Support for external json indices

.gitignore

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

.gitmodules

@@ -31,15 +31,12 @@
 [submodule "lib/mupdf"]
 	path = lib/mupdf
 	url = git://git.ghostscript.com/mupdf.git
-[submodule "lib/zstd"]
+[submodule "lib/tesseract"]
-	path = lib/zstd
+	path = lib/tesseract
-	url = https://github.com/facebook/zstd
+	url = https://github.com/tesseract-ocr/tesseract
-[submodule "lib/lz4"]
+[submodule "lib/leptonica"]
-	path = lib/lz4
+	path = lib/leptonica
-	url = https://github.com/lz4/lz4
+	url = https://github.com/danbloomberg/leptonica
-[submodule "lib/libarchive"]
+[submodule "lib/libtiff"]
-	path = lib/libarchive
+	path = lib/libtiff
-	url = https://github.com/libarchive/libarchive
+	url = https://gitlab.com/libtiff/libtiff
-[submodule "lib/libxml2"]
-	path = lib/libxml2
-	url = https://github.com/GNOME/libxml2

CMakeLists.txt

@@ -1,14 +1,10 @@
 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(
-    add_executable(
         sist2
         src/main.c
         src/sist.h
@@ -44,44 +40,7 @@ if (WITH_SIST2)
 
         # 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
-            src/parsing/arc.h src/parsing/arc.c
-            src/parsing/doc.h src/parsing/doc.c
-            src/log.h src/log.c
-
-            # 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 ()
 
 find_package(PkgConfig REQUIRED)
 set(ENV{PKG_CONFIG_PATH} "$ENV{PKG_CONFIG_PATH}:/usr/local/lib/pkgconfig/")
@@ -102,8 +61,7 @@ list(REMOVE_ITEM GLIB_LIBRARIES pcre)
 list(REMOVE_ITEM GOBJECT_LIBRARIES pcre)
 list(REMOVE_ITEM UUID_LIBRARIES pcre)
 
-if (WITH_SIST2)
+target_include_directories(
-    target_include_directories(
         sist2 PUBLIC
         ${GOBJECT_INCLUDE_DIRS}
         ${GLIB_INCLUDE_DIRS}
@@ -116,22 +74,22 @@ if (WITH_SIST2)
         ${PROJECT_SOURCE_DIR}/lib/mupdf/include/
         ${PROJECT_SOURCE_DIR}/include/
         /usr/include/libxml2/
-    )
+        ${PROJECT_SOURCE_DIR}/lib/tesseract/include/
-    target_link_directories(
+)
+target_link_directories(
         sist2 PUBLIC
         ${UUID_LIBRARY_DIRS}
-    )
+)
 
-
+target_compile_options(sist2
-    target_compile_options(sist2
         PRIVATE
         -Ofast
-#                    -march=native
+        # -march=native
         -fno-stack-protector
         -fomit-frame-pointer
         )
 
-    TARGET_LINK_LIBRARIES(
+TARGET_LINK_LIBRARIES(
         sist2
 
         ${GLIB_LIBRARIES}
@@ -162,100 +120,21 @@ if (WITH_SIST2)
         freetype
         archive
 
-#            ${PROJECT_SOURCE_DIR}/lib/libxml2.a
         xml2
         ${PROJECT_SOURCE_DIR}/lib/libopc/libmce.a
         ${PROJECT_SOURCE_DIR}/lib/libopc/libopc.a
         ${PROJECT_SOURCE_DIR}/lib/libopc/libplib.a
-    )
 
-endif ()
+        ${PROJECT_SOURCE_DIR}/lib/libtesseract.a
-
+        ${PROJECT_SOURCE_DIR}/lib/liblept.a
-if (WITH_SIST2_SCAN)
+        ${PROJECT_SOURCE_DIR}/lib/libtiff.a
-
+        png
-    set_target_properties(
+        stdc++
-            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}
-            ${PROJECT_SOURCE_DIR}/lib/ffmpeg/
-            ${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/
-            ${PROJECT_SOURCE_DIR}/include/
-            /usr/include/libxml2/
-    )
-    target_link_directories(
-            sist2_scan PUBLIC
-            ${UUID_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
-
-            ${PROJECT_SOURCE_DIR}/lib/libarchive.a
-            ${PROJECT_SOURCE_DIR}/lib/liblz4.a
-            ${PROJECT_SOURCE_DIR}/lib/liblzma.a
-            ${PROJECT_SOURCE_DIR}/lib/libzstd.a
-
-            ${PROJECT_SOURCE_DIR}/lib/libxml2.a
-            ${PROJECT_SOURCE_DIR}/lib/libopc/libmce.a
-            ${PROJECT_SOURCE_DIR}/lib/libopc/libopc.a
-            ${PROJECT_SOURCE_DIR}/lib/libopc/libplib.a
-    )
-endif ()
 
 add_custom_target(
         before_sist2
         COMMAND ${CMAKE_CURRENT_SOURCE_DIR}/scripts/before_build.sh
 )
 
-IF (WITH_SIST2)
+add_dependencies(sist2 before_sist2)
-    add_dependencies(sist2 before_sist2)
-else ()
-    add_dependencies(sist2_scan before_sist2)
-endif ()

Docker/Dockerfile

@@ -2,7 +2,17 @@ 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
+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
 

Docker/build.sh

@@ -1,5 +1,6 @@
 rm ./sist2
 cp ../sist2 .
+strip sist2
 
 version=$(./sist2 --version)
 

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
 
@@ -11,21 +12,24 @@ 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\*
+* Extracts text from common file types \*
-* Generates thumbnails\*
+* 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 [Archive files](#archive-files)
+\*\* See [Archive files](#archive-files)    
+\*\*\* See [OCR](#ocr)    
 
 ## Getting Started
 
 1. Have an [Elasticsearch](https://www.elastic.co/downloads/elasticsearch) instance running
 1. 
     1. Download the [latest sist2 release](https://github.com/simon987/sist2/releases) *
+    1. *(or)* Download an [development snapshot](https://files.simon987.net/artifacts/Sist2/Build/) *(Not recommended!)*
     1. *(or)* `docker pull simon987/sist2:latest`
    
 
@@ -85,7 +89,7 @@ docker stop sist2
 
 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` |
@@ -111,6 +115,21 @@ scan is also supported.
 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
 
@@ -124,12 +143,7 @@ binaries.
     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
+        libarchive-dev libtiff5 libpng16-16 libpango1.0-dev
-   ```
-    *(FreeBSD)*
-    ```bash
-   pkg install cmake gcc yasm gmake bash ffmpeg e2fsprogs-uuid\
-        autotools ragel libarchive
    ```
 
 2. Build

schema/mappings.json

@@ -13,19 +13,24 @@
       "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",

scripts/get_static_libs.sh

@@ -70,43 +70,30 @@ make -j $THREADS
 cd ..
 mv libmagic/src/.libs/libmagic.a .
 
-# libarchive
+# tesseract
-cd libarchive/build
+cd tesseract
+mkdir build
+cd build
+cmake -DSTATIC=on -DBUILD_TRAINING_TOOLS=off ..
+make -j $THREADS
+cd ../..
+mv tesseract/build/libtesseract.a .
+
+# leptonica
+cd leptonica
 ./autogen.sh
-cd ..
+./configure --without-zlib --without-jpeg --without-giflib \
-./configure --without-nettle --without-expat --without-xml2 --without-openssl
+  --without-giflib --without-libwebp --without-libwebpmux --without-libopenjpeg \
+  --enable-static --disable-shared
 make -j $THREADS
 cd ..
-mv libarchive/.libs/libarchive.a .
+mv leptonica/src/.libs/liblept.a .
 
-# lz4
+# tiff
-cd lz4
+cd libtiff
-make -j $THREADS
-cd ..
-mv lz4/lib/liblz4.a .
-
-# lzma
-wget https://newcontinuum.dl.sourceforge.net/project/lzmautils/xz-5.2.3.tar.gz
-tar -xzf xz-5.2.3.tar.gz
-rm xz-5.2.3.tar.gz
-cd xz-5.2.3
 ./autogen.sh
-./configure
+./configure --enable-static --disable-shared --disable-lzw --disable-jpeg --disable-webp \
+  --disable-lzma --disable-zstd --disable-jbig
 make -j $THREADS
 cd ..
-mv xz-5.2.3/src/liblzma/.libs/liblzma.a .
+mv libtiff/libtiff/.libs/libtiff.a .
-
-# zstd
-cd zstd
-make -j $THREADS
-cd ..
-mv zstd/lib/libzstd.a .
-
-# xml2
-cd libxml2
-./autogen.sh --without-zlib --without-lzma
-make
-cd ..
-mv libxml2/.libs/libxml2.a .
-
-cd ..

scripts/get_static_libs_freebsd.sh

@@ -1,58 +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
-cd bzip2-1.0.6
-make -j 4
-cd ..
-mv bzip2-1.0.6/libbz2.a .
-
-# magic
-cd libmagic
-./autogen.sh
-./configure --enable-static --disable-shared
-make -j 4
-cd ..
-mv libmagic/src/.libs/libmagic.a .
-
-cd ..

src/cli.c

@@ -1,6 +1,8 @@
 #include "cli.h"
 #include "ctx.h"
 
+#include <tesseract/capi.h>
+
 #define DEFAULT_OUTPUT "index.sist2/"
 #define DEFAULT_CONTENT_SIZE 32768
 #define DEFAULT_QUALITY 5
@@ -35,8 +37,6 @@ void scan_args_destroy(scan_args_t *args) {
     free(args);
 }
 
-#ifndef SIST_SCAN_ONLY
-
 void index_args_destroy(index_args_t *args) {
     //todo
     free(args);
@@ -47,8 +47,6 @@ void web_args_destroy(web_args_t *args) {
     free(args);
 }
 
-#endif
-
 int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
     if (argc < 2) {
         fprintf(stderr, "Required positional argument: PATH.\n");
@@ -136,6 +134,17 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
         return 1;
     }
 
+    if (args->tesseract_lang != NULL) {
+        TessBaseAPI *api = TessBaseAPICreate();
+        ret = TessBaseAPIInit3(api, TESS_DATAPATH, 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);
+    }
+
     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)
@@ -147,14 +156,15 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
     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 ocr=%s", args->tesseract_lang)
 
     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;
@@ -216,6 +226,8 @@ 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) {
 
+    LogCtx.verbose = 1;
+
     if (argc < 2) {
         fprintf(stderr, "Required positional argument: PATH.\n");
         return 1;
@@ -273,5 +285,3 @@ web_args_t *web_args_create() {
     return args;
 }
 
-#endif
-

src/cli.h

@@ -16,13 +16,13 @@ typedef struct scan_args {
     char *path;
     char *archive;
     archive_mode_t archive_mode;
+    char *tesseract_lang;
 } 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;
@@ -51,6 +51,5 @@ 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

@@ -27,6 +27,7 @@ struct {
     GHashTable *copy_table;
 
     pthread_mutex_t mupdf_mu;
+    char * tesseract_lang;
 } ScanCtx;
 
 struct {
@@ -35,7 +36,6 @@ struct {
     int no_color;
 } LogCtx;
 
-#ifndef SIST_SCAN_ONLY
 struct {
     char *es_url;
     int batch_size;
@@ -47,7 +47,6 @@ struct {
     char *b64credentials;
     struct index_t indices[16];
 } WebCtx;
-#endif
 
 
 #endif

src/index/elastic.c

@@ -27,13 +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]) {
@@ -68,7 +69,7 @@ void execute_update_script(const char *script, const char index_id[UUID_STR_LEN]
     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");
-    printf("Executed user script <%d>\n", r->status_code);
+    LOG_INFOF("elastic.c", "Executed user script <%d>", r->status_code);
     cJSON *resp = cJSON_Parse(r->body);
 
     cJSON_free(str);
@@ -79,7 +80,7 @@ void execute_update_script(const char *script, const char index_id[UUID_STR_LEN]
     if (error != NULL) {
         char *error_str = cJSON_Print(error);
 
-        fprintf(stderr, "User script error: \n%s\n", error_str);
+        LOG_ERRORF("elastic.c", "User script error: \n%s", error_str);
         cJSON_free(error_str);
     }
 
@@ -132,11 +133,10 @@ void elastic_flush() {
     response_t *r = web_post(bulk_url, buf, "Content-Type: application/x-ndjson");
 
     if (r->status_code == 0) {
-        fprintf(stderr, "Could not connect to %s, make sure that elasticsearch is running!\n", IndexCtx.es_url);
+        LOG_FATALF("elastic.c", "Could not connect to %s, make sure that elasticsearch is running!\n", IndexCtx.es_url)
-        exit(1);
     }
 
-    printf("Indexed %3d documents (%zukB) <%d>\n", count, buf_cur / 1024, r->status_code);
+    LOG_INFOF("elastic.c", "Indexed %3d 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) {
@@ -144,7 +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);
             }
         }
@@ -198,7 +198,7 @@ void destroy_indexer(char * script, char index_id[UUID_STR_LEN]) {
 
     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) {
@@ -207,12 +207,12 @@ void destroy_indexer(char * script, char index_id[UUID_STR_LEN]) {
 
     snprintf(url, sizeof(url), "%s/sist2/_refresh", IndexCtx.es_url);
     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);
 
     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) {
@@ -232,32 +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);
     }
 }

src/io/serialize.c

@@ -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);
 
@@ -56,8 +57,7 @@ index_descriptor_t read_index_descriptor(char *path) {
     int fd = open(path, O_RDONLY);
 
     if (fd == -1) {
-        fprintf(stderr, "Invalid/corrupt index (Could not find descriptor)\n");
+        LOG_FATAL("serialize.c", "Invalid/corrupt index (Could not find descriptor)\n")
-        exit(1);
     }
 
     char *buf = malloc(info.st_size + 1);
@@ -75,6 +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);
@@ -172,8 +177,8 @@ 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();
 
@@ -229,7 +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:
@@ -275,6 +280,89 @@ void read_index(const char *path, const char index_id[UUID_STR_LEN], index_func
     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

@@ -11,7 +11,7 @@ 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);
 

src/log.c

@@ -56,7 +56,7 @@ void sist_logf(char *filepath, int level, char *format, ...) {
         log_len += 1;
     }
 
-    write(STDOUT_FILENO, log_str, log_len);
+    write(STDERR_FILENO, log_str, log_len);
 }
 
 void sist_log(char *filepath, int level, char *str) {

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 = "1.1.14";
+static const char *const Version = "1.2.1";
 static const char *const usage[] = {
         "sist2 scan [OPTION]... PATH",
         "sist2 index [OPTION]... INDEX",
@@ -19,9 +15,7 @@ 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();
 }
@@ -35,6 +29,7 @@ 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);
 }
@@ -55,6 +50,7 @@ void sist2_scan(scan_args_t *args) {
     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;
 
     init_dir(ScanCtx.index.path);
 
@@ -122,8 +118,6 @@ 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;
@@ -137,8 +131,12 @@ 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;
     }
 
@@ -160,7 +158,7 @@ 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);
@@ -198,18 +196,14 @@ void sist2_web(web_args_t *args) {
     serve(args->bind, args->port);
 }
 
-#endif
-
 
 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
 
     int arg_version = 0;
 
@@ -240,8 +234,9 @@ int main(int argc, const char *argv[]) {
             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 with port. DEFAULT=http://localhost:9200"),
             OPT_BOOLEAN('p', "print", &index_args->print, "Just print JSON documents to stdout."),
@@ -255,7 +250,6 @@ int main(int argc, const char *argv[]) {
             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"),
             OPT_STRING(0, "auth", &web_args->credentials, "Basic auth in user:password format"),
-#endif
 
             OPT_END(),
     };
@@ -274,10 +268,8 @@ int main(int argc, const char *argv[]) {
         LogCtx.verbose = 1;
     }
 
-#ifndef SIST_SCAN_ONLY
     web_args->es_url = common_es_url;
     index_args->es_url = common_es_url;
-#endif
 
     if (argc == 0) {
         argparse_usage(&argparse);
@@ -292,7 +284,6 @@ 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);
@@ -310,7 +301,6 @@ int main(int argc, const char *argv[]) {
         sist2_web(web_args);
 
     }
-#endif
     else {
         fprintf(stderr, "Invalid command: '%s'\n", argv[0]);
         argparse_usage(&argparse);
@@ -320,10 +310,8 @@ int main(int argc, const char *argv[]) {
 
     scan_args_destroy(scan_args);
 
-#ifndef SIST_SCAN_ONLY
     index_args_destroy(index_args);
     web_args_destroy(web_args);
-#endif
 
     return 0;
 }

src/parsing/media.c

@@ -127,6 +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); \

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;
@@ -86,14 +90,13 @@ fz_page *render_cover(fz_context *ctx, document_t *doc, fz_document *fzdoc) {
     return cover;
 }
 
-void fz_err_callback(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;
+        document_t *doc = (document_t *) user;
         LOG_WARNINGF(doc->filepath, "FZ: %s", message)
     }
 }
 
-
 __always_inline
 void init_ctx(fz_context *ctx, document_t *doc) {
     fz_disable_icc(ctx);
@@ -124,6 +127,40 @@ 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, TESS_DATAPATH, 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) {
@@ -210,7 +247,7 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
 
     if (ScanCtx.content_size > 0) {
         fz_stext_options opts = {0};
-        text_buffer_t text_buf = text_buffer_create(ScanCtx.content_size);
+        thread_buffer = text_buffer_create(ScanCtx.content_size);
 
         for (int current_page = 0; current_page < page_count; current_page++) {
             fz_page *page = NULL;
@@ -224,7 +261,7 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
                     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(&text_buf);
+                    text_buffer_destroy(&thread_buffer);
                     fz_drop_page(ctx, page);
                     fz_drop_stream(ctx, stream);
                     fz_drop_document(ctx, fzdoc);
@@ -235,6 +272,15 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
 
             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;
+            }
 
             fz_var(err);
             fz_try(ctx)
@@ -249,7 +295,7 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
 
             if (err != 0) {
                 LOG_WARNINGF(doc->filepath, "fz_run_page() returned error code [%d] %s", err, ctx->error.message)
-                text_buffer_destroy(&text_buf);
+                text_buffer_destroy(&thread_buffer);
                 fz_drop_page(ctx, page);
                 fz_drop_stext_page(ctx, stext);
                 fz_drop_stream(ctx, stream);
@@ -260,7 +306,7 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
 
             fz_stext_block *block = stext->first_block;
             while (block != NULL) {
-                int ret = read_stext_block(block, &text_buf);
+                int ret = read_stext_block(block, &thread_buffer);
                 if (ret == TEXT_BUF_FULL) {
                     break;
                 }
@@ -269,18 +315,18 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
             fz_drop_stext_page(ctx, stext);
             fz_drop_page(ctx, page);
 
-            if (text_buf.dyn_buffer.cur >= text_buf.dyn_buffer.size) {
+            if (thread_buffer.dyn_buffer.cur >= thread_buffer.dyn_buffer.size) {
                 break;
             }
         }
-        text_buffer_terminate_string(&text_buf);
+        text_buffer_terminate_string(&thread_buffer);
 
-        meta_line_t *meta_content = malloc(sizeof(meta_line_t) + text_buf.dyn_buffer.cur);
+        meta_line_t *meta_content = malloc(sizeof(meta_line_t) + thread_buffer.dyn_buffer.cur);
         meta_content->key = MetaContent;
-        memcpy(meta_content->strval, text_buf.dyn_buffer.buf, text_buf.dyn_buffer.cur);
+        memcpy(meta_content->strval, thread_buffer.dyn_buffer.buf, thread_buffer.dyn_buffer.cur);
         APPEND_META(doc, meta_content)
 
-        text_buffer_destroy(&text_buf);
+        text_buffer_destroy(&thread_buffer);
     }
 
     fz_drop_stream(ctx, stream);

src/parsing/pdf.h

@@ -2,6 +2,7 @@
 #define SIST2_PDF_H
 
 #include "src/sist.h"
+#include <tesseract/capi.h>
 
 
 void parse_pdf(void *buf, size_t buf_len, document_t *doc);

src/parsing/text.c

@@ -25,6 +25,7 @@ void parse_text(int bytes_read, struct vfile *f, char *buf, document_t *doc) {
     }
     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/sist.h

@@ -2,6 +2,8 @@
 #define SIST_H
 
 #define UUID_STR_LEN 37
+#define UNUSED(x) __attribute__((__unused__))  x
+#define TESS_DATAPATH "/usr/share/tessdata/"
 
 #include <glib-2.0/glib.h>
 #include <unistd.h>
@@ -33,14 +35,12 @@
 #include <opc/opc.h>
 #include <libxml/xmlstring.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"
 
@@ -62,12 +62,10 @@
 #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"
 #include "web/auth_basic.h"
-#endif
 
 ;
 

src/types.h

@@ -33,6 +33,10 @@ enum metakey {
     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];
@@ -41,6 +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 {
@@ -70,8 +75,6 @@ typedef struct document {
     short ext;
     meta_line_t *meta_head;
     meta_line_t *meta_tail;
-    struct document *child_head;
-    struct document *child_tail;
     char *filepath;
 } document_t;
 
@@ -79,8 +82,6 @@ typedef struct vfile vfile_t;
 
 typedef int (*read_func_t)(struct vfile *, void *buf, size_t size);
 
-typedef int (*seek_func_t)(struct vfile *, size_t size, int whence);
-
 typedef void (*close_func_t)(struct vfile *);
 
 typedef struct vfile {

src/util.c

@@ -149,7 +149,6 @@ 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;
     }
 
@@ -158,7 +157,6 @@ 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;
 }
 
@@ -169,7 +167,6 @@ 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) {

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 {

web/search.html

@@ -11,7 +11,7 @@
 
 <nav class="navbar navbar-expand-lg">
     <a class="navbar-brand" href="/">sist2</a>
-    <span class="badge badge-pill version">v1.1.14</span>
+    <span class="badge badge-pill version">v1.2.1</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>