_new_context(NULL, NULL, FZ_STORE_UNLIMITED);
 
-    init_ctx(ctx);
+    init_ctx(ctx, doc);
 
     int err = 0;
 
@@ -171,13 +221,24 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
         err = ctx->error.errcode;
 
     if (err) {
+        LOG_WARNINGF(doc->filepath, "fz_count_pages() returned error code [%d] %s", err, ctx->error.message)
         fz_drop_stream(ctx, stream);
         fz_drop_document(ctx, fzdoc);
         fz_drop_context(ctx);
         return;
     }
 
-    fz_page *cover = render_cover(ctx, doc, fzdoc);
+    fz_page *cover = NULL;
+    if (ScanCtx.tn_size > 0) {
+        cover = render_cover(ctx, doc, fzdoc);
+    } else {
+        fz_var(cover);
+        fz_try(ctx)
+            cover = fz_load_page(ctx, fzdoc, 0);
+        fz_catch(ctx)
+            cover = NULL;
+    }
+
     if (cover == NULL) {
         fz_drop_stream(ctx, stream);
         fz_drop_document(ctx, fzdoc);
@@ -185,79 +246,91 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
         return;
     }
 
-    fz_stext_options opts = {0};
-    text_buffer_t text_buf = text_buffer_create(ScanCtx.content_size);
+    if (ScanCtx.content_size > 0) {
+        fz_stext_options opts = {0};
+        thread_buffer = text_buffer_create(ScanCtx.content_size);
+
+        for (int current_page = 0; current_page < page_count; current_page++) {
+            fz_page *page = NULL;
+            if (current_page == 0) {
+                page = cover;
+            } else {
+                fz_var(err);
+                fz_try(ctx)
+                            page = fz_load_page(ctx, fzdoc, current_page);
+                fz_catch(ctx)
+                    err = ctx->error.errcode;
+                if (err != 0) {
+                    LOG_WARNINGF(doc->filepath, "fz_load_page() returned error code [%d] %s", err, ctx->error.message)
+                    text_buffer_destroy(&thread_buffer);
+                    fz_drop_page(ctx, page);
+                    fz_drop_stream(ctx, stream);
+                    fz_drop_document(ctx, fzdoc);
+                    fz_drop_context(ctx);
+                    return;
+                }
+            }
+
+            fz_stext_page *stext = fz_new_stext_page(ctx, fz_bound_page(ctx, page));
+            fz_device *dev = fz_new_stext_device(ctx, stext, &opts);
+            dev->stroke_path = NULL;
+            dev->stroke_text = NULL;
+            dev->clip_text = NULL;
+            dev->clip_stroke_path = NULL;
+            dev->clip_stroke_text = NULL;
+
+            if (ScanCtx.tesseract_lang != NULL) {
+                dev->fill_image = fill_image;
+            }
 
-    for (int current_page = 0; current_page < page_count; current_page++) {
-        fz_page *page = NULL;
-        if (current_page == 0) {
-            page = cover;
-        } else {
             fz_var(err);
             fz_try(ctx)
-                page = fz_load_page(ctx, fzdoc, current_page);
+                        fz_run_page(ctx, page, dev, fz_identity, NULL);
+            fz_always(ctx)
+                {
+                    fz_close_device(ctx, dev);
+                    fz_drop_device(ctx, dev);
+                }
             fz_catch(ctx)
                 err = ctx->error.errcode;
+
             if (err != 0) {
-                text_buffer_destroy(&text_buf);
+                LOG_WARNINGF(doc->filepath, "fz_run_page() returned error code [%d] %s", err, ctx->error.message)
+                text_buffer_destroy(&thread_buffer);
                 fz_drop_page(ctx, page);
+                fz_drop_stext_page(ctx, stext);
                 fz_drop_stream(ctx, stream);
                 fz_drop_document(ctx, fzdoc);
                 fz_drop_context(ctx);
                 return;
             }
-        }
 
-        fz_stext_page *stext = fz_new_stext_page(ctx, fz_bound_page(ctx, page));
-        fz_device *dev = fz_new_stext_device(ctx, stext, &opts);
-
-        fz_var(err);
-        fz_try(ctx)
-            fz_run_page(ctx, page, dev, fz_identity, NULL);
-        fz_always(ctx)
-        {
-            fz_close_device(ctx, dev);
-            fz_drop_device(ctx, dev);
-        }
-        fz_catch(ctx)
-            err = ctx->error.errcode;
-
-        if (err != 0) {
-            text_buffer_destroy(&text_buf);
-            fz_drop_page(ctx, page);
+            fz_stext_block *block = stext->first_block;
+            while (block != NULL) {
+                int ret = read_stext_block(block, &thread_buffer);
+                if (ret == TEXT_BUF_FULL) {
+                    break;
+                }
+                block = block->next;
+            }
             fz_drop_stext_page(ctx, stext);
-            fz_drop_stream(ctx, stream);
-            fz_drop_document(ctx, fzdoc);
-            fz_drop_context(ctx);
-            return;
-        }
+            fz_drop_page(ctx, page);
 
-        fz_stext_block *block = stext->first_block;
-        while (block != NULL) {
-            int ret = read_stext_block(block, &text_buf);
-            if (ret == TEXT_BUF_FULL) {
+            if (thread_buffer.dyn_buffer.cur >= thread_buffer.dyn_buffer.size) {
                 break;
             }
-            block = block->next;
         }
-        fz_drop_stext_page(ctx, stext);
-        fz_drop_page(ctx, page);
+        text_buffer_terminate_string(&thread_buffer);
 
-        if (text_buf.dyn_buffer.cur >= text_buf.dyn_buffer.size) {
-            break;
-        }
+        meta_line_t *meta_content = malloc(sizeof(meta_line_t) + thread_buffer.dyn_buffer.cur);
+        meta_content->key = MetaContent;
+        memcpy(meta_content->strval, thread_buffer.dyn_buffer.buf, thread_buffer.dyn_buffer.cur);
+        APPEND_META(doc, meta_content)
+
+        text_buffer_destroy(&thread_buffer);
     }
-    text_buffer_terminate_string(&text_buf);
-
-    meta_line_t *meta_content = malloc(sizeof(meta_line_t) + text_buf.dyn_buffer.cur);
-    meta_content->key = MetaContent;
-    memcpy(meta_content->strval, text_buf.dyn_buffer.buf, text_buf.dyn_buffer.cur);
-    APPEND_META(doc, meta_content)
 
     fz_drop_stream(ctx, stream);
     fz_drop_document(ctx, fzdoc);
     fz_drop_context(ctx);
-
-    text_buffer_destroy(&text_buf);
 }
-

src/parsing/text.c

@@ -1,7 +1,7 @@
 #include "text.h"
 #include "src/ctx.h"
 
-void parse_text(int bytes_read, int *fd, char *buf, document_t *doc) {
+void parse_text(int bytes_read, struct vfile *f, char *buf, document_t *doc) {
 
     char *intermediate_buf;
     int intermediate_buf_len;
@@ -13,10 +13,6 @@ void parse_text(int bytes_read, int *fd, char *buf, document_t *doc) {
         memcpy(intermediate_buf, buf, to_copy);
 
     } else {
-        if (*fd == -1) {
-            *fd = open(doc->filepath, O_RDONLY);
-        }
-
         int to_read = MIN(ScanCtx.content_size, doc->size) - bytes_read;
 
         intermediate_buf = malloc(to_read + bytes_read);
@@ -25,10 +21,11 @@ void parse_text(int bytes_read, int *fd, char *buf, document_t *doc) {
             memcpy(intermediate_buf, buf, bytes_read);
         }
 
-        read(*fd, intermediate_buf + bytes_read, to_read);
+        f->read(f, intermediate_buf + bytes_read, to_read);
     }
     text_buffer_t tex = text_buffer_create(ScanCtx.content_size);
     text_buffer_append_string(&tex, intermediate_buf, intermediate_buf_len);
+    text_buffer_terminate_string(&tex);
 
     meta_line_t *meta = malloc(sizeof(meta_line_t) + tex.dyn_buffer.cur);
     meta->key = MetaContent;

src/parsing/text.h

@@ -3,6 +3,6 @@
 
 #include "src/sist.h"
 
-void parse_text(int bytes_read, int *fd, char *buf, document_t *doc);
+void parse_text(int bytes_read, struct vfile *f, char *buf, document_t *doc);
 
 #endif

src/sist.h

@@ -2,6 +2,7 @@
 #define SIST_H
 
 #define UUID_STR_LEN 37
+#define UNUSED(x) __attribute__((__unused__))  x
 
 #include <glib-2.0/glib.h>
 #include <unistd.h>
@@ -12,11 +13,11 @@
 #include <ftw.h>
 #include <uuid.h>
 #include <magic.h>
-#include <libavformat/avformat.h>
-#include <libswscale/swscale.h>
-#include <libswresample/swresample.h>
-#include <libavcodec/avcodec.h>
-#include <libavutil/imgutils.h>
+#include "libavformat/avformat.h"
+#include "libswscale/swscale.h"
+#include "libswresample/swresample.h"
+#include "libavcodec/avcodec.h"
+#include "libavutil/imgutils.h"
 #include <ctype.h>
 #include <mupdf/fitz.h>
 #include <mupdf/pdf.h>
@@ -26,14 +27,21 @@
 #include <pthread.h>
 #include <sys/stat.h>
 #include <wordexp.h>
+#include "ft2build.h"
+#include "freetype/freetype.h"
+#include <archive.h>
+#include <archive_entry.h>
+#include <opc/opc.h>
+#include <libxml/xmlstring.h>
+#define BOOL int
+#include <tesseract/capi.h>
 
-#ifndef SIST_SCAN_ONLY
 #include <onion/onion.h>
 #include <onion/handler.h>
 #include <onion/block.h>
 #include <onion/shortcuts.h>
+#include <onion/codecs.h>
 #include <curl/curl.h>
-#endif
 
 #include "cJSON/cJSON.h"
 
@@ -49,14 +57,16 @@
 #include "parsing/pdf.h"
 #include "parsing/media.h"
 #include "parsing/font.h"
+#include "parsing/arc.h"
+#include "parsing/doc.h"
 #include "cli.h"
+#include "log.h"
 #include "utf8.h/utf8.h"
 
-#ifndef SIST_SCAN_ONLY
 #include "src/index/elastic.h"
 #include "index/web.h"
 #include "web/serve.h"
-#endif
+#include "web/auth_basic.h"
 
 ;
 

src/tpool.c

@@ -114,12 +114,19 @@ static void *tpool_worker(void *arg) {
         pthread_mutex_unlock(&(pool->work_mutex));
 
         if (work != NULL) {
+            if (pool->stop) {
+                break;
+            }
+
             work->func(work->arg);
+            free(work->arg);
             free(work);
         }
 
         pthread_mutex_lock(&(pool->work_mutex));
-        pool->done_cnt++;
+        if (work != NULL) {
+            pool->done_cnt++;
+        }
 
         progress_bar_print((double) pool->done_cnt / pool->work_cnt, ScanCtx.stat_tn_size, ScanCtx.stat_index_size);
 
@@ -129,6 +136,7 @@ static void *tpool_worker(void *arg) {
         pthread_mutex_unlock(&(pool->work_mutex));
     }
 
+    LOG_INFO("tpool.c", "Executing cleaup function")
     pool->cleanup_func();
 
     pthread_cond_signal(&(pool->working_cond));
@@ -137,17 +145,24 @@ static void *tpool_worker(void *arg) {
 }
 
 void tpool_wait(tpool_t *pool) {
+    LOG_INFO("tpool.c", "Waiting for worker threads to finish")
     pthread_mutex_lock(&(pool->work_mutex));
     while (1) {
         if (pool->done_cnt < pool->work_cnt) {
             pthread_cond_wait(&(pool->working_cond), &(pool->work_mutex));
         } else {
-            pool->stop = 1;
-            break;
+            usleep(500000);
+            if (pool->done_cnt == pool->work_cnt) {
+                pool->stop = 1;
+                usleep(1000000);
+                break;
+            }
         }
-        progress_bar_print(100.0, ScanCtx.stat_tn_size, ScanCtx.stat_index_size);
     }
+    progress_bar_print(1.0, ScanCtx.stat_tn_size, ScanCtx.stat_index_size);
     pthread_mutex_unlock(&(pool->work_mutex));
+
+    LOG_INFO("tpool.c", "Worker threads finished")
 }
 
 void tpool_destroy(tpool_t *pool) {
@@ -155,6 +170,8 @@ void tpool_destroy(tpool_t *pool) {
         return;
     }
 
+    LOG_INFO("tpool.c", "Destroying thread pool")
+
     pthread_mutex_lock(&(pool->work_mutex));
     tpool_work_t *work = pool->work_head;
     while (work != NULL) {
@@ -169,10 +186,13 @@ void tpool_destroy(tpool_t *pool) {
     for (size_t i = 0; i < pool->thread_cnt; i++) {
         pthread_t thread = pool->threads[i];
         if (thread != 0) {
-            pthread_cancel(thread);
+            void *_;
+            pthread_join(thread, &_);
         }
     }
 
+    LOG_INFO("tpool.c", "Final cleanup")
+
     pthread_mutex_destroy(&(pool->work_mutex));
     pthread_cond_destroy(&(pool->has_work_cond));
     pthread_cond_destroy(&(pool->working_cond));
@@ -208,9 +228,9 @@ tpool_t *tpool_create(size_t thread_cnt, void cleanup_func()) {
 
 void tpool_start(tpool_t *pool) {
 
+    LOG_INFOF("tpool.c", "Starting thread pool with %d threads", pool->thread_cnt)
+
     for (size_t i = 0; i < pool->thread_cnt; i++) {
-        pthread_t thread = pool->threads[i];
-        pthread_create(&thread, NULL, tpool_worker, pool);
-        pthread_detach(thread);
+        pthread_create(&pool->threads[i], NULL, tpool_worker, pool);
     }
 }

src/types.h

@@ -9,6 +9,12 @@
 #define IS_META_LONG(key) (key & META_LONG_MASK) == META_LONG_MASK
 #define IS_META_STR(meta) (meta->key & META_STR_MASK) == META_STR_MASK
 
+#define ARC_MODE_SKIP 0
+#define ARC_MODE_LIST 1
+#define ARC_MODE_SHALLOW 2
+#define ARC_MODE_RECURSE 3
+typedef int archive_mode_t;
+
 // This is written to file as a 8bit char!
 enum metakey {
     MetaContent = 1 | META_STR_MASK,
@@ -24,8 +30,13 @@ enum metakey {
     MetaGenre = 11 | META_STR_MASK,
     MetaTitle = 12 | META_STR_MASK,
     MetaFontName = 13 | META_STR_MASK,
+    MetaParent = 14 | META_STR_MASK,
 };
 
+#define INDEX_TYPE_BIN "binary"
+#define INDEX_TYPE_JSON "json"
+#define INDEX_VERSION_EXTERNAL "_external_v1"
+
 typedef struct index_descriptor {
     char uuid[UUID_STR_LEN];
     char version[6];
@@ -34,6 +45,7 @@ typedef struct index_descriptor {
     char rewrite_url[8196];
     short root_len;
     char name[1024];
+    char type[64];
 } index_descriptor_t;
 
 typedef struct index_t {
@@ -66,10 +78,32 @@ typedef struct document {
     char *filepath;
 } document_t;
 
+typedef struct vfile vfile_t;
+
+typedef int (*read_func_t)(struct vfile *, void *buf, size_t size);
+
+typedef void (*close_func_t)(struct vfile *);
+
+typedef struct vfile {
+
+    union {
+        int fd;
+        struct archive *arc;
+    };
+
+    int is_fs_file;
+    char *filepath;
+
+    read_func_t read;
+    close_func_t close;
+} vfile_t;
+
 typedef struct parse_job_t {
     int base;
     int ext;
     struct stat info;
+    struct vfile vfile;
+    uuid_t parent;
     char filepath[1];
 } parse_job_t;
 

src/util.c

@@ -1,4 +1,5 @@
 #include "util.h"
+#include "src/ctx.h"
 
 dyn_buffer_t dyn_buffer_create() {
     dyn_buffer_t buf;
@@ -46,6 +47,10 @@ void dyn_buffer_write_str(dyn_buffer_t *buf, char *str) {
     dyn_buffer_write_char(buf, '\0');
 }
 
+void dyn_buffer_append_string(dyn_buffer_t *buf, char *str) {
+    dyn_buffer_write(buf, str, strlen(str));
+}
+
 void dyn_buffer_write_int(dyn_buffer_t *buf, int d) {
     grow_buffer_small(buf);
 
@@ -145,7 +150,6 @@ int text_buffer_append_string(text_buffer_t *buf, char *str, size_t len) {
         (0xe0 == (0xf0 & str[0]) && len < 3) ||
         (0xc0 == (0xe0 & str[0]) && len == 1) ||
         *(str) == 0) {
-        text_buffer_terminate_string(buf);
         return 0;
     }
 
@@ -154,7 +158,6 @@ int text_buffer_append_string(text_buffer_t *buf, char *str, size_t len) {
             text_buffer_append_char(buf, c);
         }
     }
-    text_buffer_terminate_string(buf);
     return 0;
 }
 
@@ -165,7 +168,6 @@ int text_buffer_append_string0(text_buffer_t *buf, char *str) {
             text_buffer_append_char(buf, c);
         }
     }
-    text_buffer_terminate_string(buf);
 }
 
 int text_buffer_append_char(text_buffer_t *buf, int c) {
@@ -316,4 +318,29 @@ GHashTable *incremental_get_table() {
     return file_table;
 }
 
+const char *find_file_in_paths(const char *paths[], const char *filename) {
+
+    for (int i = 0; paths[i] != NULL; i++) {
+
+        char *apath = abspath(paths[i]);
+        if (apath == NULL) {
+            continue;
+        }
+
+        char path[PATH_MAX];
+        snprintf(path, sizeof(path), "%s%s", apath, filename);
+
+        LOG_DEBUGF("util.c", "Looking for '%s' in folder '%s'", filename, apath)
+        free(apath);
+
+        struct stat info;
+        int ret = stat(path, &info);
+        if (ret != -1) {
+            return paths[i];
+        }
+    }
+
+    return NULL;
+}
+
 

src/util.h

@@ -7,7 +7,7 @@
 #define INITIAL_BUF_SIZE 1024 * 16
 
 #define SHOULD_IGNORE_CHAR(c) !(SHOULD_KEEP_CHAR(c))
-#define SHOULD_KEEP_CHAR(c) (c >= (int)'!')
+#define SHOULD_KEEP_CHAR(c) ((c >= '0' && c <= '9') || (c >= 'A' && c <= 'z') || (c > 127))
 
 
 typedef struct dyn_buffer {
@@ -47,6 +47,8 @@ void dyn_buffer_write_char(dyn_buffer_t *buf, char c);
 
 void dyn_buffer_write_str(dyn_buffer_t *buf, char *str);
 
+void dyn_buffer_append_string(dyn_buffer_t *buf, char *str);
+
 void dyn_buffer_write_int(dyn_buffer_t *buf, int d);
 
 void dyn_buffer_write_short(dyn_buffer_t *buf, short s);
@@ -72,5 +74,6 @@ int incremental_get(GHashTable *table, unsigned long inode_no);
 
 int incremental_mark_file_for_copy(GHashTable *table, unsigned long inode_no);
 
+const char *find_file_in_paths(const char **paths, const char *filename);
 
 #endif

src/web/auth_basic.c

@@ -0,0 +1,59 @@
+#include "auth_basic.h"
+
+#define UNAUTHORIZED_TEXT "Unauthorized"
+
+typedef struct auth_basic_data {
+    onion_handler *inside;
+    const char *b64credentials;
+} auth_basic_data_t;
+
+
+int authenticate(const char *expected, const char *credentials) {
+
+    if (expected == NULL) {
+        return TRUE;
+    }
+
+    if (credentials && strncmp(credentials, "Basic ", 6) == 0) {
+        if (strcmp((credentials + 6), expected) == 0) {
+            return TRUE;
+        }
+    }
+
+    return FALSE;
+}
+
+int auth_basic_handler(auth_basic_data_t *d,
+                       onion_request *req,
+                       onion_response *res) {
+
+    const char *credentials = onion_request_get_header(req, "Authorization");
+
+    if (authenticate(d->b64credentials, credentials)) {
+        return onion_handler_handle(d->inside, req, res);
+    }
+
+    onion_response_set_header(res, "WWW-Authenticate", "Basic realm=\"sist2\"");
+    onion_response_set_code(res, HTTP_UNAUTHORIZED);
+    onion_response_write(res, UNAUTHORIZED_TEXT, sizeof(UNAUTHORIZED_TEXT));
+    onion_response_set_length(res, sizeof(UNAUTHORIZED_TEXT));
+
+    return OCS_PROCESSED;
+}
+
+void auth_basic_free(auth_basic_data_t *data) {
+    onion_handler_free(data->inside);
+    free(data);
+}
+
+onion_handler *auth_basic(const char *b64credentials, onion_handler *inside_level) {
+
+    auth_basic_data_t *privdata = malloc(sizeof(auth_basic_data_t));
+
+    privdata->b64credentials = b64credentials;
+    privdata->inside = inside_level;
+
+    return onion_handler_new((onion_handler_handler) auth_basic_handler, privdata,
+                             (onion_handler_private_data_free) auth_basic_free);
+}
+

src/web/auth_basic.h

@@ -0,0 +1,4 @@
+#include "src/sist.h"
+
+
+onion_handler *auth_basic(const char *b64credentials, onion_handler *inside_level);

src/web/serve.c

@@ -110,7 +110,7 @@ int thumbnail(void *p, onion_request *req, onion_response *res) {
     int written = onion_response_write(res, data, data_len);
     onion_response_flush(res);
     if (written != data_len || data_len == 0) {
-        printf("Couldn't write thumb\n");
+        LOG_DEBUG("serve.c", "Couldn't write thumbnail");
     }
     free(data);
 
@@ -214,7 +214,7 @@ int chunked_response_file(const char *filename, const char *mime,
     return OCS_PROCESSED;
 }
 
-int search(void *p, onion_request *req, onion_response *res) {
+int search(UNUSED(void *p), onion_request *req, onion_response *res) {
 
     int flags = onion_request_get_flags(req);
     if ((flags & OR_METHODS) != OR_POST) {
@@ -245,6 +245,8 @@ int search(void *p, onion_request *req, onion_response *res) {
 
     if (r->status_code == 200) {
         onion_response_write(res, r->body, r->size);
+    } else {
+        onion_response_set_code(res, HTTP_INTERNAL_ERROR);
     }
 
     free_response(r);
@@ -252,7 +254,7 @@ int search(void *p, onion_request *req, onion_response *res) {
     return OCS_PROCESSED;
 }
 
-int scroll(void *p, onion_request *req, onion_response *res) {
+int scroll(UNUSED(void *p), onion_request *req, onion_response *res) {
 
     int flags = onion_request_get_flags(req);
     if ((flags & OR_METHODS) != OR_GET) {
@@ -325,7 +327,7 @@ int serve_file_from_disk(cJSON *json, index_t *idx, onion_request *req, onion_re
     return chunked_response_file(full_path, mime, 1, req, res);
 }
 
-int index_info(void *p, onion_request *req, onion_response *res) {
+int index_info(UNUSED(void *p), onion_request *req, onion_response *res) {
     cJSON *json = cJSON_CreateObject();
     cJSON *arr = cJSON_AddArrayToObject(json, "indices");
 
@@ -351,7 +353,8 @@ int index_info(void *p, onion_request *req, onion_response *res) {
     return OCS_PROCESSED;
 }
 
-int file(void *p, onion_request *req, onion_response *res) {
+
+int document_info(UNUSED(void *p), onion_request *req, onion_response *res) {
 
     const char *arg_uuid = onion_request_get_query(req, "1");
     if (arg_uuid == NULL) {
@@ -360,12 +363,56 @@ int file(void *p, onion_request *req, onion_response *res) {
 
     cJSON *doc = elastic_get_document(arg_uuid);
     cJSON *source = cJSON_GetObjectItem(doc, "_source");
+
     cJSON *index_id = cJSON_GetObjectItem(source, "index");
     if (index_id == NULL) {
         cJSON_Delete(doc);
         return OCS_NOT_PROCESSED;
     }
 
+    index_t *idx = get_index_by_id(index_id->valuestring);
+    if (idx == NULL) {
+        cJSON_Delete(doc);
+        return OCS_NOT_PROCESSED;
+    }
+
+    onion_response_set_header(res, "Content-Type", "application/json");
+
+    char *json_str = cJSON_PrintUnformatted(source);
+    onion_response_write0(res, json_str);
+    free(json_str);
+    cJSON_Delete(doc);
+
+    return OCS_PROCESSED;
+}
+
+int file(UNUSED(void *p), onion_request *req, onion_response *res) {
+
+    const char *arg_uuid = onion_request_get_query(req, "1");
+    if (arg_uuid == NULL) {
+        return OCS_PROCESSED;
+    }
+
+    const char *next = arg_uuid;
+    cJSON *doc = NULL;
+    cJSON *index_id = NULL;
+    cJSON *source = NULL;
+
+    while (true) {
+        doc = elastic_get_document(next);
+        source = cJSON_GetObjectItem(doc, "_source");
+        index_id = cJSON_GetObjectItem(source, "index");
+        if (index_id == NULL) {
+            cJSON_Delete(doc);
+            return OCS_NOT_PROCESSED;
+        }
+        cJSON *parent = cJSON_GetObjectItem(source, "parent");
+        if (parent == NULL) {
+            break;
+        }
+        next = parent->valuestring;
+    }
+
     index_t *idx = get_index_by_id(index_id->valuestring);
 
     if (idx == NULL) {
@@ -384,6 +431,23 @@ int file(void *p, onion_request *req, onion_response *res) {
     return ret;
 }
 
+int status(UNUSED(void *p), UNUSED(onion_request *req), onion_response *res) {
+    set_default_headers(res);
+
+    onion_response_set_header(res, "Content-Type", "application/x-empty");
+
+    char *status = elastic_get_status();
+    if (strcmp(status, "open") == 0) {
+        onion_response_set_code(res, 204);
+    } else {
+        onion_response_set_code(res, 500);
+    }
+
+    free(status);
+
+    return OCS_PROCESSED;
+}
+
 void serve(const char *hostname, const char *port) {
     onion *o = onion_new(O_POOL);
     onion_set_timeout(o, 3500);
@@ -391,9 +455,11 @@ void serve(const char *hostname, const char *port) {
     onion_set_hostname(o, hostname);
     onion_set_port(o, port);
 
-    onion_url *urls = onion_root_url(o);
+    onion_url *urls = onion_url_new();
 
     // Static paths
+    onion_set_root_handler(o, auth_basic(WebCtx.b64credentials, onion_url_to_handler(urls)));
+
     onion_url_add(urls, "", search_index);
     onion_url_add(urls, "css", style);
     onion_url_add(urls, "js", javascript);
@@ -401,6 +467,7 @@ void serve(const char *hostname, const char *port) {
 
     onion_url_add(urls, "es", search);
     onion_url_add(urls, "scroll", scroll);
+    onion_url_add(urls, "status", status);
     onion_url_add(
             urls,
             "^t/([a-fA-F0-9]{8}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{12})/"
@@ -408,8 +475,10 @@ void serve(const char *hostname, const char *port) {
             thumbnail
     );
     onion_url_add(urls, "^f/([a-fA-F0-9]{8}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{12})$", file);
+    onion_url_add(urls, "^d/([a-fA-F0-9]{8}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-[a-fA-F0-9]{12})$", document_info);
     onion_url_add(urls, "i", index_info);
 
+
     printf("Starting web server @ http://%s:%s\n", hostname, port);
 
     onion_listen(o);

web/css/dark.css

@@ -1,3 +1,44 @@
+*:focus {
+    outline: 0;
+}
+
+.info-icon {
+    width: 1rem;
+    margin-right: 0.2rem;
+    cursor: pointer;
+    color: #757575;
+    line-height: 1rem;
+    height: 1.1rem;
+}
+
+.info-icon:hover {
+    color: inherit;
+}
+
+.modal-title {
+    max-width: calc(100% - 2rem);
+    overflow: hidden;
+    text-overflow: ellipsis;
+}
+
+
+.path-row {
+    display: -ms-flexbox;
+    display: flex;
+    -ms-flex-align: start;
+    align-items: flex-start;
+}
+
+.tag-container {
+    margin-left: 0.3rem;
+}
+
+.path-line {
+    color: #BBB;
+    text-overflow: ellipsis;
+    overflow: hidden;
+}
+
 a {
     color: #00BCD4;
 }
@@ -11,7 +52,7 @@ body {
     margin-top: 1em;
 }
 
-.card {
+.card, .modal-content {
     margin-top: 1em;
     background: #212121;
     color: #e0e0e0;
@@ -19,6 +60,54 @@ body {
     border: none;
 }
 
+.table {
+    color: #e0e0e0;
+}
+
+.table td, .table th {
+    border: none;
+}
+
+.table thead th {
+    border-bottom: 1px solid #646464;
+}
+
+.modal-header .close {
+    color: #e0e0e0;
+    text-shadow: none;
+}
+
+.modal-header {
+    border-bottom: 1px solid #646464;
+}
+
+.sub-document {
+    background: #37474F !important;
+}
+
+.list-group-item.sub-document {
+    border-top: 1px solid #646464 !important;
+}
+
+.sub-document .text-muted {
+    color: #8a949c !important;
+}
+
+
+.list-group-item {
+    background: #212121;
+    color: #e0e0e0;
+
+    border-top: 1px solid #424242;
+    border-bottom: none;
+    border-left: none;
+    border-right: none;
+}
+
+.list-group-item:first-child {
+    border-top: none;
+}
+
 .navbar-brand {
     font-size: 1.75rem;
     padding: 0;
@@ -85,16 +174,25 @@ body {
 }
 
 .file-title {
+    width: 100%;
+    line-height: 1rem;
+    height: 1.1rem;
     font-size: 10pt;
     white-space: nowrap;
     text-overflow: ellipsis;
     overflow: hidden;
+    color: #00BCD4;
 }
 
 .badge {
     margin-right: 3px;
 }
 
+.badge-user {
+    color: #212529;
+    background-color: #e0e0e0;
+}
+
 .fit {
     display: block;
     min-width: 64px;
@@ -106,6 +204,15 @@ body {
     height: auto;
 }
 
+.fit-sm {
+    display: block;
+    max-width: 64px;
+    max-height: 64px;
+    margin: 0 auto 0;
+    width: auto;
+    height: auto;
+}
+
 .audio-fit {
     height: 39px;
     vertical-align: bottom;
@@ -149,6 +256,8 @@ mark {
     border: 1px solid #616161;
     border-radius: 4px;
     margin: 3px;
+    white-space: normal;
+    color: rgb(224, 224, 224);
 }
 
 .irs-single, .irs-from, .irs-to {
@@ -164,6 +273,7 @@ mark {
     margin-top: 1em;
     margin-bottom: 1em;
 }
+
 .custom-select {
     overflow: auto;
     background-color: #37474F;
@@ -229,6 +339,7 @@ option {
     padding: 0.5rem;
     background: #212121;
     color: #eee;
+    margin-top: 1em;
 }
 
 .btn-xs {
@@ -239,4 +350,76 @@ option {
 
 .btn {
     color: #eee;
-}
+}
+
+.nav-tabs .nav-link {
+    color: #e0e0e0;
+}
+
+.nav-tabs .nav-item.show .nav-link, .nav-tabs .nav-link.active {
+    background-color: #212121;
+    border-color: #616161 #616161 #212121;
+    color: #e0e0e0;
+}
+
+.nav-tabs .nav-link:focus, .nav-tabs .nav-link:focus {
+    border-color: #616161 #616161 #212121;
+    color: #e0e0e0;
+}
+
+.nav-tabs .nav-link:focus, .nav-tabs .nav-link:hover {
+    border-color: #e0e0e0 #e0e0e0 #212121;
+    color: #e0e0e0;
+}
+
+.nav-tabs {
+    border-bottom: #616161;
+}
+
+.nav {
+    margin-top: 0.5rem;
+}
+
+@media (max-width: 800px) {
+    #treeTabs {
+        flex-basis: inherit;
+        flex-grow: inherit;
+    }
+}
+
+.list-group {
+    margin-top: 1em;
+}
+
+.list-group-item {
+    padding: .25rem 0.5rem;
+}
+
+.wrapper-sm {
+    min-width: 64px;
+}
+
+.media-expanded {
+    display: inherit;
+}
+
+.media-expanded .fit {
+    max-height: 250px;
+}
+
+@media (max-width: 600px) {
+    .media-expanded .fit {
+        max-height: none;
+    }
+
+    .tagline {
+        display: none;
+    }
+}
+
+.version {
+    color: #00BCD4;
+    margin-left: -18px;
+    margin-top: -14px;
+    font-size: 11px;
+}

web/css/jquery.toast.min.css

@@ -0,0 +1 @@
+.jq-toast-wrap,.jq-toast-wrap *{margin:0;padding:0}.jq-toast-wrap{display:block;position:fixed;width:250px;pointer-events:none!important;letter-spacing:normal;z-index:9000!important}.jq-toast-wrap.bottom-left{bottom:20px;left:20px}.jq-toast-wrap.bottom-right{bottom:20px;right:40px}.jq-toast-wrap.top-left{top:20px;left:20px}.jq-toast-wrap.top-right{top:20px;right:40px}.jq-toast-single{display:block;width:100%;padding:10px;margin:0 0 5px;border-radius:4px;font-size:12px;font-family:arial,sans-serif;line-height:17px;position:relative;pointer-events:all!important;background-color:#444;color:#fff}.jq-toast-single h2{font-family:arial,sans-serif;font-size:14px;margin:0 0 7px;background:0 0;color:inherit;line-height:inherit;letter-spacing:normal}.jq-toast-single a{color:#eee;text-decoration:none;font-weight:700;border-bottom:1px solid #fff;padding-bottom:3px;font-size:12px}.jq-toast-single ul{margin:0 0 0 15px;background:0 0;padding:0}.jq-toast-single ul li{list-style-type:disc!important;line-height:17px;background:0 0;margin:0;padding:0;letter-spacing:normal}.close-jq-toast-single{position:absolute;top:3px;right:7px;font-size:14px;cursor:pointer}.jq-toast-loader{display:block;position:absolute;top:-2px;height:5px;width:0;left:0;border-radius:5px;background:red}.jq-toast-loaded{width:100%}.jq-has-icon{padding:10px 10px 10px 50px;background-repeat:no-repeat;background-position:10px}.jq-icon-info{background-image:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABgAAAAYCAYAAADgdz34AAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAADsMAAA7DAcdvqGQAAAGwSURBVEhLtZa9SgNBEMc9sUxxRcoUKSzSWIhXpFMhhYWFhaBg4yPYiWCXZxBLERsLRS3EQkEfwCKdjWJAwSKCgoKCcudv4O5YLrt7EzgXhiU3/4+b2ckmwVjJSpKkQ6wAi4gwhT+z3wRBcEz0yjSseUTrcRyfsHsXmD0AmbHOC9Ii8VImnuXBPglHpQ5wwSVM7sNnTG7Za4JwDdCjxyAiH3nyA2mtaTJufiDZ5dCaqlItILh1NHatfN5skvjx9Z38m69CgzuXmZgVrPIGE763Jx9qKsRozWYw6xOHdER+nn2KkO+Bb+UV5CBN6WC6QtBgbRVozrahAbmm6HtUsgtPC19tFdxXZYBOfkbmFJ1VaHA1VAHjd0pp70oTZzvR+EVrx2Ygfdsq6eu55BHYR8hlcki+n+kERUFG8BrA0BwjeAv2M8WLQBtcy+SD6fNsmnB3AlBLrgTtVW1c2QN4bVWLATaIS60J2Du5y1TiJgjSBvFVZgTmwCU+dAZFoPxGEEs8nyHC9Bwe2GvEJv2WXZb0vjdyFT4Cxk3e/kIqlOGoVLwwPevpYHT+00T+hWwXDf4AJAOUqWcDhbwAAAAASUVORK5CYII=);background-color:#31708f;color:#d9edf7;border-color:#bce8f1}.jq-icon-warning{background-image:url(data:image/png;base64,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);background-color:#8a6d3b;color:#fcf8e3;border-color:#faebcc}.jq-icon-error{background-image:url(data:image/png;base64,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);background-color:#a94442;color:#f2dede;border-color:#ebccd1}.jq-icon-success{background-image:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABgAAAAYCAYAAADgdz34AAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAADsMAAA7DAcdvqGQAAADsSURBVEhLY2AYBfQMgf///3P8+/evAIgvA/FsIF+BavYDDWMBGroaSMMBiE8VC7AZDrIFaMFnii3AZTjUgsUUWUDA8OdAH6iQbQEhw4HyGsPEcKBXBIC4ARhex4G4BsjmweU1soIFaGg/WtoFZRIZdEvIMhxkCCjXIVsATV6gFGACs4Rsw0EGgIIH3QJYJgHSARQZDrWAB+jawzgs+Q2UO49D7jnRSRGoEFRILcdmEMWGI0cm0JJ2QpYA1RDvcmzJEWhABhD/pqrL0S0CWuABKgnRki9lLseS7g2AlqwHWQSKH4oKLrILpRGhEQCw2LiRUIa4lwAAAABJRU5ErkJggg==);color:#dff0d8;background-color:#3c763d;border-color:#d6e9c6}

web/css/light.css

@@ -1,4 +1,46 @@
-body {overflow-y:scroll;}
+*:focus {
+    outline: 0;
+}
+
+.info-icon {
+    width: 1rem;
+    margin-right: 0.2rem;
+    cursor: pointer;
+    color: #757575;
+    line-height: 1rem;
+    height: 1rem;
+}
+
+.info-icon:hover {
+    color: inherit;
+}
+
+.modal-title {
+    max-width: calc(100% - 2rem);
+    overflow: hidden;
+    text-overflow: ellipsis;
+}
+
+.path-row {
+    display: -ms-flexbox;
+    display: flex;
+    -ms-flex-align: start;
+    align-items: flex-start;
+}
+
+.tag-container {
+    margin-left: 0.3rem;
+}
+
+.path-line {
+    color: #444;
+    text-overflow: ellipsis;
+    overflow: hidden;
+}
+
+body {
+    overflow-y: scroll;
+}
 
 .progress {
     margin-top: 1em;
@@ -6,15 +48,23 @@ body {overflow-y:scroll;}
 
 .card {
     margin-top: 1em;
-    box-shadow: 0 .125rem .25rem rgba(0,0,0,.075) !important;
+    box-shadow: 0 .125rem .25rem rgba(0, 0, 0, .075) !important;
 }
+
+.sub-document {
+    background: #AB47BC1F !important;
+}
+
 .navbar-brand {
     font-size: 1.75rem;
     padding: 0;
 }
+
 .navbar {
-    background: #F7F7F7; border-bottom: solid 1px #dfdfdf;
+    background: #F7F7F7;
+    border-bottom: solid 1px #dfdfdf;
 }
+
 .document {
     padding: 0.5rem;
 }
@@ -47,6 +97,11 @@ body {overflow-y:scroll;}
     background-color: #FFC107;
 }
 
+.badge-user {
+    color: #212529;
+    background-color: #e0e0e0;
+}
+
 .badge-text {
     color: #FFFFFF;
     background-color: #FAAB3C;
@@ -63,6 +118,9 @@ body {overflow-y:scroll;}
 }
 
 .file-title {
+    width: 100%;
+    line-height: 1rem;
+    height: 1.1rem;
     font-size: 10pt;
     white-space: nowrap;
     text-overflow: ellipsis;
@@ -84,6 +142,15 @@ body {overflow-y:scroll;}
     height: auto;
 }
 
+.fit-sm {
+    display: block;
+    max-width: 64px;
+    max-height: 64px;
+    margin: 0 auto 0;
+    width: auto;
+    height: auto;
+}
+
 .audio-fit {
     height: 39px;
     vertical-align: bottom;
@@ -98,16 +165,17 @@ body {overflow-y:scroll;}
 }
 
 @media (min-width: 1500px) {
-    .container  {
+    .container {
         max-width: 1440px;
     }
+
     .card-columns {
         column-count: 5;
     }
 }
 
 @media (min-width: 1800px) {
-    .container  {
+    .container {
         max-width: 1550px;
     }
 }
@@ -119,13 +187,15 @@ mark {
 }
 
 .content-div {
-    font-family: SFMono-Regular,Menlo,Monaco,Consolas,"Liberation Mono","Courier New",monospace;
+    font-family: SFMono-Regular, Menlo, Monaco, Consolas, "Liberation Mono", "Courier New", monospace;
     font-size: 13px;
     padding: 1em;
     background-color: #f5f5f5;
     border: 1px solid #ccc;
     border-radius: 4px;
     margin: 3px;
+    white-space: normal;
+    color: #000;
 }
 
 .irs-single, .irs-from, .irs-to {
@@ -145,8 +215,7 @@ mark {
     margin-bottom: 1em;
 }
 
-.inspire-tree .selected > .wholerow, .inspire-tree .selected > .title-wrap:hover + .wholerow
-{
+.inspire-tree .selected > .wholerow, .inspire-tree .selected > .title-wrap:hover + .wholerow {
     background: none;
 }
 
@@ -162,10 +231,59 @@ mark {
     line-height: 1rem;
     padding: 0.5rem;
     background: #f8f9fa;
+    margin-top: 1em;
 }
 
 .btn-xs {
     padding: .1rem .3rem;
     font-size: .875rem;
     border-radius: .2rem;
-}
+}
+
+.nav {
+    margin-top: 0.5rem;
+}
+
+@media (max-width: 800px) {
+    #treeTabs {
+        flex-basis: inherit;
+        flex-grow: inherit;
+    }
+}
+
+.list-group {
+    margin-top: 1em;
+}
+
+.list-group-item {
+    padding: .25rem 0.5rem;
+}
+
+.wrapper-sm {
+    min-width: 64px;
+}
+
+.media-expanded {
+    display: inherit;
+}
+
+.media-expanded .fit {
+    max-height: 250px;
+}
+
+@media (max-width: 600px) {
+    .media-expanded .fit {
+        max-height: none;
+    }
+
+    .tagline {
+        display: none;
+    }
+}
+
+.version {
+    color: #007bff;
+    margin-left: -18px;
+    margin-top: -14px;
+    font-size: 11px;
+}

web/js/dom.js

@@ -75,11 +75,122 @@ function shouldPlayVideo(hit) {
     return videoc !== "hevc" && videoc !== "mpeg2video" && videoc !== "wmv3";
 }
 
-/**
- *
- * @param hit
- * @returns {Element}
- */
+function makePlaceholder(w, h, small) {
+    let calc;
+    if (small) {
+        calc = w > h
+            ? (64 / w / h) >= 100
+                ? (64 * w / h)
+                : 64
+            : 64;
+    } else {
+        calc = w > h
+            ? (175 / w / h) >= 272
+                ? (175 * w / h)
+                : 175
+            : 175;
+    }
+
+    const el = document.createElement("div");
+    el.setAttribute("style", `height: ${calc}px`);
+    return el;
+}
+
+function makeTitle(hit) {
+    let title = document.createElement("div");
+    title.setAttribute("class", "file-title");
+    let extension = hit["_source"].hasOwnProperty("extension") && hit["_source"]["extension"] !== "" ? "." + hit["_source"]["extension"] : "";
+
+    applyNameToTitle(hit, title, extension);
+
+    title.setAttribute("title", hit["_source"]["path"] + "/" + hit["_source"]["name"] + extension);
+    return title;
+}
+
+function getTags(hit, mimeCategory) {
+
+    let tags = [];
+    switch (mimeCategory) {
+        case "video":
+        case "image":
+            if (hit["_source"].hasOwnProperty("videoc")) {
+                const formatTag = document.createElement("span");
+                formatTag.setAttribute("class", "badge badge-pill badge-video");
+                formatTag.appendChild(document.createTextNode(hit["_source"]["videoc"].replace(" ", "")));
+                tags.push(formatTag);
+            }
+            break;
+        case "audio": {
+            if (hit["_source"].hasOwnProperty("audioc")) {
+                let formatTag = document.createElement("span");
+                formatTag.setAttribute("class", "badge badge-pill badge-audio");
+                formatTag.appendChild(document.createTextNode(hit["_source"]["audioc"]));
+                tags.push(formatTag);
+            }
+        }
+            break;
+    }
+    // User tags
+    if (hit["_source"].hasOwnProperty("tag")) {
+        hit["_source"]["tag"].forEach(tag => {
+            const userTag = document.createElement("span");
+            userTag.setAttribute("class", "badge badge-pill badge-user");
+
+            const tokens = tag.split("#");
+
+            if (tokens.length > 1) {
+                const bg = "#" + tokens[1];
+                const fg = lum(tokens[1]) > 40 ? "#000" : "#fff";
+                userTag.setAttribute("style", `background-color: ${bg}; color: ${fg}`);
+            }
+
+            const name = tokens[0].split(".")[tokens[0].split(".").length - 1];
+            userTag.appendChild(document.createTextNode(name));
+            tags.push(userTag);
+        })
+    }
+
+    return tags
+}
+
+function infoButtonCb(hit) {
+    return () => {
+        getDocumentInfo(hit["_id"]).then(doc => {
+            $("#modal-title").text(doc["name"] + (doc["extension"] ? "." + doc["extension"] : ""));
+
+            const tbody = $("<tbody>");
+            $("#modal-body").empty()
+                .append($("<table class='table table-sm'>")
+                    .append($("<thead>")
+                        .append($("<tr>")
+                            .append($("<th>").text("Field"))
+                            .append($("<th>").text("Value"))
+                        )
+                    )
+                    .append(tbody)
+                );
+
+            const displayFields = new Set([
+                "mime", "size", "mtime", "path", "title", "width", "height", "duration", "audioc", "videoc",
+                "bitrate", "artist", "album", "album_artist", "genre", "title", "font_name", "tag"
+            ]);
+            Object.keys(doc)
+                .filter(key => key.startsWith("_keyword.") || key.startsWith("_text.") || displayFields.has(key))
+                .forEach(key => {
+                    tbody.append($("<tr>")
+                        .append($("<td>").text(key))
+                        .append($("<td>").text(doc[key]))
+                    );
+                });
+            if (doc.hasOwnProperty("content") && doc["content"]) {
+                $("#modal-body").append($("<div class='content-div'>").text(doc["content"]))
+            }
+
+            $("#modal").modal();
+        });
+    }
+}
+
 function createDocCard(hit) {
     let docCard = document.createElement("div");
     docCard.setAttribute("class", "card");
@@ -87,27 +198,26 @@ function createDocCard(hit) {
     let docCardBody = document.createElement("div");
     docCardBody.setAttribute("class", "card-body document");
 
+    //Title
+    let title = makeTitle(hit);
+    let isSubDocument = false;
+
     let link = document.createElement("a");
     link.setAttribute("href", "f/" + hit["_id"]);
     link.setAttribute("target", "_blank");
+    link.style.maxWidth = "calc(100% - 1.2rem)";
+    link.appendChild(title);
 
-    //Title
-    let title = document.createElement("p");
-    title.setAttribute("class", "file-title");
-    let extension = hit["_source"].hasOwnProperty("extension") && hit["_source"]["extension"] !== "" ? "." + hit["_source"]["extension"] : "";
-
-    applyNameToTitle(hit, title, extension);
-
-    title.setAttribute("title", hit["_source"]["path"] + "/" + hit["_source"]["name"] + extension);
-    docCard.appendChild(title);
+    if (hit["_source"].hasOwnProperty("parent")) {
+        docCard.classList.add("sub-document");
+        isSubDocument = true;
+    }
 
     let tagContainer = document.createElement("div");
     tagContainer.setAttribute("class", "card-text");
 
     if (hit["_source"].hasOwnProperty("mime") && hit["_source"]["mime"] !== null) {
 
-        let tags = [];
-        let thumbnail = null;
         let thumbnailOverlay = null;
         let imgWrapper = document.createElement("div");
         imgWrapper.setAttribute("style", "position: relative");
@@ -115,31 +225,7 @@ function createDocCard(hit) {
         let mimeCategory = hit["_source"]["mime"].split("/")[0];
 
         //Thumbnail
-        if (mimeCategory === "video" && shouldPlayVideo(hit)) {
-            thumbnail = document.createElement("video");
-            addVidSrc("f/" + hit["_id"], hit["_source"]["mime"], thumbnail);
-
-            thumbnail.setAttribute("class", "fit");
-            thumbnail.setAttribute("loop", "");
-            thumbnail.setAttribute("controls", "");
-            thumbnail.setAttribute("preload", "none");
-            thumbnail.setAttribute("poster", `t/${hit["_source"]["index"]}/${hit["_id"]}`);
-            thumbnail.addEventListener("dblclick", function () {
-                thumbnail.webkitRequestFullScreen();
-            });
-        } else if ((hit["_source"].hasOwnProperty("width") && hit["_source"]["width"] > 20 && hit["_source"]["height"] > 20)
-            || hit["_source"]["mime"] === "application/pdf"
-            || hit["_source"]["mime"] === "application/epub+zip"
-            || hit["_source"]["mime"] === "application/x-cbz"
-            || hit["_source"].hasOwnProperty("font_name")
-        ) {
-            thumbnail = document.createElement("img");
-            thumbnail.setAttribute("class", "card-img-top fit");
-            thumbnail.setAttribute("src", `t/${hit["_source"]["index"]}/${hit["_id"]}`);
-            thumbnail.addEventListener("error", () => {
-                imgWrapper.remove();
-            });
-        }
+        let thumbnail = makeThumbnail(mimeCategory, hit, imgWrapper, false);
 
         //Thumbnail overlay
         switch (mimeCategory) {
@@ -149,15 +235,17 @@ function createDocCard(hit) {
                 thumbnailOverlay.setAttribute("class", "card-img-overlay");
 
                 //Resolution
-                let resolutionBadge = document.createElement("span");
-                resolutionBadge.setAttribute("class", "badge badge-resolution");
-                if (hit["_source"].hasOwnProperty("width")) {
-                    resolutionBadge.appendChild(document.createTextNode(hit["_source"]["width"] + "x" + hit["_source"]["height"]));
+                if (hit["_source"].hasOwnProperty("width") && hit["_source"]["width"] > 32 && hit["_source"]["height"] > 32) {
+                    let resolutionBadge = document.createElement("span");
+                    resolutionBadge.setAttribute("class", "badge badge-resolution");
+                    if (hit["_source"].hasOwnProperty("width")) {
+                        resolutionBadge.appendChild(document.createTextNode(hit["_source"]["width"] + "x" + hit["_source"]["height"]));
+                    }
+                    thumbnailOverlay.appendChild(resolutionBadge);
                 }
-                thumbnailOverlay.appendChild(resolutionBadge);
 
                 // Hover
-                if (thumbnail && hit["_source"]["videoc"] === "gif") {
+                if (thumbnail && hit["_source"]["videoc"] === "gif" && !isSubDocument) {
                     gifOver(thumbnail, hit);
                 }
                 break;
@@ -167,51 +255,34 @@ function createDocCard(hit) {
                 if (hit["_source"].hasOwnProperty("duration")) {
                     thumbnailOverlay = document.createElement("div");
                     thumbnailOverlay.setAttribute("class", "card-img-overlay");
-                    let durationBadge = document.createElement("span");
+                    const durationBadge = document.createElement("span");
                     durationBadge.setAttribute("class", "badge badge-resolution");
                     durationBadge.appendChild(document.createTextNode(humanTime(hit["_source"]["duration"])));
                     thumbnailOverlay.appendChild(durationBadge);
                 }
         }
 
-        //Tags
-        switch (mimeCategory) {
-            case "video":
-            case "image":
-                if (hit["_source"].hasOwnProperty("videoc")) {
-                    let formatTag = document.createElement("span");
-                    formatTag.setAttribute("class", "badge badge-pill badge-video");
-                    formatTag.appendChild(document.createTextNode(hit["_source"]["videoc"].replace(" ", "")));
-                    tags.push(formatTag);
-                }
-                break;
-            case "audio": {
-                if (hit["_source"].hasOwnProperty("audioc")) {
-                    let formatTag = document.createElement("span");
-                    formatTag.setAttribute("class", "badge badge-pill badge-audio");
-                    formatTag.appendChild(document.createTextNode(hit["_source"]["audioc"]));
-                    tags.push(formatTag);
-                }
-            }
-                break;
+        // Tags
+        let tags = getTags(hit, mimeCategory);
+        for (let i = 0; i < tags.length; i++) {
+            tagContainer.appendChild(tags[i]);
         }
 
         //Content
         let contentHl = getContentHighlight(hit);
         if (contentHl !== undefined) {
-            let contentDiv = document.createElement("div");
+            const contentDiv = document.createElement("div");
             contentDiv.setAttribute("class", "content-div");
             contentDiv.insertAdjacentHTML('afterbegin', contentHl);
             docCard.appendChild(contentDiv);
         }
 
         if (thumbnail !== null) {
-            imgWrapper.appendChild(thumbnail);
             docCard.appendChild(imgWrapper);
         }
 
         //Audio
-        if (mimeCategory === "audio" && hit["_source"].hasOwnProperty("audioc")) {
+        if (mimeCategory === "audio" && hit["_source"].hasOwnProperty("audioc") && !isSubDocument) {
 
             let audio = document.createElement("audio");
             audio.setAttribute("preload", "none");
@@ -226,10 +297,6 @@ function createDocCard(hit) {
         if (thumbnailOverlay !== null) {
             imgWrapper.appendChild(thumbnailOverlay);
         }
-
-        for (let i = 0; i < tags.length; i++) {
-            tagContainer.appendChild(tags[i]);
-        }
     }
 
     //Size tag
@@ -238,15 +305,186 @@ function createDocCard(hit) {
     sizeTag.setAttribute("class", "text-muted");
     tagContainer.appendChild(sizeTag);
 
-    docCardBody.appendChild(link);
+    const titleWrapper = document.createElement("div");
+    titleWrapper.style.display = "flex";
+
+    const infoButton = makeInfoButton(hit);
+
+    titleWrapper.appendChild(infoButton);
+    titleWrapper.appendChild(link);
+
+    docCardBody.appendChild(titleWrapper);
     docCard.appendChild(docCardBody);
 
-    link.appendChild(title);
     docCardBody.appendChild(tagContainer);
 
     return docCard;
 }
 
+function makeThumbnail(mimeCategory, hit, imgWrapper, small) {
+    let thumbnail;
+    let isSubDocument = hit["_source"].hasOwnProperty("parent");
+
+    if (mimeCategory === "video" && shouldPlayVideo(hit) && !isSubDocument) {
+        thumbnail = document.createElement("video");
+        addVidSrc("f/" + hit["_id"], hit["_source"]["mime"], thumbnail);
+
+        const placeholder = makePlaceholder(hit["_source"]["width"], hit["_source"]["height"], small);
+        imgWrapper.appendChild(placeholder);
+
+        if (small) {
+            thumbnail.setAttribute("class", "fit-sm");
+        } else {
+            thumbnail.setAttribute("class", "fit");
+        }
+        if (small) {
+            thumbnail.style.cursor = "pointer";
+            thumbnail.title = "Enlarge";
+            thumbnail.addEventListener("click", function () {
+                imgWrapper.classList.remove("wrapper-sm", "mr-1");
+                imgWrapper.parentElement.classList.add("media-expanded");
+                thumbnail.setAttribute("class", "fit");
+                thumbnail.setAttribute("controls", "");
+            });
+        } else {
+            thumbnail.setAttribute("controls", "");
+        }
+        thumbnail.setAttribute("preload", "none");
+        thumbnail.setAttribute("poster", `t/${hit["_source"]["index"]}/${hit["_id"]}`);
+        thumbnail.addEventListener("dblclick", function () {
+            thumbnail.setAttribute("controls", "");
+            if (thumbnail.webkitRequestFullScreen) {
+                thumbnail.webkitRequestFullScreen();
+            } else {
+                thumbnail.requestFullscreen();
+            }
+        });
+        const poster = new Image();
+        poster.src = thumbnail.getAttribute('poster');
+        poster.addEventListener("load", function () {
+            placeholder.remove();
+            imgWrapper.appendChild(thumbnail);
+        });
+    } else if ((hit["_source"].hasOwnProperty("width") && hit["_source"]["width"] > 32 && hit["_source"]["height"] > 32)
+        || hit["_source"]["mime"] === "application/pdf"
+        || hit["_source"]["mime"] === "application/epub+zip"
+        || hit["_source"]["mime"] === "application/x-cbz"
+        || hit["_source"].hasOwnProperty("font_name")
+    ) {
+        thumbnail = document.createElement("img");
+        if (small) {
+            thumbnail.setAttribute("class", "fit-sm");
+        } else {
+            thumbnail.setAttribute("class", "card-img-top fit");
+        }
+        thumbnail.setAttribute("src", `t/${hit["_source"]["index"]}/${hit["_id"]}`);
+
+        const placeholder = makePlaceholder(hit["_source"]["width"], hit["_source"]["height"], small);
+        imgWrapper.appendChild(placeholder);
+
+        thumbnail.addEventListener("error", () => {
+            imgWrapper.remove();
+        });
+        thumbnail.addEventListener("load", () => {
+            placeholder.remove();
+            imgWrapper.appendChild(thumbnail);
+        });
+    }
+
+    return thumbnail;
+}
+
+function makeInfoButton(hit) {
+    const infoButton = document.createElement("span");
+    infoButton.appendChild(document.createTextNode("🛈"));
+    infoButton.setAttribute("class", "info-icon");
+    infoButton.addEventListener("click", infoButtonCb(hit));
+    return infoButton;
+}
+
+function createDocLine(hit) {
+
+    const mime = hit["_source"]["mime"];
+    let mimeCategory = mime ? mime.split("/")[0] : null;
+    let tags = getTags(hit, mimeCategory);
+
+    let imgWrapper = document.createElement("div");
+    imgWrapper.setAttribute("class", "align-self-start mr-1 wrapper-sm");
+
+    let media = document.createElement("div");
+    media.setAttribute("class", "media");
+
+    const line = document.createElement("div");
+    line.setAttribute("class", "list-group-item flex-column align-items-start");
+
+    if (hit["_source"].hasOwnProperty("parent")) {
+        line.classList.add("sub-document");
+        isSubDocument = true;
+    }
+
+    const infoButton = makeInfoButton(hit);
+
+    const title = makeTitle(hit);
+
+    let link = document.createElement("a");
+    link.setAttribute("href", "f/" + hit["_id"]);
+    link.setAttribute("target", "_blank");
+    link.appendChild(title);
+
+    const titleDiv = document.createElement("div");
+
+    const titleWrapper = document.createElement("div");
+    titleWrapper.style.display = "flex";
+    titleWrapper.appendChild(infoButton);
+    titleWrapper.appendChild(link);
+
+    titleDiv.appendChild(titleWrapper);
+
+    line.appendChild(media);
+
+    let thumbnail = makeThumbnail(mimeCategory, hit, imgWrapper, true);
+    if (thumbnail) {
+        media.appendChild(imgWrapper);
+    }
+    media.appendChild(titleDiv);
+
+    // Content
+    let contentHl = getContentHighlight(hit);
+    if (contentHl !== undefined) {
+        const contentDiv = document.createElement("div");
+        contentDiv.setAttribute("class", "content-div");
+        contentDiv.insertAdjacentHTML('afterbegin', contentHl);
+        titleDiv.appendChild(contentDiv);
+    }
+
+    let pathLine = document.createElement("div");
+    pathLine.setAttribute("class", "path-row");
+
+    let path = document.createElement("div");
+    path.setAttribute("class", "path-line");
+    path.setAttribute("title", hit["_source"]["path"] + "/");
+    path.appendChild(document.createTextNode(hit["_source"]["path"] + "/"));
+
+    let tagContainer = document.createElement("div");
+    tagContainer.setAttribute("class", "tag-container");
+
+    for (let i = 0; i < tags.length; i++) {
+        tagContainer.appendChild(tags[i]);
+    }
+
+    //Size tag
+    let sizeTag = document.createElement("small");
+    sizeTag.appendChild(document.createTextNode(humanFileSize(hit["_source"]["size"])));
+    sizeTag.setAttribute("class", "text-muted");
+    tagContainer.appendChild(sizeTag);
+
+    titleDiv.appendChild(pathLine);
+    pathLine.appendChild(path);
+    pathLine.appendChild(tagContainer);
+
+    return line;
+}
+
 function makePreloader() {
     const elem = document.createElement("div");
     elem.setAttribute("class", "progress");
@@ -271,18 +509,53 @@ function makePageIndicator(searchResult) {
 function makeStatsCard(searchResult) {
 
     let statsCard = document.createElement("div");
-    statsCard.setAttribute("class", "card");
+    statsCard.setAttribute("class", "card stat");
     let statsCardBody = document.createElement("div");
     statsCardBody.setAttribute("class", "card-body");
 
-    let stat = document.createElement("p");
+    const resultMode = document.createElement("div");
+    resultMode.setAttribute("class", "btn-group btn-group-toggle");
+    resultMode.setAttribute("data-toggle", "buttons");
+    resultMode.style.cssFloat = "right";
+
+    const listMode = document.createElement("label");
+    listMode.setAttribute("class", "btn btn-primary");
+    listMode.appendChild(document.createTextNode("List"));
+
+    const gridMode = document.createElement("label");
+    gridMode.setAttribute("class", "btn btn-primary");
+    gridMode.appendChild(document.createTextNode("Grid"));
+
+    resultMode.appendChild(gridMode);
+    resultMode.appendChild(listMode);
+
+    if (mode === "grid") {
+        gridMode.classList.add("active")
+    } else {
+        listMode.classList.add("active")
+    }
+
+    gridMode.addEventListener("click", () => {
+        mode = "grid";
+        localStorage.setItem("mode", mode);
+        searchDebounced();
+    });
+    listMode.addEventListener("click", () => {
+        mode = "list";
+        localStorage.setItem("mode", mode);
+        searchDebounced();
+    });
+
+    let stat = document.createElement("span");
     const totalHits = searchResult["hits"]["total"].hasOwnProperty("value")
         ? searchResult["hits"]["total"]["value"] : searchResult["hits"]["total"];
     stat.appendChild(document.createTextNode(totalHits + " results in " + searchResult["took"] + "ms"));
+
     statsCardBody.appendChild(stat);
+    statsCardBody.appendChild(resultMode);
 
     if (totalHits !== 0) {
-        let sizeStat = document.createElement("span");
+        let sizeStat = document.createElement("div");
         sizeStat.appendChild(document.createTextNode(humanFileSize(searchResult["aggregations"]["total_size"]["value"])));
         statsCardBody.appendChild(sizeStat);
     }
@@ -294,7 +567,11 @@ function makeStatsCard(searchResult) {
 
 function makeResultContainer() {
     let resultContainer = document.createElement("div");
-    resultContainer.setAttribute("class", "card-columns");
 
+    if (mode === "grid") {
+        resultContainer.setAttribute("class", "card-columns");
+    } else {
+        resultContainer.setAttribute("class", "list-group");
+    }
     return resultContainer;
 }

web/js/search.js

@@ -1,6 +1,8 @@
 const SIZE = 40;
 let mimeMap = [];
-let tree;
+let tagMap = [];
+let mimeTree;
+let tagTree;
 
 let searchBar = document.getElementById("searchBar");
 let pathBar = document.getElementById("pathBar");
@@ -10,6 +12,26 @@ let coolingDown = false;
 let searchBusy = true;
 let selectedIndices = [];
 
+let mode;
+if (localStorage.getItem("mode") === null) {
+    mode = "grid";
+} else {
+    mode = localStorage.getItem("mode")
+}
+
+function showEsError() {
+    $.toast({
+        heading: "Elasticsearch connection error",
+        text: "sist2 web module encountered an error while connecting " +
+            "to Elasticsearch. See server logs for more information.",
+        stack: false,
+        bgColor: "#a94442",
+        textColor: "#f2dede",
+        position: 'bottom-right',
+        hideAfter: false
+    });
+}
+
 jQuery["jsonPost"] = function (url, data) {
     return jQuery.ajax({
         url: url,
@@ -17,6 +39,7 @@ jQuery["jsonPost"] = function (url, data) {
         data: JSON.stringify(data),
         contentType: "application/json"
     }).fail(err => {
+        showEsError();
         console.log(err);
     });
 };
@@ -37,18 +60,51 @@ function toggleFuzzy() {
 }
 
 $.jsonPost("i").then(resp => {
+
+    const urlIndices = (new URLSearchParams(location.search)).get("i");
+
     resp["indices"].forEach(idx => {
         const opt = $("<option>")
             .attr("value", idx.id)
             .append(idx.name);
-        if (!idx.name.includes("(nsfw)")) {
-            opt.attr("selected", !idx.name.includes("(nsfw)"));
+
+        if (urlIndices) {
+            if (urlIndices.split(",").indexOf(idx.name) !== -1) {
+                opt.attr("selected", true);
+                selectedIndices.push(idx.id);
+            }
+        } else if (!idx.name.includes("(nsfw)")) {
+            opt.attr("selected", true);
             selectedIndices.push(idx.id);
         }
         $("#indices").append(opt);
     });
 });
 
+function getDocumentInfo(id) {
+    return $.getJSON("d/" + id).fail(e => {
+        console.log(e);
+        showEsError();
+    })
+}
+
+function handleTreeClick(tree) {
+    return (event, node, handler) => {
+        event.preventTreeDefault();
+
+        if (node.id === "any") {
+            if (!node.itree.state.checked) {
+                tree.deselect();
+            }
+        } else {
+            tree.node("any").deselect();
+        }
+
+        handler();
+        searchDebounced();
+    }
+}
+
 $.jsonPost("es", {
     aggs: {
         mimeTypes: {
@@ -85,34 +141,86 @@ $.jsonPost("es", {
     });
     mimeMap.push({"text": "All", "id": "any"});
 
-    tree = new InspireTree({
+    mimeTree = new InspireTree({
         selection: {
             mode: 'checkbox'
         },
         data: mimeMap
     });
-    new InspireTreeDOM(tree, {
-        target: '.tree'
+    new InspireTreeDOM(mimeTree, {
+        target: '#mimeTree'
     });
-    tree.on("node.click", function (event, node, handler) {
-        event.preventTreeDefault();
+    mimeTree.on("node.click", handleTreeClick(mimeTree));
+    mimeTree.select();
+    mimeTree.node("any").deselect();
+});
 
-        if (node.id === "any") {
-            if (!node.itree.state.checked) {
-                tree.deselect();
+function leafTag(tag) {
+    const tokens = tag.split(".");
+    return tokens[tokens.length - 1]
+}
+
+// Tags tree
+$.jsonPost("es", {
+    aggs: {
+        tags: {
+            terms: {
+                field: "tag",
+                size: 10000
             }
-        } else {
-            tree.node("any").deselect();
         }
+    },
+    size: 0,
+}).then(resp => {
+    resp["aggregations"]["tags"]["buckets"]
+        .sort((a, b) => a["key"].localeCompare(b["key"]))
+        .forEach(bucket => {
+            addTag(tagMap, bucket["key"], bucket["key"], bucket["doc_count"])
+        });
 
-        handler();
-        searchDebounced();
+    tagMap.push({"text": "All", "id": "any"});
+    tagTree = new InspireTree({
+        selection: {
+            mode: 'checkbox'
+        },
+        data: tagMap
     });
-    tree.select();
-    tree.node("any").deselect();
+    new InspireTreeDOM(tagTree, {
+        target: '#tagTree'
+    });
+    tagTree.on("node.click", handleTreeClick(tagTree));
+    tagTree.node("any").select();
     searchBusy = false;
 });
 
+function addTag(map, tag, id, count) {
+    let tags = tag.split("#")[0].split(".");
+
+    let child = {
+        id: id,
+        text: tags.length !== 1 ? tags[0] : `${tags[0]} (${count})`,
+        children: []
+    };
+
+    let found = false;
+    map.forEach(node => {
+        if (node.text === child.text) {
+            found = true;
+            if (tags.length !== 1) {
+                addTag(node.children, tags.slice(1).join("."), id, count);
+            }
+        }
+    });
+    if (!found) {
+        if (tags.length !== 1) {
+            addTag(child.children, tags.slice(1).join("."), id, count);
+            map.push(child);
+        } else {
+            map.push(child);
+        }
+    }
+}
+
 new autoComplete({
     selector: '#pathBar',
     minChars: 1,
@@ -140,7 +248,12 @@ new autoComplete({
 
 function insertHits(resultContainer, hits) {
     for (let i = 0; i < hits.length; i++) {
-        resultContainer.appendChild(createDocCard(hits[i]));
+
+        if (mode === "grid") {
+            resultContainer.appendChild(createDocCard(hits[i]));
+        } else {
+            resultContainer.appendChild(createDocLine(hits[i]));
+        }
         docCount++;
     }
 }
@@ -181,8 +294,8 @@ function doScroll() {
         })
 }
 
-function getSelectedMimeTypes() {
-    let mimeTypes = [];
+function getSelectedNodes(tree) {
+    let selectedNodes = [];
 
     let selected = tree.selected();
 
@@ -194,11 +307,11 @@ function getSelectedMimeTypes() {
 
         //Only get children
         if (selected[i].text.indexOf("(") !== -1) {
-            mimeTypes.push(selected[i].id);
+            selectedNodes.push(selected[i].id);
         }
     }
 
-    return mimeTypes
+    return selectedNodes
 }
 
 function search() {
@@ -239,11 +352,16 @@ function search() {
     if (path !== "") {
         filters.push([{term: {path: path}}])
     }
-    let mimeTypes = getSelectedMimeTypes();
+    let mimeTypes = getSelectedNodes(mimeTree);
     if (!mimeTypes.includes("any")) {
         filters.push([{terms: {"mime": mimeTypes}}]);
     }
 
+    let tags = getSelectedNodes(tagTree);
+    if (!tags.includes("any")) {
+        filters.push([{terms: {"tag": tags}}]);
+    }
+
     $.jsonPost("es?scroll=1", {
         "_source": {
             excludes: ["content"]
@@ -269,6 +387,7 @@ function search() {
             post_tags: ["</mark>"],
             fields: {
                 content: {},
+                // "content.nGram": {},
                 name: {},
                 "name.nGram": {},
                 font_name: {},

web/js/util.js

@@ -3,7 +3,7 @@
  */
 function humanFileSize(bytes) {
     if (bytes === 0) {
-        return "? B"
+        return "0 B"
     }
 
     let thresh = 1000;
@@ -43,9 +43,9 @@ function humanTime(sec_num) {
 
 function debounce(func, wait) {
     let timeout;
-    return function() {
+    return function () {
         let context = this, args = arguments;
-        let later = function() {
+        let later = function () {
             timeout = null;
             func.apply(context, args);
         };
@@ -54,3 +54,13 @@ function debounce(func, wait) {
         func.apply(context, args);
     };
 }
+
+function lum(c) {
+    c = c.substring(1);
+    let rgb = parseInt(c, 16);
+    let r = (rgb >> 16) & 0xff;
+    let g = (rgb >> 8) & 0xff;
+    let b = (rgb >> 0) & 0xff;
+
+    return 0.2126 * r + 0.7152 * g + 0.0722 * b;
+}

web/search.html

@@ -11,6 +11,7 @@
 
 <nav class="navbar navbar-expand-lg">
     <a class="navbar-brand" href="/">sist2</a>
+    <span class="badge badge-pill version">v1.2.4</span>
     <span class="tagline">Lightning-fast file system indexer and search tool </span>
     <a style="margin-left: auto" id="theme" class="btn" title="Toggle theme" href="/">Theme</a>
 </nav>
@@ -24,7 +25,7 @@
             <div class="input-group">
                 <div class="input-group-prepend">
                     <div class="input-group-text">
-                        <span onclick="document.getElementById('fuzzyToggle').click()">Fuzzy&nbsp</span>
+                        <span title="Toggle fuzzy searching"  onclick="document.getElementById('fuzzyToggle').click()">Fuzzy&nbsp</span>
                         <input title="Toggle fuzzy searching" type="checkbox" id="fuzzyToggle"
                                onclick="toggleFuzzy()" checked>
                     </div>
@@ -41,11 +42,39 @@
                     <select class="custom-select" id="indices" multiple size="6"></select>
                 </div>
 
-                <div class="col">
-                    <label>Mime types</label>
-
-                    <div class="tree"></div>
+                <div class="col" id="treeTabs">
+                    <ul class="nav nav-tabs" role="tablist">
+                        <li class="nav-item">
+                            <a class="nav-link active" data-toggle="tab" href="#mime" role="tab" aria-controls="home" aria-selected="true">Mime Types</a>
+                        </li>
+                        <li class="nav-item">
+                            <a class="nav-link" data-toggle="tab" href="#tag" role="tab" aria-controls="profile" aria-selected="false" title="User-defined tags">Tags</a>
+                        </li>
+                    </ul>
+                    <div class="tab-content" id="myTabContent">
+                        <div class="tab-pane fade show active" id="mime" role="tabpanel" aria-labelledby="home-tab">
+                            <div id="mimeTree" class="tree"></div>
+                        </div>
+                        <div class="tab-pane fade" id="tag" role="tabpanel" aria-labelledby="profile-tab">
+                            <div id="tagTree" class="tree"></div>
+                        </div>
+                    </div>
                 </div>
+
+            </div>
+        </div>
+    </div>
+
+    <div class="modal" id="modal" tabindex="-1" role="dialog" aria-labelledby="modal-title" aria-hidden="true">
+        <div class="modal-dialog modal-lg modal-dialog-centered" role="document">
+            <div class="modal-content">
+                <div class="modal-header">
+                    <h5 class="modal-title" id="modal-title"></h5>
+                    <button type="button" class="close" data-dismiss="modal" aria-label="Close">
+                        <span aria-hidden="true">&times;</span>
+                    </button>
+                </div>
+                <div class="modal-body" id="modal-body"></div>
             </div>
         </div>
     </div>