Fix bug with media files, don't encode thumbnail when not necessary

Update USAGE.md

CMakeLists.txt

@@ -40,6 +40,7 @@ find_package(cJSON CONFIG REQUIRED)
 find_package(unofficial-glib CONFIG REQUIRED)
 find_package(unofficial-mongoose CONFIG REQUIRED)
 find_library(UUID_LIB NAMES uuid)
+find_package(CURL CONFIG REQUIRED)
 
 #find_package(OpenSSL REQUIRED)
 
@@ -56,7 +57,6 @@ target_compile_options(
         sist2
         PRIVATE
         -fPIC
-        -Werror
 )
 
 if (SIST_DEBUG)
@@ -67,6 +67,7 @@ if (SIST_DEBUG)
             -fstack-protector
             -fno-omit-frame-pointer
             -fsanitize=address
+            -O2
     )
     target_link_options(
             sist2
@@ -80,6 +81,7 @@ if (SIST_DEBUG)
             OUTPUT_NAME sist2_debug
     )
 else ()
+#    set(VCPKG_BUILD_TYPE release)
     target_compile_options(
             sist2
             PRIVATE
@@ -105,6 +107,7 @@ target_link_libraries(
         unofficial::glib::glib
         unofficial::mongoose::mongoose
 #        OpenSSL::SSL OpenSSL::Crypto
+        CURL::libcurl
 
         ${UUID_LIB}
         pthread

README.md

@@ -15,10 +15,10 @@ sist2 (Simple incremental search tool)
 * Fast, low memory usage, multi-threaded
 * Mobile-friendly Web interface
 * Portable (all its features are packaged in a single executable)
-* Extracts text from common file types \*
+* Extracts text and metadata from common file types \*
 * Generates thumbnails \*
 * Incremental scanning
-* Automatic tagging from file attributes via [user scripts](docs/scripting.md)
+* Manual tagging from the UI and automatic tagging based on file attributes via [user scripts](docs/scripting.md)
 * Recursive scan inside archive files \*\*
 * OCR support with tesseract \*\*\*
 * Stats page & disk utilisation visualization
@@ -72,15 +72,17 @@ See [Usage guide](docs/USAGE.md) for more details
 
 File type | Library | Content | Thumbnail | Metadata
 :---|:---|:---|:---|:---
-pdf,xps,cbz,cbr,fb2,epub | MuPDF | text+ocr | yes, `png` | title |
+pdf,xps,fb2,epub | MuPDF | text+ocr | yes | title |
-`audio/*` | ffmpeg | - | yes, `jpeg` | ID3 tags |
+cbz,cbr | *(none)* | - | yes | - |
-`video/*` | ffmpeg | - | yes, `jpeg` | title, comment, artist |
+`audio/*` | ffmpeg | - | yes | ID3 tags |
-`image/*` | ffmpeg | - | yes, `jpeg` | [Common EXIF tags](https://github.com/simon987/sist2/blob/efdde2734eca9b14a54f84568863b7ffd59bdba3/src/parsing/media.c#L190) |
+`video/*` | ffmpeg | - | yes | title, comment, artist |
+`image/*` | ffmpeg | - | yes | [Common EXIF tags](https://github.com/simon987/sist2/blob/efdde2734eca9b14a54f84568863b7ffd59bdba3/src/parsing/media.c#L190) |
+raw, rw2, dng, cr2, crw, dcr, k25, kdc, mrw, pef, xf3, arw, sr2, srf, erf  | LibRaw | - | yes | Common EXIF tags |
 ttf,ttc,cff,woff,fnt,otf | Freetype2 | - | yes, `bmp` | Name & style |
 `text/plain` | *(none)* | yes | no | - |
 html, xml | *(none)* | yes | no | - |
 tar, zip, rar, 7z, ar ...  | Libarchive | yes\* | - | no |
-docx, xlsx, pptx | *(none)* | yes | no | creator, modified_by, title |
+docx, xlsx, pptx | *(none)* | yes | if embedded | creator, modified_by, title |
 mobi, azw, azw3 | libmobi | yes | no | author, title |
 
 \* *See [Archive files](#archive-files)*
@@ -100,7 +102,7 @@ scan is also supported.
  
 ### OCR
 
-You can enable OCR support for pdf,xps,cbz,cbr,fb2,epub file types with the
+You can enable OCR support for pdf,xps,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).
 
@@ -122,7 +124,7 @@ binaries (GCC 7+ required).
 1. Install compile-time dependencies
 
    ```bash
-   vcpkg install lmdb cjson glib libarchive[core,bzip2,libxml2,lz4,lzma,lzo] pthread tesseract libxml2 ffmpeg zstd gtest mongoose libuuid libmagic
+   vcpkg install lmdb cjson glib libarchive[core,bzip2,libxml2,lz4,lzma,lzo] pthread tesseract libxml2 ffmpeg zstd gtest mongoose libuuid libmagic libraw
    ```
 
 2. Build

ci/build.sh

@@ -1,16 +1,17 @@
 #!/usr/bin/env bash
 
+VCPKG_ROOT="/vcpkg"
 
 rm *.gz
 
 rm -rf CMakeFiles CMakeCache.txt
-cmake -DSIST_DEBUG=off -DCMAKE_TOOLCHAIN_FILE=/vcpkg/scripts/buildsystems/vcpkg.cmake .
+cmake -DSIST_DEBUG=off -DVCPKG_BUILD_TYPE=release -DCMAKE_TOOLCHAIN_FILE="${VCPKG_ROOT}/scripts/buildsystems/vcpkg.cmake" .
-make
+make -j 12
 strip sist2
 gzip -9 sist2
 
 rm -rf CMakeFiles CMakeCache.txt
-cmake -DSIST_DEBUG=on -DCMAKE_TOOLCHAIN_FILE=/vcpkg/scripts/buildsystems/vcpkg.cmake .
+cmake -DSIST_DEBUG=on -DVCPKG_BUILD_TYPE=debug -DCMAKE_TOOLCHAIN_FILE="${VCPKG_ROOT}/scripts/buildsystems/vcpkg.cmake" .
-make
+make -j 12
 cp /usr/lib/x86_64-linux-gnu/libasan.so.2.0.0 libasan.so.2
 tar -czf sist2_debug.tar.gz sist2_debug libasan.so.2

docs/USAGE.md

@@ -15,6 +15,7 @@
     * [rewrite_url](#rewrite_url)
     * [link to specific indices](#link-to-specific-indices)
 * [exec-script](#exec-script)
+* [tagging](#tagging)
 
 ```
 Usage: sist2 scan [OPTION]... PATH
@@ -46,6 +47,7 @@ Scan options
     --mem-buffer=<int>            Maximum memory buffer size per thread in MB for files inside archives (see USAGE.md). DEFAULT: 2000
 
 Index options
+    -t, --threads=<int>           Number of threads. DEFAULT=1
     --es-url=<str>                Elasticsearch url with port. DEFAULT=http://localhost:9200
     -p, --print                   Just print JSON documents to stdout.
     --script-file=<str>           Path to user script.
@@ -56,6 +58,7 @@ Web options
     --es-url=<str>                Elasticsearch url. DEFAULT=http://localhost:9200
     --bind=<str>                  Listen on this address. DEFAULT=localhost:4090
     --auth=<str>                  Basic auth in user:password format
+    --tag-auth=<str>              Basic auth in user:password format for tagging
 
 Exec-script options
     --script-file=<str>           Path to user script.
@@ -67,7 +70,7 @@ Made by simon987 <me@simon987.net>. Released under GPL-3.0
 ### Scan options
 
 * `-t, --threads` 
-      Number of threads for file parsing. **Do not set a number higher than `$(nproc)`!**.
+      Number of threads for file parsing. **Do not set a number higher than `$(nproc)` or `$(Get-WmiObject Win32_ComputerSystem).NumberOfLogicalProcessors` in Windows!**
 * `-q, --quality` 
     Thumbnail quality, on a scale of 1.0 to 31.0, 1.0 being the best. *Does not affect PDF thumbnails quality*
 * `--size` 
@@ -144,7 +147,10 @@ documents.idx/
 ├── agg_mime.csv
 ├── agg_date.csv
 ├── add_size.csv
-└── thumbs
+├── thumbs
+|   ├── data.mdb
+|   └── lock.mdb
+└── tags
     ├── data.mdb
     └── lock.mdb
 ```
@@ -269,6 +275,8 @@ sist2 index --print ./my_index/ | jq | less
  * `--es-url=<str>` Elasticsearch url.
  * `--bind=<str>` Listen on this address.
  * `--auth=<str>` Basic auth in user:password format
+ * `--tag-auth=<str>` Basic auth in user:password format. Works the same way as the 
+    `--auth` argument, but authentication is only applied the `/tag/` endpoint.
  
 ### Web examples
 
@@ -300,3 +308,31 @@ not displayed.
 ## exec-script
 
 The `exec-script` command is used to execute a user script for an index that has already been imported to Elasticsearch with the `index` command. Note that the documents will not be reset to their default state before each execution as the `index` command does: if you make undesired changes to the documents by accident, you will need to run `index` again to revert to the original state.
+
+
+# Tagging
+
+### Manual tagging
+
+You can modify tags of individual documents directly from the 
+ `web` interface. Note that you can setup authentication for this feature
+ with the `--tag-auth` option (See [web options](#web-options))
+
+![manual_tag](manual_tag.png)
+
+Tags that are manually added are saved both in the 
+ index folder (in `/tags/`) and in Elasticsearch*. When re-`index`ing, 
+ they are read from the index and automatically applied.
+ 
+You can safely copy the `/tags/` database to another index.
+
+See [Automatic tagging](#automatic-tagging) for information about tag 
+ hierarchies and tag colors.
+
+\* *It can take a few seconds to take effect in new search queries, and the page needs 
+    to be reloaded for the tag tab to update*
+
+
+### Automatic tagging
+
+See [scripting](docs/scripting.md) documentation.

schema/mappings.json

@@ -126,7 +126,12 @@
       }
     },
     "tag": {
-      "type": "keyword"
+      "type": "keyword",
+      "copy_to": "suggest-tag"
+    },
+    "suggest-tag": {
+      "type": "completion",
+      "analyzer": "case_insensitive_kw_analyzer"
     },
     "exif_make": {
       "type": "text"

schema/settings.json

@@ -1,7 +1,8 @@
 {
   "index": {
     "refresh_interval": "30s",
-    "codec": "best_compression"
+    "codec": "best_compression",
+    "number_of_replicas": 0
   },
   "analysis": {
     "tokenizer": {

scripts/mime.py

@@ -18,7 +18,6 @@ major_mime = {
 
 pdf = (
     "application/pdf",
-    "application/x-cbz",
     "application/epub+zip",
     "application/vnd.ms-xpsdocument",
 )

src/cli.c

@@ -41,6 +41,9 @@ void scan_args_destroy(scan_args_t *args) {
     if (args->name != NULL) {
         free(args->name);
     }
+    if (args->incremental != NULL) {
+        free(args->incremental);
+    }
     if (args->path != NULL) {
         free(args->path);
     }
@@ -79,7 +82,7 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
     }
 
     if (args->incremental != NULL) {
-        abs_path = abspath(args->incremental);
+        args->incremental = abspath(args->incremental);
         if (abs_path == NULL) {
             sist_log("main.c", SIST_WARNING, "Could not open original index! Disabled incremental scan feature.");
             args->incremental = NULL;
@@ -260,6 +263,13 @@ int index_args_validate(index_args_t *args, int argc, const char **argv) {
         return 1;
     }
 
+    if (args->threads == 0) {
+        args->threads = 1;
+    } else if (args->threads < 0) {
+        fprintf(stderr, "Invalid threads: %d\n", args->threads);
+        return 1;
+    }
+
     char *index_path = abspath(argv[1]);
     if (index_path == NULL) {
         fprintf(stderr, "File not found: %s\n", argv[1]);
@@ -319,7 +329,7 @@ int web_args_validate(web_args_t *args, int argc, const char **argv) {
         }
 
         strncpy(args->auth_user, args->credentials, (ptr - args->credentials));
-        strncpy(args->auth_pass, ptr + 1, strlen(ptr + 1));
+        strcpy(args->auth_pass, ptr + 1);
 
         if (strlen(args->auth_user) == 0) {
             fprintf(stderr, "--auth username must be at least one character long");
@@ -331,6 +341,31 @@ int web_args_validate(web_args_t *args, int argc, const char **argv) {
         args->auth_enabled = FALSE;
     }
 
+    if (args->tag_credentials != NULL && args->credentials != NULL) {
+        fprintf(stderr, "--auth and --tag-auth are mutually exclusive");
+        return 1;
+    }
+
+    if (args->tag_credentials != NULL) {
+        char *ptr = strstr(args->tag_credentials, ":");
+        if (ptr == NULL) {
+            fprintf(stderr, "Invalid --tag-auth format, see usage\n");
+            return 1;
+        }
+
+        strncpy(args->auth_user, args->tag_credentials, (ptr - args->tag_credentials));
+        strcpy(args->auth_pass, ptr + 1);
+
+        if (strlen(args->auth_user) == 0) {
+            fprintf(stderr, "--tag-auth username must be at least one character long");
+            return 1;
+        }
+
+        args->tag_auth_enabled = TRUE;
+    } else {
+        args->tag_auth_enabled = FALSE;
+    }
+
     args->index_count = argc - 1;
     args->indices = argv + 1;
 
@@ -345,6 +380,7 @@ int web_args_validate(web_args_t *args, int argc, const char **argv) {
     LOG_DEBUGF("cli.c", "arg es_url=%s", args->es_url)
     LOG_DEBUGF("cli.c", "arg listen=%s", args->listen_address)
     LOG_DEBUGF("cli.c", "arg credentials=%s", args->credentials)
+    LOG_DEBUGF("cli.c", "arg tag_credentials=%s", args->tag_credentials)
     LOG_DEBUGF("cli.c", "arg auth_user=%s", args->auth_user)
     LOG_DEBUGF("cli.c", "arg auth_pass=%s", args->auth_pass)
     LOG_DEBUGF("cli.c", "arg index_count=%d", args->index_count)
@@ -367,6 +403,11 @@ web_args_t *web_args_create() {
 
 int exec_args_validate(exec_args_t *args, int argc, const char **argv) {
 
+    if (argc < 2) {
+        fprintf(stderr, "Required positional argument: PATH.\n");
+        return 1;
+    }
+
     char *index_path = abspath(argv[1]);
     if (index_path == NULL) {
         fprintf(stderr, "File not found: %s\n", argv[1]);

src/cli.h

@@ -41,15 +41,18 @@ typedef struct index_args {
     int print;
     int batch_size;
     int force_reset;
+    int threads;
 } index_args_t;
 
 typedef struct web_args {
     char *es_url;
     char *listen_address;
     char *credentials;
+    char *tag_credentials;
     char auth_user[256];
     char auth_pass[256];
     int auth_enabled;
+    int tag_auth_enabled;
     int index_count;
     const char **indices;
 } web_args_t;

src/ctx.h

@@ -5,7 +5,7 @@
 #include "tpool.h"
 #include "libscan/scan.h"
 #include "libscan/arc/arc.h"
-#include "libscan/cbr/cbr.h"
+#include "libscan/comic/comic.h"
 #include "libscan/ebook/ebook.h"
 #include "libscan/font/font.h"
 #include "libscan/media/media.h"
@@ -13,6 +13,7 @@
 #include "libscan/text/text.h"
 #include "libscan/mobi/scan_mobi.h"
 #include "libscan/raw/raw.h"
+#include "src/io/store.h"
 
 #include <glib.h>
 #include <pcre.h>
@@ -39,7 +40,7 @@ typedef struct {
     int fast;
 
     scan_arc_ctx_t arc_ctx;
-    scan_cbr_ctx_t cbr_ctx;
+    scan_comic_ctx_t comic_ctx;
     scan_ebook_ctx_t ebook_ctx;
     scan_font_ctx_t font_ctx;
     scan_media_ctx_t media_ctx;
@@ -58,6 +59,9 @@ typedef struct {
 typedef struct {
     char *es_url;
     int batch_size;
+    tpool_t *pool;
+    store_t *tag_store;
+    GHashTable *tags;
 } IndexCtx_t;
 
 typedef struct {
@@ -66,6 +70,7 @@ typedef struct {
     char *auth_user;
     char *auth_pass;
     int auth_enabled;
+    int tag_auth_enabled;
     struct index_t indices[64];
 } WebCtx_t;
 

src/index/elastic.c

@@ -14,9 +14,18 @@ typedef struct es_indexer {
 } es_indexer_t;
 
 
-static es_indexer_t *Indexer;
+static __thread es_indexer_t *Indexer;
 
 void delete_queue(int max);
+void elastic_flush();
+
+void elastic_cleanup() {
+    elastic_flush();
+    if (Indexer != NULL) {
+        free(Indexer->es_url);
+        free(Indexer);
+    }
+}
 
 void print_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
 
@@ -35,8 +44,12 @@ void print_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
     cJSON_Delete(line);
 }
 
-void index_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
+void index_json_func(void *arg) {
+    es_bulk_line_t *line = arg;
+    elastic_index_line(line);
+}
 
+void index_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
     char *json = cJSON_PrintUnformatted(document);
 
     size_t json_len = strlen(json);
@@ -48,7 +61,7 @@ void index_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
     bulk_line->next = NULL;
 
     cJSON_free(json);
-    elastic_index_line(bulk_line);
+    tpool_add_work(IndexCtx.pool, index_json_func, bulk_line);
 }
 
 void execute_update_script(const char *script, const char index_id[UUID_STR_LEN]) {
@@ -69,7 +82,7 @@ void execute_update_script(const char *script, const char index_id[UUID_STR_LEN]
     char *str = cJSON_Print(body);
 
     char bulk_url[4096];
-    snprintf(bulk_url, 4096, "%s/sist2/_update_by_query?pretty", Indexer->es_url);
+    snprintf(bulk_url, 4096, "%s/sist2/_update_by_query?wait_for_completion=false", Indexer->es_url);
     response_t *r = web_post(bulk_url, str);
     LOG_INFOF("elastic.c", "Executed user script <%d>", r->status_code);
     cJSON *resp = cJSON_Parse(r->body);
@@ -89,33 +102,44 @@ void execute_update_script(const char *script, const char index_id[UUID_STR_LEN]
     cJSON_Delete(resp);
 }
 
+#define ACTION_STR_LEN 91
+
 void *create_bulk_buffer(int max, int *count, size_t *buf_len) {
     es_bulk_line_t *line = Indexer->line_head;
     *count = 0;
 
     size_t buf_size = 0;
     size_t buf_cur = 0;
-    char *buf = malloc(1);
+    char *buf = malloc(8192);
+    size_t buf_capacity = 8192;
 
     while (line != NULL && *count < max) {
-        char action_str[512];
+        char action_str[256];
-        snprintf(action_str, 512,
+        snprintf(action_str, 256,
                  "{\"index\":{\"_id\":\"%s\", \"_type\":\"_doc\", \"_index\":\"sist2\"}}\n", line->uuid_str);
-        size_t action_str_len = strlen(action_str);
 
         size_t line_len = strlen(line->line);
-        buf = realloc(buf, buf_size + line_len + action_str_len);
-        buf_size += line_len + action_str_len;
 
-        memcpy(buf + buf_cur, action_str, action_str_len);
+        while (buf_size + line_len + ACTION_STR_LEN > buf_capacity) {
-        buf_cur += action_str_len;
+            buf_capacity *= 2;
+            buf = realloc(buf, buf_capacity);
+        }
+
+        buf_size += line_len + ACTION_STR_LEN;
+
+        memcpy(buf + buf_cur, action_str, ACTION_STR_LEN);
+        buf_cur += ACTION_STR_LEN;
         memcpy(buf + buf_cur, line->line, line_len);
         buf_cur += line_len;
 
         line = line->next;
         (*count)++;
     }
-    buf = realloc(buf, buf_size + 1);
+
+    if (buf_size + 1 > buf_capacity) {
+        buf = realloc(buf, buf_capacity + 1);
+    }
+
     *(buf + buf_cur) = '\0';
 
     *buf_len = buf_cur;
@@ -123,7 +147,7 @@ void *create_bulk_buffer(int max, int *count, size_t *buf_len) {
 }
 
 void print_errors(response_t *r) {
-    char * tmp = malloc(r->size + 1);
+    char *tmp = malloc(r->size + 1);
     memcpy(tmp, r->body, r->size);
     *(tmp + r->size) = '\0';
 
@@ -181,6 +205,15 @@ void _elastic_flush(int max) {
         _elastic_flush(max / 2);
         return;
 
+    } else if (r->status_code == 429) {
+
+        free_response(r);
+        free(buf);
+        LOG_WARNING("elastic.c", "Got 429 status, will retry after delay")
+        usleep(1000000 * 20);
+        _elastic_flush(max);
+        return;
+
     } else if (r->status_code != 200) {
         print_errors(r);
         delete_queue(Indexer->queued);
@@ -257,7 +290,7 @@ es_indexer_t *create_indexer(const char *url) {
     return indexer;
 }
 
-void destroy_indexer(char *script, char index_id[UUID_STR_LEN]) {
+void finish_indexer(char *script, char *index_id) {
 
     char url[4096];
 
@@ -269,22 +302,22 @@ void destroy_indexer(char *script, char index_id[UUID_STR_LEN]) {
     if (script != NULL) {
         execute_update_script(script, index_id);
         free(script);
-    }
 
-    snprintf(url, sizeof(url), "%s/sist2/_refresh", IndexCtx.es_url);
+        snprintf(url, sizeof(url), "%s/sist2/_refresh", IndexCtx.es_url);
-    r = web_post(url, "");
+        r = web_post(url, "");
-    LOG_INFOF("elastic.c", "Refresh index <%d>", r->status_code);
+        LOG_INFOF("elastic.c", "Refresh index <%d>", r->status_code);
-    free_response(r);
+        free_response(r);
+    }
 
     snprintf(url, sizeof(url), "%s/sist2/_forcemerge", IndexCtx.es_url);
     r = web_post(url, "");
     LOG_INFOF("elastic.c", "Merge index <%d>", r->status_code);
     free_response(r);
 
-    if (Indexer != NULL) {
+    snprintf(url, sizeof(url), "%s/sist2/_settings", IndexCtx.es_url);
-        free(Indexer->es_url);
+    r = web_put(url, "{\"index\":{\"refresh_interval\":\"1s\"}}");
-        free(Indexer);
+    LOG_INFOF("elastic.c", "Set refresh interval <%d>", r->status_code);
-    }
+    free_response(r);
 }
 
 void elastic_init(int force_reset) {
@@ -292,7 +325,7 @@ void elastic_init(int force_reset) {
     // Check if index exists
     char url[4096];
     snprintf(url, 4096, "%s/sist2", IndexCtx.es_url);
-    response_t *r = web_get(url);
+    response_t *r = web_get(url, 30);
     int index_exists = r->status_code == 200;
     free_response(r);
 
@@ -337,7 +370,7 @@ cJSON *elastic_get_document(const char *uuid_str) {
     char url[4096];
     snprintf(url, 4096, "%s/sist2/_doc/%s", WebCtx.es_url, uuid_str);
 
-    response_t *r = web_get(url);
+    response_t *r = web_get(url, 3);
     cJSON *json = NULL;
     if (r->status_code == 200) {
         json = cJSON_Parse(r->body);
@@ -351,7 +384,7 @@ char *elastic_get_status() {
     snprintf(url, 4096,
              "%s/_cluster/state/metadata/sist2?filter_path=metadata.indices.*.state", WebCtx.es_url);
 
-    response_t *r = web_get(url);
+    response_t *r = web_get(url, 30);
     cJSON *json = NULL;
     char *status = malloc(128 * sizeof(char));
     status[0] = '\0';

src/index/elastic.h

@@ -16,15 +16,14 @@ typedef struct es_indexer es_indexer_t;
 
 void elastic_index_line(es_bulk_line_t *line);
 
-void elastic_flush();
-
 void print_json(cJSON *document, const char uuid_str[UUID_STR_LEN]);
 
 void index_json(cJSON *document, const char uuid_str[UUID_STR_LEN]);
 
 es_indexer_t *create_indexer(const char* es_url);
 
-void destroy_indexer(char *script, char index_id[UUID_STR_LEN]);
+void elastic_cleanup();
+void finish_indexer(char *script, char *index_id);
 
 void elastic_init(int force_reset);
 

src/index/web.c

@@ -1,11 +1,19 @@
 #include "web.h"
 #include "src/sist.h"
-#include "src/ctx.h"
 
 #include <mongoose.h>
 #include <pthread.h>
+#include <curl/curl.h>
 
 
+size_t write_cb(char *ptr, size_t size, size_t nmemb, void *user_data) {
+
+    size_t real_size = size * nmemb;
+    dyn_buffer_t *buf = user_data;
+    dyn_buffer_write(buf, ptr, real_size);
+    return real_size;
+}
+
 void free_response(response_t *resp) {
     if (resp->body != NULL) {
         free(resp->body);
@@ -75,7 +83,7 @@ subreq_ctx_t *http_req(const char *url, const char *extra_headers, const char *p
     subreq_ctx_t *ctx = malloc(sizeof(subreq_ctx_t));
     mg_mgr_init(&ctx->mgr, NULL);
 
-    char address[8196];
+    char address[8192];
     snprintf(address, sizeof(address), "tcp://%.*s:%u", (int) host.len, host.p, port);
     struct mg_connection *nc = mg_connect(&ctx->mgr, address, http_req_ev);
     nc->user_data = &ctx->ev_data;
@@ -100,55 +108,126 @@ subreq_ctx_t *http_req(const char *url, const char *extra_headers, const char *p
     return ctx;
 }
 
-response_t *web_get(const char *url) {
-    subreq_ctx_t *ctx = http_req(url, SIST2_HEADERS, NULL, "GET");
-    while (ctx->ev_data.done == FALSE) {
-        mg_mgr_poll(&ctx->mgr, 50);
-    }
-    mg_mgr_free(&ctx->mgr);
-
-    response_t *ret = ctx->ev_data.resp;
-    free(ctx);
-    return ret;
-}
-
 subreq_ctx_t *web_post_async(const char *url, const char *data) {
     return http_req(url, SIST2_HEADERS, data, "POST");
 }
 
-response_t *web_post(const char *url, const char *data) {
+response_t *web_get(const char *url, int timeout) {
-    subreq_ctx_t *ctx = http_req(url, SIST2_HEADERS, data, "POST");
+    response_t *resp = malloc(sizeof(response_t));
 
-    while (ctx->ev_data.done == FALSE) {
+    CURL *curl;
-        mg_mgr_poll(&ctx->mgr, 50);
+    dyn_buffer_t buffer = dyn_buffer_create();
-    }
-    mg_mgr_free(&ctx->mgr);
 
-    response_t *ret = ctx->ev_data.resp;
+    curl = curl_easy_init();
-    free(ctx);
+    curl_easy_setopt(curl, CURLOPT_URL, url);
-    return ret;
+    curl_easy_setopt(curl, CURLOPT_WRITEDATA, (void *) (&buffer));
+    curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, write_cb);
+    curl_easy_setopt(curl, CURLOPT_USERAGENT, "sist2");
+    curl_easy_setopt(curl, CURLOPT_TIMEOUT, timeout);
+
+    struct curl_slist *headers = curl_slist_append(headers, "Content-Type: application/json");
+    curl_easy_setopt(curl, CURLOPT_HTTPHEADER, headers);
+
+    curl_easy_perform(curl);
+    curl_easy_getinfo(curl, CURLINFO_RESPONSE_CODE, &resp->status_code);
+
+    curl_easy_cleanup(curl);
+    curl_slist_free_all(headers);
+
+    resp->body = buffer.buf;
+    resp->size = buffer.cur;
+    return resp;
 }
 
-response_t *web_put(const char *url, const char *data) {
+response_t *web_post(const char *url, const char *data) {
-    subreq_ctx_t *ctx = http_req(url, SIST2_HEADERS, data, "PUT");
-    while (ctx->ev_data.done == FALSE) {
-        mg_mgr_poll(&ctx->mgr, 50);
-    }
-    mg_mgr_free(&ctx->mgr);
 
-    response_t *ret = ctx->ev_data.resp;
+    response_t *resp = malloc(sizeof(response_t));
-    free(ctx);
+
-    return ret;
+    CURL *curl;
+    dyn_buffer_t buffer = dyn_buffer_create();
+
+    curl = curl_easy_init();
+    curl_easy_setopt(curl, CURLOPT_URL, url);
+    curl_easy_setopt(curl, CURLOPT_WRITEDATA, (void *) (&buffer));
+    curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, write_cb);
+    curl_easy_setopt(curl, CURLOPT_POST, 1);
+    curl_easy_setopt(curl, CURLOPT_USERAGENT, "sist2");
+
+    struct curl_slist *headers = curl_slist_append(headers, "Content-Type: application/json");
+    curl_easy_setopt(curl, CURLOPT_HTTPHEADER, headers);
+
+    curl_easy_setopt(curl, CURLOPT_POSTFIELDS, data);
+
+    curl_easy_perform(curl);
+    curl_easy_getinfo(curl, CURLINFO_RESPONSE_CODE, &resp->status_code);
+
+    curl_easy_cleanup(curl);
+    curl_slist_free_all(headers);
+
+    resp->body = buffer.buf;
+    resp->size = buffer.cur;
+
+    return resp;
+}
+
+
+response_t *web_put(const char *url, const char *data) {
+
+    response_t *resp = malloc(sizeof(response_t));
+
+    CURL *curl;
+    dyn_buffer_t buffer = dyn_buffer_create();
+
+    curl = curl_easy_init();
+    curl_easy_setopt(curl, CURLOPT_URL, url);
+    curl_easy_setopt(curl, CURLOPT_WRITEDATA, (void *) (&buffer));
+    curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, write_cb);
+    curl_easy_setopt(curl, CURLOPT_CUSTOMREQUEST, "PUT");
+    curl_easy_setopt(curl, CURLOPT_USERAGENT, "sist2");
+    curl_easy_setopt(curl, CURLOPT_DNS_USE_GLOBAL_CACHE, 0);
+    curl_easy_setopt(curl, CURLOPT_IPRESOLVE, CURLOPT_DNS_LOCAL_IP4 );
+
+    struct curl_slist *headers = curl_slist_append(headers, "Content-Type: application/json");
+    curl_easy_setopt(curl, CURLOPT_HTTPHEADER, headers);
+
+    curl_easy_setopt(curl, CURLOPT_POSTFIELDS, data);
+
+    curl_easy_perform(curl);
+    curl_easy_getinfo(curl, CURLINFO_RESPONSE_CODE, &resp->status_code);
+
+    curl_easy_cleanup(curl);
+    curl_slist_free_all(headers);
+
+    resp->body = buffer.buf;
+    resp->size = buffer.cur;
+    return resp;
 }
 
 response_t *web_delete(const char *url) {
-    subreq_ctx_t *ctx = http_req(url, SIST2_HEADERS, NULL, "DELETE");
-    while (ctx->ev_data.done == FALSE) {
-        mg_mgr_poll(&ctx->mgr, 50);
-    }
-    mg_mgr_free(&ctx->mgr);
 
-    response_t *ret = ctx->ev_data.resp;
+    response_t *resp = malloc(sizeof(response_t));
-    free(ctx);
+
-    return ret;
+    CURL *curl;
+    dyn_buffer_t buffer = dyn_buffer_create();
+
+    curl = curl_easy_init();
+    curl_easy_setopt(curl, CURLOPT_URL, url);
+    curl_easy_setopt(curl, CURLOPT_WRITEDATA, (void *) (&buffer));
+    curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, write_cb);
+    curl_easy_setopt(curl, CURLOPT_CUSTOMREQUEST, "DELETE");
+    curl_easy_setopt(curl, CURLOPT_USERAGENT, "sist2");
+
+    curl_easy_setopt(curl, CURLOPT_POSTFIELDS, "");
+    struct curl_slist *headers = curl_slist_append(headers, "Content-Type: application/json");
+    curl_easy_setopt(curl, CURLOPT_HTTPHEADER, headers);
+
+    curl_easy_perform(curl);
+    curl_easy_getinfo(curl, CURLINFO_RESPONSE_CODE, &resp->status_code);
+
+    curl_easy_cleanup(curl);
+    curl_slist_free_all(headers);
+
+    resp->body = buffer.buf;
+    resp->size = buffer.cur;
+    return resp;
 }

src/index/web.h

@@ -20,7 +20,7 @@ typedef struct {
     struct mg_mgr mgr;
 } subreq_ctx_t;
 
-response_t *web_get(const char *url);
+response_t *web_get(const char *url, int timeout);
 response_t *web_post(const char * url, const char * data);
 subreq_ctx_t *web_post_async(const char *url, const char *data);
 response_t *web_put(const char *url, const char *data);

src/io/serialize.c

@@ -62,7 +62,7 @@ index_descriptor_t read_index_descriptor(char *path) {
     int fd = open(path, O_RDONLY);
 
     if (fd == -1) {
-        LOG_FATALF("serialize.c", "Invalid/corrupt index (Could not find descriptor): %s: %s\n", path ,strerror(errno))
+        LOG_FATALF("serialize.c", "Invalid/corrupt index (Could not find descriptor): %s: %s\n", path, strerror(errno))
     }
 
     char *buf = malloc(info.st_size + 1);
@@ -172,8 +172,8 @@ void write_document(document_t *doc) {
     dyn_buffer_t buf = dyn_buffer_create();
 
     // Ignore root directory in the file path
-    doc->ext = doc->ext - ScanCtx.index.desc.root_len;
+    doc->ext = (short) (doc->ext - ScanCtx.index.desc.root_len);
-    doc->base = doc->base - ScanCtx.index.desc.root_len;
+    doc->base = (short) (doc->base - ScanCtx.index.desc.root_len);
     doc->filepath += ScanCtx.index.desc.root_len;
 
     dyn_buffer_write(&buf, doc, sizeof(line_t));
@@ -230,7 +230,7 @@ void read_index_bin(const char *path, const char *index_id, index_func func) {
         char uuid_str[UUID_STR_LEN];
         uuid_unparse(line.uuid, uuid_str);
 
-        const char* mime_text = mime_get_mime_text(line.mime);
+        const char *mime_text = mime_get_mime_text(line.mime);
         if (mime_text == NULL) {
             cJSON_AddNullToObject(document, "mime");
         } else {
@@ -239,12 +239,20 @@ void read_index_bin(const char *path, const char *index_id, index_func func) {
         cJSON_AddNumberToObject(document, "size", (double) line.size);
         cJSON_AddNumberToObject(document, "mtime", line.mtime);
 
-        int c;
+        int c = 0;
         while ((c = getc(file)) != 0) {
             dyn_buffer_write_char(&buf, (char) c);
         }
         dyn_buffer_write_char(&buf, '\0');
 
+        if (IndexCtx.tags != NULL) {
+            const char *tags_string = g_hash_table_lookup(IndexCtx.tags, buf.buf);
+            if (tags_string != NULL) {
+                cJSON *tags_arr = cJSON_Parse(tags_string);
+                cJSON_AddItemToObject(document, "tag", tags_arr);
+            }
+        }
+
         cJSON_AddStringToObject(document, "extension", buf.buf + line.ext);
         if (*(buf.buf + line.ext - 1) == '.') {
             *(buf.buf + line.ext - 1) = '\0';

src/io/store.c

@@ -1,9 +1,10 @@
 #include "store.h"
 #include "src/ctx.h"
 
-store_t *store_create(char *path) {
+store_t *store_create(char *path, size_t chunk_size) {
 
     store_t *store = malloc(sizeof(struct store_t));
+    store->chunk_size = chunk_size;
     pthread_rwlock_init(&store->lock, NULL);
 
     mdb_env_create(&store->env);
@@ -18,7 +19,7 @@ store_t *store_create(char *path) {
         LOG_FATALF("store.c", "Error while opening store: %s (%s)\n", mdb_strerror(open_ret), path)
     }
 
-    store->size = (size_t) 1024 * 1024 * 5;
+    store->size = (size_t) store->chunk_size;
     ScanCtx.stat_tn_size = 0;
     mdb_env_set_mapsize(store->env, store->size);
 
@@ -69,7 +70,7 @@ void store_write(store_t *store, char *key, size_t key_len, char *buf, size_t bu
         // Cannot resize when there is a opened transaction.
         //  Resize take effect on the next commit.
         pthread_rwlock_wrlock(&store->lock);
-        store->size += 1024 * 1024 * 50;
+        store->size += store->chunk_size;
         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);
@@ -110,3 +111,40 @@ char *store_read(store_t *store, char *key, size_t key_len, size_t *ret_vallen)
     return buf;
 }
 
+GHashTable *store_read_all(store_t *store) {
+
+    int count = 0;
+
+    GHashTable *table = g_hash_table_new_full(g_str_hash, g_str_equal, free, free);
+
+    MDB_txn *txn = NULL;
+    mdb_txn_begin(store->env, NULL, MDB_RDONLY, &txn);
+
+    MDB_cursor *cur = NULL;
+    mdb_cursor_open(txn, store->dbi, &cur);
+
+    MDB_val key;
+    MDB_val value;
+
+    while (mdb_cursor_get(cur, &key, &value, MDB_NEXT) == 0) {
+        char *key_str = malloc(key.mv_size);
+        memcpy(key_str, key.mv_data, key.mv_size);
+        char *val_str = malloc(value.mv_size);
+        memcpy(val_str, value.mv_data, value.mv_size);
+
+        g_hash_table_insert(table, key_str, val_str);
+        count += 1;
+    }
+
+    LOG_DEBUGF("store.c", "Read tags for %d documents", count);
+
+    mdb_cursor_close(cur);
+    mdb_txn_abort(txn);
+    return table;
+}
+
+
+void store_copy(store_t *store, const char *destination) {
+    mkdir(destination, S_IWUSR | S_IRUSR | S_IXUSR);
+    mdb_env_copy(store->env, destination);
+}

src/io/store.h

@@ -4,14 +4,20 @@
 #include <pthread.h>
 #include <lmdb.h>
 
+#include <glib.h>
+
+#define STORE_SIZE_TN 1024 * 1024 * 5
+#define STORE_SIZE_TAG 1024 * 16
+
 typedef struct store_t {
     MDB_dbi dbi;
     MDB_env *env;
     size_t size;
+    size_t chunk_size;
     pthread_rwlock_t lock;
 } store_t;
 
-store_t *store_create(char *path);
+store_t *store_create(char *path, size_t chunk_size);
 
 void store_destroy(store_t *store);
 
@@ -19,4 +25,8 @@ void store_write(store_t *store, char *key, size_t key_len, char *buf, size_t bu
 
 char *store_read(store_t *store, char *key, size_t key_len, size_t *ret_vallen);
 
+GHashTable *store_read_all(store_t *store);
+
+void store_copy(store_t *store, const char *destination);
+
 #endif

src/main.c

@@ -21,7 +21,7 @@
 #define EPILOG "Made by simon987 <me@simon987.net>. Released under GPL-3.0"
 
 
-static const char *const Version = "2.5.0";
+static const char *const Version = "2.7.4";
 static const char *const usage[] = {
         "sist2 scan [OPTION]... PATH",
         "sist2 index [OPTION]... INDEX",
@@ -75,7 +75,7 @@ void _logf(const char *filepath, int level, char *format, ...) {
 
     va_start(args, format);
     if (level == LEVEL_FATAL) {
-        sist_logf(filepath, level, format, args);
+        vsist_logf(filepath, level, format, args);
         exit(-1);
     }
 
@@ -85,7 +85,7 @@ void _logf(const char *filepath, int level, char *format, ...) {
                 vsist_logf(filepath, level, format, args);
             }
         } else {
-            sist_logf(filepath, level, format, args);
+            vsist_logf(filepath, level, format, args);
         }
     }
     va_end(args);
@@ -99,11 +99,14 @@ void initialize_scan_context(scan_args_t *args) {
     ScanCtx.arc_ctx.logf = _logf;
     ScanCtx.arc_ctx.parse = (parse_callback_t) parse;
 
-    // Cbr
+    // Comic
-    ScanCtx.cbr_ctx.log = _log;
+    ScanCtx.comic_ctx.log = _log;
-    ScanCtx.cbr_ctx.logf = _logf;
+    ScanCtx.comic_ctx.logf = _logf;
-    ScanCtx.cbr_ctx.store = _store;
+    ScanCtx.comic_ctx.store = _store;
-    ScanCtx.cbr_ctx.cbr_mime = mime_get_mime_by_string(ScanCtx.mime_table, "application/x-cbr");
+    ScanCtx.comic_ctx.tn_size = args->size;
+    ScanCtx.comic_ctx.tn_qscale = args->quality;
+    ScanCtx.comic_ctx.cbr_mime = mime_get_mime_by_string(ScanCtx.mime_table, "application/x-cbr");
+    ScanCtx.comic_ctx.cbz_mime = mime_get_mime_by_string(ScanCtx.mime_table, "application/x-cbz");
 
     // Ebook
     pthread_mutex_init(&ScanCtx.ebook_ctx.mupdf_mutex, NULL);
@@ -134,6 +137,7 @@ void initialize_scan_context(scan_args_t *args) {
     ScanCtx.ooxml_ctx.content_size = args->content_size;
     ScanCtx.ooxml_ctx.log = _log;
     ScanCtx.ooxml_ctx.logf = _logf;
+    ScanCtx.ooxml_ctx.store = _store;
 
     // MOBI
     ScanCtx.mobi_ctx.content_size = args->content_size;
@@ -176,7 +180,7 @@ void sist2_scan(scan_args_t *args) {
     char store_path[PATH_MAX];
     snprintf(store_path, PATH_MAX, "%sthumbs", ScanCtx.index.path);
     mkdir(store_path, S_IWUSR | S_IRUSR | S_IXUSR);
-    ScanCtx.index.store = store_create(store_path);
+    ScanCtx.index.store = store_create(store_path, STORE_SIZE_TN);
 
     scan_print_header();
 
@@ -211,7 +215,7 @@ void sist2_scan(scan_args_t *args) {
         LOG_INFOF("main.c", "Loaded %d items in to mtime table.", g_hash_table_size(ScanCtx.original_table))
     }
 
-    ScanCtx.pool = tpool_create(args->threads, thread_cleanup);
+    ScanCtx.pool = tpool_create(args->threads, thread_cleanup, TRUE);
     tpool_start(ScanCtx.pool);
     walk_directory_tree(ScanCtx.index.desc.root);
     tpool_wait(ScanCtx.pool);
@@ -223,7 +227,7 @@ void sist2_scan(scan_args_t *args) {
         char dst_path[PATH_MAX];
         snprintf(store_path, PATH_MAX, "%sthumbs", args->incremental);
         snprintf(dst_path, PATH_MAX, "%s_index_original", ScanCtx.index.path);
-        store_t *source = store_create(store_path);
+        store_t *source = store_create(store_path, STORE_SIZE_TN);
 
         DIR *dir = opendir(args->incremental);
         if (dir == NULL) {
@@ -240,6 +244,13 @@ void sist2_scan(scan_args_t *args) {
         }
         closedir(dir);
         store_destroy(source);
+
+        snprintf(store_path, PATH_MAX, "%stags", args->incremental);
+        snprintf(dst_path, PATH_MAX, "%stags", ScanCtx.index.path);
+        mkdir(store_path, S_IWUSR | S_IRUSR | S_IXUSR);
+        store_t *source_tags = store_create(store_path, STORE_SIZE_TAG);
+        store_copy(source_tags, dst_path);
+        store_destroy(source_tags);
     }
 
     store_destroy(ScanCtx.index.store);
@@ -271,6 +282,12 @@ void sist2_index(index_args_t *args) {
         LOG_FATALF("main.c", "Could not open index %s: %s", args->index_path, strerror(errno))
     }
 
+    char path_tmp[PATH_MAX];
+    snprintf(path_tmp, sizeof(path_tmp), "%s/tags", args->index_path);
+    mkdir(path_tmp, S_IWUSR | S_IRUSR | S_IXUSR);
+    IndexCtx.tag_store = store_create(path_tmp, STORE_SIZE_TAG);
+    IndexCtx.tags = store_read_all(IndexCtx.tag_store);
+
     index_func f;
     if (args->print) {
         f = print_json;
@@ -278,6 +295,16 @@ void sist2_index(index_args_t *args) {
         f = index_json;
     }
 
+    void (*cleanup)();
+    if (args->print) {
+        cleanup = NULL;
+    } else {
+        cleanup = elastic_cleanup;
+    }
+
+    IndexCtx.pool = tpool_create(args->threads, cleanup, FALSE);
+    tpool_start(IndexCtx.pool);
+
     struct dirent *de;
     while ((de = readdir(dir)) != NULL) {
         if (strncmp(de->d_name, "_index_", sizeof("_index_") - 1) == 0) {
@@ -288,10 +315,16 @@ void sist2_index(index_args_t *args) {
     }
     closedir(dir);
 
+    tpool_wait(IndexCtx.pool);
+
     if (!args->print) {
-        elastic_flush();
+        finish_indexer(args->script, desc.uuid);
-        destroy_indexer(args->script, desc.uuid);
     }
+    tpool_destroy(IndexCtx.pool);
+
+    store_destroy(IndexCtx.tag_store);
+    g_hash_table_remove_all(IndexCtx.tags);
+    g_hash_table_destroy(IndexCtx.tags);
 }
 
 void sist2_exec_script(exec_args_t *args) {
@@ -317,6 +350,7 @@ void sist2_web(web_args_t *args) {
     WebCtx.auth_user = args->auth_user;
     WebCtx.auth_pass = args->auth_pass;
     WebCtx.auth_enabled = args->auth_enabled;
+    WebCtx.tag_auth_enabled = args->tag_auth_enabled;
 
     for (int i = 0; i < args->index_count; i++) {
         char *abs_path = abspath(args->indices[i]);
@@ -326,7 +360,11 @@ void sist2_web(web_args_t *args) {
         char path_tmp[PATH_MAX];
 
         snprintf(path_tmp, PATH_MAX, "%sthumbs", abs_path);
-        WebCtx.indices[i].store = store_create(path_tmp);
+        WebCtx.indices[i].store = store_create(path_tmp, STORE_SIZE_TN);
+
+        snprintf(path_tmp, PATH_MAX, "%stags", abs_path);
+        mkdir(path_tmp, S_IWUSR | S_IRUSR | S_IXUSR);
+        WebCtx.indices[i].tag_store = store_create(path_tmp, STORE_SIZE_TAG);
 
         snprintf(path_tmp, PATH_MAX, "%sdescriptor.json", abs_path);
         WebCtx.indices[i].desc = read_index_descriptor(path_tmp);
@@ -352,6 +390,7 @@ int main(int argc, const char *argv[]) {
 
     char *common_es_url = NULL;
     char *common_script_path = NULL;
+    int common_threads = 0;
 
     struct argparse_option options[] = {
             OPT_HELP(),
@@ -361,7 +400,7 @@ int main(int argc, const char *argv[]) {
             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_INTEGER('t', "threads", &common_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=5"),
             OPT_INTEGER(0, "size", &scan_args->size,
@@ -389,6 +428,7 @@ int main(int argc, const char *argv[]) {
                         "(see USAGE.md). DEFAULT: 2000"),
 
             OPT_GROUP("Index options"),
+            OPT_INTEGER('t', "threads", &common_threads, "Number of threads. DEFAULT=1"),
             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", &common_script_path, "Path to user script."),
@@ -400,6 +440,7 @@ int main(int argc, const char *argv[]) {
             OPT_STRING(0, "es-url", &common_es_url, "Elasticsearch url. DEFAULT=http://localhost:9200"),
             OPT_STRING(0, "bind", &web_args->listen_address, "Listen on this address. DEFAULT=localhost:4090"),
             OPT_STRING(0, "auth", &web_args->credentials, "Basic auth in user:password format"),
+            OPT_STRING(0, "tag-auth", &web_args->tag_credentials, "Basic auth in user:password format for tagging"),
 
             OPT_GROUP("Exec-script options"),
             OPT_STRING(0, "script-file", &common_script_path, "Path to user script."),
@@ -423,8 +464,11 @@ int main(int argc, const char *argv[]) {
 
     web_args->es_url = common_es_url;
     index_args->es_url = common_es_url;
+    exec_args->es_url = common_es_url;
     index_args->script_path = common_script_path;
     exec_args->script_path = common_script_path;
+    index_args->threads = common_threads;
+    scan_args->threads = common_threads;
 
     if (argc == 0) {
         argparse_usage(&argparse);

src/parsing/mime_generated.c

@@ -108,7 +108,7 @@ enum mime {
     application_x_bzip=655460,
     application_x_bzip2=655461 | 0x08000000,
     application_x_cbr=655462,
-    application_x_cbz=655463 | 0x40000000,
+    application_x_cbz=655463,
     application_x_cdlink=655464,
     application_x_chat=655465,
     application_x_chrome_extension=655466,

src/parsing/parse.c

@@ -145,11 +145,10 @@ void parse(void *arg) {
                     (IS_ARC_FILTER(doc.mime) && should_parse_filtered_file(doc.filepath, doc.ext))
             )) {
         parse_archive(&ScanCtx.arc_ctx, &job->vfile, &doc);
-    } else if (ScanCtx.ooxml_ctx.content_size > 0 && IS_DOC(doc.mime)) {
+    } else if ((ScanCtx.ooxml_ctx.content_size > 0 || ScanCtx.media_ctx.tn_size > 0) && IS_DOC(doc.mime)) {
         parse_ooxml(&ScanCtx.ooxml_ctx, &job->vfile, &doc);
-
+    } else if (is_cbr(&ScanCtx.comic_ctx, doc.mime) || is_cbz(&ScanCtx.comic_ctx, doc.mime)) {
-    } else if (is_cbr(&ScanCtx.cbr_ctx, doc.mime)) {
+        parse_comic(&ScanCtx.comic_ctx, &job->vfile, &doc);
-        parse_cbr(&ScanCtx.cbr_ctx, &job->vfile, &doc);
     } else if (IS_MOBI(doc.mime)) {
         parse_mobi(&ScanCtx.mobi_ctx, &job->vfile, &doc);
     }

src/static/css/dark.css

@@ -121,7 +121,7 @@ body {
     background: #546b7a;
 }
 
-a:hover,.btn:hover {
+a:hover, .btn:hover {
     color: #fff;
 }
 
@@ -130,7 +130,11 @@ a:hover,.btn:hover {
 }
 
 .document {
-    padding: 0.5rem;
+    padding: 0.3rem;
+}
+
+.card-text:last-child {
+    margin-top: -1px;
 }
 
 .document p {
@@ -166,6 +170,12 @@ a:hover,.btn:hover {
     background-color: #FAAB3C;
 }
 
+.add-tag-button {
+    cursor: pointer;
+    color: #212529;
+    background-color: #e0e0e0;
+}
+
 .card-img-overlay {
     pointer-events: none;
     padding: 0.75rem;
@@ -191,6 +201,18 @@ a:hover,.btn:hover {
     margin-right: 3px;
 }
 
+.badge-delete {
+    margin-right: -2px;
+    margin-left: 2px;
+    margin-top: -1px;
+    font-family: monospace;
+    font-size: 90%;
+    background: rgba(0, 0, 0, 0.2);
+    padding: 0.1em 0.4em;
+    color: white;
+    cursor: pointer;
+}
+
 .badge-user {
     color: #212529;
     background-color: #e0e0e0;
@@ -200,7 +222,7 @@ a:hover,.btn:hover {
     display: block;
     min-width: 64px;
     max-width: 100%;
-    max-height: 175px;
+    max-height: 240px;
     margin: 0 auto 0;
     padding: 3px 3px 0;
     width: auto;
@@ -223,20 +245,6 @@ a:hover,.btn:hover {
     width: 100%;
 }
 
-@media screen and (min-width: 1500px) {
-    .container {
-        max-width: 1440px;
-    }
-
-    .bricklayer-column-sizer {
-        width: 20% !important;
-    }
-
-    .bricklayer-column {
-        max-width: 20%;
-    }
-}
-
 @media screen and (min-width: 1800px) {
     .container {
         max-width: 1550px;
@@ -433,6 +441,7 @@ option {
     .small-btn {
         display: none;
     }
+
     .large-btn {
         display: inherit;
     }
@@ -442,6 +451,7 @@ option {
     .small-btn {
         display: inherit;
     }
+
     .large-btn {
         display: none;
     }
@@ -512,3 +522,11 @@ svg {
 #graphs-card svg text {
     fill: #eee;
 }
+
+.wholerow {
+    outline: none !important;
+}
+
+.stat > .card-body {
+    padding: 0.7em 1.25em;
+}

src/static/css/light.css

@@ -70,7 +70,11 @@ body {
 }
 
 .document {
-    padding: 0.5rem;
+   padding: 0.3rem;
+}
+
+.card-text:last-child {
+    margin-top: -1px;
 }
 
 .document p {
@@ -106,11 +110,33 @@ body {
     background-color: #e0e0e0;
 }
 
+.badge {
+    margin-right: 3px;
+}
+
+.badge-delete {
+    margin-right: -2px;
+    margin-left: 2px;
+    margin-top: -1px;
+    font-family: monospace;
+    font-size: 90%;
+    background: rgba(0,0,0,0.2);
+    padding: 0.1em 0.4em;
+    color: white;
+    cursor: pointer;
+}
+
 .badge-text {
     color: #FFFFFF;
     background-color: #FAAB3C;
 }
 
+.add-tag-button {
+    cursor: pointer;
+    color: #212529;
+    background-color: #e0e0e0;
+}
+
 .card-img-overlay {
     pointer-events: none;
     padding: 0.75rem;
@@ -131,15 +157,12 @@ body {
     overflow: hidden;
 }
 
-.badge {
-    margin-right: 3px;
-}
 
 .fit {
     display: block;
     min-width: 64px;
     max-width: 100%;
-    max-height: 175px;
+    max-height: 240px;
     margin: 0 auto 0;
     padding: 3px 3px 0 3px;
     width: auto;
@@ -162,6 +185,10 @@ body {
     width: 100%;
 }
 
+.bricklayer {
+    /*max-width: 100%;*/
+}
+
 @media screen and (max-width: 1200px) {
     .bricklayer-column {
         max-width: 100%;
@@ -375,3 +402,11 @@ mark {
     float: right;
     margin-bottom: 10px;
 }
+
+.wholerow {
+    outline: none !important;
+}
+
+.stat > .card-body {
+    padding: 0.7em 1.25em;
+}

src/static/js/dom.js

@@ -153,26 +153,44 @@ function getTags(hit, mimeCategory) {
     // User tags
     if (hit["_source"].hasOwnProperty("tag")) {
         hit["_source"]["tag"].forEach(tag => {
-            const userTag = document.createElement("span");
+            tags.push(makeUserTag(tag, hit));
-            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 makeUserTag(tag, hit) {
+    const userTag = document.createElement("span");
+    userTag.setAttribute("class", "badge badge-pill badge-user");
+    userTag.setAttribute("title", tag.split("#")[0])
+
+    const tokens = tag.split("#");
+
+    if (tokens.length > 1) {
+        const bg = "#" + tokens[1];
+        const fg = lum(tokens[1]) > 50 ? "#000" : "#fff";
+        userTag.setAttribute("style", `background-color: ${bg}; color: ${fg}`);
+    }
+
+    const deleteButton = document.createElement("span");
+    deleteButton.setAttribute("class", "badge badge-pill badge-delete")
+    deleteButton.setAttribute("title", "Delete tag")
+    deleteButton.appendChild(document.createTextNode("X"));
+    deleteButton.addEventListener("click", () => {
+        deleteTag(tag, hit).then(() => {
+            userTag.remove();
+        });
+    });
+    userTag.addEventListener("mouseenter", () => userTag.appendChild(deleteButton));
+    userTag.addEventListener("mouseleave", () => deleteButton.remove());
+
+    const name = tokens[0].split(".")[tokens[0].split(".").length - 1];
+    userTag.appendChild(document.createTextNode(name));
+
+    return userTag;
+}
+
 function infoButtonCb(hit) {
     return () => {
         getDocumentInfo(hit["_id"]).then(doc => {
@@ -192,7 +210,8 @@ function infoButtonCb(hit) {
 
             const displayFields = new Set([
                 "mime", "size", "mtime", "path", "title", "width", "height", "duration", "audioc", "videoc",
-                "bitrate", "artist", "album", "album_artist", "genre", "title", "font_name", "tag"
+                "bitrate", "artist", "album", "album_artist", "genre", "title", "font_name", "tag", "author",
+                "modified_by"
             ]);
             Object.keys(doc)
                 .filter(key => key.startsWith("_keyword.") || key.startsWith("_text.") || displayFields.has(key) || key.startsWith("exif_"))
@@ -338,9 +357,31 @@ function createDocCard(hit) {
 
     docCardBody.appendChild(tagContainer);
 
+    attachTagContainerEventListener(tagContainer, hit);
     return docCard;
 }
 
+function attachTagContainerEventListener(tagContainer, hit) {
+    const sizeTag = Array.from(tagContainer.children).find(child => child.tagName === "SMALL");
+
+    const addTagButton = document.createElement("span");
+    addTagButton.setAttribute("class", "badge badge-pill add-tag-button");
+    addTagButton.appendChild(document.createTextNode("+Add"));
+
+    tagContainer.addEventListener("mouseenter", () => tagContainer.insertBefore(addTagButton, sizeTag));
+    tagContainer.addEventListener("mouseleave", () => addTagButton.remove());
+
+    addTagButton.addEventListener("click", () => {
+        tagBar.value = "";
+        currentDocToTag = hit;
+        currentTagCallback = tag => {
+            tagContainer.insertBefore(makeUserTag(tag, hit), sizeTag);
+        }
+        $("#tagModal").modal("show");
+        tagBar.focus();
+    });
+}
+
 function makeThumbnail(mimeCategory, hit, imgWrapper, small) {
 
     if (!hit["_source"].hasOwnProperty("thumbnail")) {
@@ -413,7 +454,6 @@ function createDocLine(hit) {
 
     if (hit["_source"].hasOwnProperty("parent")) {
         line.classList.add("sub-document");
-        isSubDocument = true;
     }
 
     const infoButton = makeInfoButton(hit);
@@ -486,6 +526,8 @@ function createDocLine(hit) {
     pathLine.appendChild(path);
     pathLine.appendChild(tagContainer);
 
+    attachTagContainerEventListener(tagContainer, hit);
+
     return line;
 }
 

src/static/js/search.js

@@ -1,4 +1,4 @@
-const SIZE = 40;
+const SIZE = 60;
 let mimeMap = [];
 let tagMap = [];
 let mimeTree;
@@ -6,6 +6,9 @@ let tagTree;
 
 let searchBar = document.getElementById("searchBar");
 let pathBar = document.getElementById("pathBar");
+let tagBar = document.getElementById("tagBar");
+let currentDocToTag = null;
+let currentTagCallback = null;
 let lastDoc = null;
 let reachedEnd = false;
 let docCount = 0;
@@ -109,13 +112,112 @@ window.onload = () => {
             searchDebounced();
         }
     });
+    new autoComplete({
+        selector: '#tagBar',
+        minChars: 1,
+        delay: 200,
+        renderItem: function (item) {
+            return '<div class="autocomplete-suggestion" data-val="' + item + '">' + item.split("#")[0] + '</div>';
+        },
+        source: async function (term, suggest) {
+            term = term.toLowerCase();
+
+            const choices = await getTagChoices();
+
+            let matches = [];
+            for (let i = 0; i < choices.length; i++) {
+                if (~choices[i].toLowerCase().indexOf(term)) {
+                    matches.push(choices[i]);
+                }
+            }
+            suggest(matches.sort());
+        },
+        onSelect: function (e, item) {
+            const name = item.split("#")[0];
+            const color = "#" + item.split("#")[1];
+            $("#tag-color").val(color);
+            $("#tag-color").trigger("keyup", color);
+            tagBar.value = name;
+            e.preventDefault();
+        }
+    });
+    [tagBar, document.getElementById("tag-color")].forEach(elem => {
+        elem.addEventListener("keyup", e => {
+            if (e.key === "Enter" && tagBar.value.length > 0) {
+                const tag = tagBar.value + document.getElementById("tag-color").value;
+                saveTag(tag, currentDocToTag).then(() => currentTagCallback(tag));
+            }
+        });
+    })
+    $("#tag-color").colorpicker({
+        format: "hex",
+        sliders: {
+            saturation: {
+                selector: '.colorpicker-saturation',
+                callLeft: 'setSaturationRatio',
+                callTop: 'setValueRatio'
+            },
+            hue: {
+                selector: '.colorpicker-hue',
+                maxLeft: 0,
+                callLeft: false,
+                callTop: 'setHueRatio'
+            }
+        }
+    });
 };
 
+function saveTag(tag, hit) {
+    const relPath = hit["_source"]["path"] + "/" + hit["_source"]["name"] + ext(hit);
+
+    return $.jsonPost("/tag/" + hit["_source"]["index"], {
+        delete: false,
+        name: tag,
+        doc_id: hit["_id"],
+        relpath: relPath
+    }).then(() => {
+        tagBar.blur();
+        $("#tagModal").modal("hide");
+        $.toast({
+            heading: "Tag added",
+            text: "Tag saved to index storage and updated in ElasticSearch",
+            stack: 3,
+            bgColor: "#00a4bc",
+            textColor: "#fff",
+            position: 'bottom-right',
+            hideAfter: 3000,
+            loaderBg: "#08c7e8",
+        });
+    })
+}
+
+function deleteTag(tag, hit) {
+    const relPath = hit["_source"]["path"] + "/" + hit["_source"]["name"] + ext(hit);
+
+    return $.jsonPost("/tag/" + hit["_source"]["index"], {
+        delete: true,
+        name: tag,
+        doc_id: hit["_id"],
+        relpath: relPath
+    }).then(() => {
+        $.toast({
+            heading: "Tag deleted",
+            text: "Tag deleted index storage and updated in ElasticSearch",
+            stack: 3,
+            bgColor: "#00a4bc",
+            textColor: "#fff",
+            position: 'bottom-right',
+            hideAfter: 3000,
+            loaderBg: "#08c7e8",
+        });
+    })
+}
+
 function toggleFuzzy() {
     searchDebounced();
 }
 
-$.jsonPost("i").then(resp => {
+$.get("i").then(resp => {
 
     const urlIndices = (new URLSearchParams(location.search)).get("i");
     resp["indices"].forEach(idx => {
@@ -151,7 +253,7 @@ function handleTreeClick(tree) {
 
         if (node.id === "any") {
             if (!node.itree.state.checked) {
-                tree.deselect();
+                tree.deselectDeep();
             }
         } else {
             tree.node("any").deselect();
@@ -234,6 +336,9 @@ $.jsonPost("es", {
         selection: {
             mode: 'checkbox'
         },
+        checkbox: {
+            autoCheckChildren: false
+        },
         data: tagMap
     });
     new InspireTreeDOM(tagTree, {
@@ -245,21 +350,70 @@ $.jsonPost("es", {
 });
 
 function addTag(map, tag, id, count) {
-    let tags = tag.split("#")[0].split(".");
+    // let tags = tag.split("#")[0].split(".");
+    let tags = tag.split(".");
 
     let child = {
         id: id,
-        text: tags.length !== 1 ? tags[0] : `${tags[0]} (${count})`,
+        values: [id],
+        count: count,
+        text: tags.length !== 1 ? tags[0] : `${tags[0].split("#")[0]} (${count})`,
         name: tags[0],
-        children: []
+        children: [],
+        isLeaf: tags.length === 1,
+        //Overwrite base functions
+        blur: function () {
+        },
+        select: function () {
+            this.state("selected", true);
+            return this.check()
+        },
+        deselect: function () {
+            this.state("selected", false);
+            return this.uncheck()
+        },
+        uncheck: function () {
+            if (!this.isLeaf) {
+                return;
+            }
+
+            baseStateChange('checked', false, 'unchecked', this, false);
+            this.state('indeterminate', false);
+
+            if (this.hasParent()) {
+                this.getParent().refreshIndeterminateState();
+            }
+
+            this._tree.end();
+            return this;
+        },
+        check: function () {
+            if (!this.isLeaf) {
+                return;
+            }
+
+            baseStateChange('checked', true, 'checked', this, false);
+
+            if (this.hasParent()) {
+                this.getParent().refreshIndeterminateState();
+            }
+
+            this._tree.end();
+            return this;
+        }
     };
 
     let found = false;
     map.forEach(node => {
-        if (node.name === child.name) {
+        if (node.name.split("#")[0] === child.name.split("#")[0]) {
             found = true;
             if (tags.length !== 1) {
                 addTag(node.children, tags.slice(1).join("."), id, count);
+            } else {
+                // Same name, different color
+                node.count += count;
+                node.text = `${tags[0].split("#")[0]} (${node.count})`;
+                node.values.push(id);
             }
         }
     });
@@ -311,7 +465,11 @@ function getSelectedNodes(tree) {
 
         //Only get children
         if (selected[i].text.indexOf("(") !== -1) {
-            selectedNodes.push(selected[i].id);
+            if (selected[i].values) {
+                selectedNodes.push(selected[i].values);
+            } else {
+                selectedNodes.push(selected[i].id);
+            }
         }
     }
 
@@ -374,7 +532,9 @@ function search(after = null) {
 
     let tags = getSelectedNodes(tagTree);
     if (!tags.includes("any")) {
-        filters.push({terms: {"tag": tags}});
+        tags.forEach(tagGroup => {
+            filters.push({terms: {"tag": tagGroup}})
+        })
     }
 
     if (date_min && date_max) {
@@ -618,6 +778,7 @@ function getNextDepth(node) {
                     text: `${name}/ (${bucket.doc_count})`,
                     depth: node.depth + 1,
                     index: node.index,
+                    values: [bucket.key],
                     children: true,
                 }
             }).filter(x => x !== null)
@@ -648,6 +809,7 @@ function createPathTree(target) {
     selectedIndices.forEach(index => {
         pathTree.addNode({
             id: "/" + index,
+            values: ["/" + index],
             text: `/[${indexMap[index]}]`,
             index: index,
             depth: 0,
@@ -676,5 +838,34 @@ function getPathChoices() {
                 }
             }
         }).then(resp => getPaths(resp["suggest"]["path"][0]["options"].map(opt => opt["_source"]["path"])));
-    })
+    });
+}
+
+
+function getTagChoices() {
+    return new Promise(getPaths => {
+        $.jsonPost("es", {
+            suggest: {
+                tag: {
+                    prefix: tagBar.value,
+                    completion: {
+                        field: "suggest-tag",
+                        skip_duplicates: true,
+                        size: 10000
+                    }
+                }
+            }
+        }).then(resp => {
+            const result = [];
+            resp["suggest"]["tag"][0]["options"].map(opt => opt["_source"]["tag"]).forEach(tags => {
+                tags.forEach(tag => {
+                    const t = tag.split("#")[0];
+                    if (!result.find(x => x.split("#")[0] === t)) {
+                        result.push(tag);
+                    }
+                });
+            });
+            getPaths(result);
+        });
+    });
 }

src/static/js/util.js

@@ -101,7 +101,8 @@ const _defaults = {
     treemapColor: "PuBuGn",
     treemapSize: "large",
     suggestPath: true,
-    fragmentSize: 100
+    fragmentSize: 100,
+    columns: 5
 };
 
 function loadSettings() {
@@ -118,6 +119,7 @@ function loadSettings() {
     $("#settingTreemapType").val(CONF.options.treemapType);
     $("#settingSuggestPath").prop("checked", CONF.options.suggestPath);
     $("#settingFragmentSize").val(CONF.options.fragmentSize);
+    $("#settingColumns").val(CONF.options.columns);
 }
 
 function Settings() {
@@ -125,6 +127,7 @@ function Settings() {
 
     this._onUpdate = function () {
         $("#fuzzyToggle").prop("checked", this.options.fuzzy);
+        updateColumnStyle();
     };
 
     this.load = function () {
@@ -161,6 +164,7 @@ function updateSettings() {
     CONF.options.treemapType = $("#settingTreemapType").val();
     CONF.options.suggestPath = $("#settingSuggestPath").prop("checked");
     CONF.options.fragmentSize = $("#settingFragmentSize").val();
+    CONF.options.columns = $("#settingColumns").val();
     CONF.save();
 
     if (typeof searchDebounced !== "undefined") {
@@ -203,3 +207,26 @@ function toggleTheme() {
     }
     window.location.reload();
 }
+
+function updateColumnStyle() {
+    const style = document.getElementById("style");
+    if (style) {
+        style.innerHTML =
+        `
+@media screen and (min-width: 1500px) {
+    .container {
+            max-width: 1440px;
+        }
+
+    .bricklayer-column-sizer {
+            width: ${100 / CONF.options.columns}% !important;
+        }
+
+    .bricklayer-column {
+            max-width: ${100 / CONF.options.columns}%;
+        }
+    }
+}
+        `
+    }
+}

src/static/search.html

@@ -6,15 +6,17 @@
     <meta name='viewport' content='width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no'/>
 
     <link href="css" rel="stylesheet" type="text/css">
+    <style id="style"></style>
 </head>
 <body>
 
 <nav class="navbar navbar-expand-lg">
     <a class="navbar-brand" href="/">sist2</a>
-    <span class="badge badge-pill version">2.5.0</span>
+    <span class="badge badge-pill version">2.7.4</span>
     <span class="tagline">Lightning-fast file system indexer and search tool </span>
     <a class="btn ml-auto" href="/stats">Stats</a>
-    <button class="btn" type="button" data-toggle="modal" data-target="#settings" onclick="loadSettings()">Settings</button>
+    <button class="btn" type="button" data-toggle="modal" data-target="#settings" onclick="loadSettings()">Settings
+    </button>
     <button class="btn" title="Toggle theme" onclick="toggleTheme()">Theme</button>
 </nav>
 
@@ -48,8 +50,11 @@
                 <div class="col">
                     <div class="input-group" style="margin-bottom: 0.5em; margin-top: 1em">
                         <div class="input-group-prepend">
-                            <button id="pathBarHelper" class="btn btn-outline-secondary" data-toggle="modal" data-target="#pathTreeModal">
+                            <button id="pathBarHelper" class="btn btn-outline-secondary" data-toggle="modal"
-                                <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 576 512" width="20px"><path d="M288 224h224a32 32 0 0 0 32-32V64a32 32 0 0 0-32-32H400L368 0h-80a32 32 0 0 0-32 32v64H64V8a8 8 0 0 0-8-8H40a8 8 0 0 0-8 8v392a16 16 0 0 0 16 16h208v64a32 32 0 0 0 32 32h224a32 32 0 0 0 32-32V352a32 32 0 0 0-32-32H400l-32-32h-80a32 32 0 0 0-32 32v64H64V128h192v64a32 32 0 0 0 32 32zm0 96h66.74l32 32H512v128H288zm0-288h66.74l32 32H512v128H288z"/></svg>
+                                    data-target="#pathTreeModal">
+                                <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 576 512" width="20px">
+                                    <path d="M288 224h224a32 32 0 0 0 32-32V64a32 32 0 0 0-32-32H400L368 0h-80a32 32 0 0 0-32 32v64H64V8a8 8 0 0 0-8-8H40a8 8 0 0 0-8 8v392a16 16 0 0 0 16 16h208v64a32 32 0 0 0 32 32h224a32 32 0 0 0 32-32V352a32 32 0 0 0-32-32H400l-32-32h-80a32 32 0 0 0-32 32v64H64V128h192v64a32 32 0 0 0 32 32zm0 96h66.74l32 32H512v128H288zm0-288h66.74l32 32H512v128H288z"/>
+                                </svg>
                             </button>
                         </div>
                         <input id="pathBar" type="search" class="form-control" placeholder="Filter path">
@@ -156,7 +161,8 @@
                         <i>fried eggs</i> and either <i>eggplant</i> or <i>potato</i>, but will ignore results
                         containing <i>frittata</i>.</p>
 
-                    <p>When neither <code>+</code> or <code>|</code> is specified, the default operator is <code>+</code> (and).</p>
+                    <p>When neither <code>+</code> or <code>|</code> is specified, the default operator is
+                        <code>+</code> (and).</p>
                     <p>When the <b>Fuzzy</b> option is checked, partial matches are also returned.</p>
                     <br>
                     <p>For more information, see <a target="_blank"
@@ -189,12 +195,14 @@
 
                     <div class="custom-control custom-checkbox">
                         <input type="checkbox" class="custom-control-input" id="settingSearchInPath">
-                        <label class="custom-control-label" for="settingSearchInPath">Enable matching query against document path</label>
+                        <label class="custom-control-label" for="settingSearchInPath">Enable matching query against
+                            document path</label>
                     </div>
 
                     <div class="custom-control custom-checkbox">
                         <input type="checkbox" class="custom-control-input" id="settingSuggestPath">
-                        <label class="custom-control-label" for="settingSuggestPath">Enable auto-complete in path filter bar</label>
+                        <label class="custom-control-label" for="settingSuggestPath">Enable auto-complete in path filter
+                            bar</label>
                     </div>
 
                     <br/>
@@ -209,6 +217,20 @@
                         <option value="list">List</option>
                     </select>
 
+                    <div class="form-group">
+                        <label for="settingColumns">Maximum column count</label>
+                        <select id="settingColumns" class="form-control form-control-sm">
+                            <option value="3">3</option>
+                            <option value="4">4</option>
+                            <option value="5">5</option>
+                            <option value="6">6</option>
+                            <option value="7">7</option>
+                            <option value="8">8</option>
+                            <option value="9">9</option>
+                            <option value="13">13</option>
+                        </select>
+                    </div>
+
                     <hr/>
                     <h4>Stats</h4>
 
@@ -288,6 +310,32 @@
         </div>
     </div>
 
+    <div class="modal" id="tagModal" tabindex="-1" role="dialog" aria-labelledby="modal-title" aria-hidden="true">
+        <div class="modal-dialog modal-dialog-centered" role="document">
+            <div class="modal-content">
+                <div class="modal-header">
+                    <h5 class="modal-title">Add tag</h5>
+                    <button type="button" class="close" data-dismiss="modal" aria-label="Close">
+                        <span aria-hidden="true">&times;</span>
+                    </button>
+                </div>
+
+                <div class="modal-body">
+                    <div class="form-group">
+                        <div class="row">
+                            <div class="col col-8">
+                                <input type="text" id="tagBar" class="form-control">
+                            </div>
+                            <div class="col col-4">
+                                <input type="text" id="tag-color" value="" class="form-control"/>
+                            </div>
+                        </div>
+                    </div>
+                </div>
+            </div>
+        </div>
+    </div>
+
     <div id="searchResults"></div>
 </div>
 

src/static/stats.html

@@ -10,7 +10,7 @@
 
 <nav class="navbar navbar-expand-lg">
     <a class="navbar-brand" href="/">sist2</a>
-    <span class="badge badge-pill version">2.5.0</span>
+    <span class="badge badge-pill version">2.7.4</span>
     <span class="tagline">Lightning-fast file system indexer and search tool </span>
     <a style="margin-left: auto" class="btn" href="/">Back</a>
     <button class="btn" type="button" data-toggle="modal" data-target="#settings"
@@ -94,6 +94,19 @@
                     <option value="list">List</option>
                 </select>
 
+                <div class="form-group">
+                    <label for="settingColumns">Maximum column count</label>
+                    <select id="settingColumns" class="form-control form-control-sm">
+                        <option value="3">3</option>
+                        <option value="4">4</option>
+                        <option value="5">5</option>
+                        <option value="6">6</option>
+                        <option value="7">7</option>
+                        <option value="8">8</option>
+                        <option value="9">9</option>
+                    </select>
+                </div>
+
                 <hr/>
                 <h4>Stats</h4>
 

src/tpool.c

@@ -3,6 +3,8 @@
 #include "sist.h"
 #include <pthread.h>
 
+#define MAX_QUEUE_SIZE 10000
+
 typedef void (*thread_func_t)(void *arg);
 
 typedef struct tpool_work {
@@ -26,6 +28,7 @@ typedef struct tpool {
     int work_cnt;
     int done_cnt;
 
+    int free_arg;
     int stop;
 
     void (*cleanup_func)();
@@ -79,6 +82,10 @@ int tpool_add_work(tpool_t *pool, thread_func_t func, void *arg) {
         return 0;
     }
 
+    while ((pool->work_cnt - pool->done_cnt) >= MAX_QUEUE_SIZE) {
+        usleep(100000);
+    }
+
     pthread_mutex_lock(&(pool->work_mutex));
     if (pool->work_head == NULL) {
         pool->work_head = work;
@@ -121,7 +128,9 @@ static void *tpool_worker(void *arg) {
             }
 
             work->func(work->arg);
-            free(work->arg);
+            if (pool->free_arg) {
+                free(work->arg);
+            }
             free(work);
         }
 
@@ -138,8 +147,10 @@ static void *tpool_worker(void *arg) {
         pthread_mutex_unlock(&(pool->work_mutex));
     }
 
-    LOG_INFO("tpool.c", "Executing cleaup function")
+    if (pool->cleanup_func != NULL) {
-    pool->cleanup_func();
+        LOG_INFO("tpool.c", "Executing cleanup function")
+        pool->cleanup_func();
+    }
 
     pthread_cond_signal(&(pool->working_cond));
     pthread_mutex_unlock(&(pool->work_mutex));
@@ -207,13 +218,14 @@ void tpool_destroy(tpool_t *pool) {
  * Create a thread pool
  * @param thread_cnt Worker threads count
  */
-tpool_t *tpool_create(size_t thread_cnt, void cleanup_func()) {
+tpool_t *tpool_create(size_t thread_cnt, void cleanup_func(), int free_arg) {
 
     tpool_t *pool = malloc(sizeof(tpool_t));
     pool->thread_cnt = thread_cnt;
     pool->work_cnt = 0;
     pool->done_cnt = 0;
     pool->stop = 0;
+    pool->free_arg = free_arg;
     pool->cleanup_func = cleanup_func;
     pool->threads = calloc(sizeof(pthread_t), thread_cnt);
 

src/tpool.h

@@ -8,7 +8,7 @@ typedef struct tpool tpool_t;
 
 typedef void (*thread_func_t)(void *arg);
 
-tpool_t *tpool_create(size_t num, void (*cleanup_func)());
+tpool_t *tpool_create(size_t num, void (*cleanup_func)(), int free_arg);
 void tpool_start(tpool_t *pool);
 void tpool_destroy(tpool_t *tm);
 

src/types.h

@@ -10,7 +10,7 @@ typedef struct index_descriptor {
     char version[64];
     long timestamp;
     char root[PATH_MAX];
-    char rewrite_url[8196];
+    char rewrite_url[8192];
     short root_len;
     char name[1024];
     char type[64];
@@ -19,6 +19,7 @@ typedef struct index_descriptor {
 typedef struct index_t {
     struct index_descriptor desc;
     struct store_t *store;
+    struct store_t *tag_store;
     char path[PATH_MAX];
 } index_t;
 

src/web/serve.c

@@ -53,6 +53,14 @@ store_t *get_store(const char *index_id) {
     return NULL;
 }
 
+store_t *get_tag_store(const char *index_id) {
+    index_t *idx = get_index_by_id(index_id);
+    if (idx != NULL) {
+        return idx->tag_store;
+    }
+    return NULL;
+}
+
 void search_index(struct mg_connection *nc) {
     send_response_line(nc, 200, sizeof(search_html), "Content-Type: text/html");
     mg_send(nc, search_html, sizeof(search_html));
@@ -103,7 +111,7 @@ void stats_files(struct mg_connection *nc, struct http_message *hm, struct mg_st
             return;
     }
 
-    char disposition[8196];
+    char disposition[8192];
     snprintf(disposition, sizeof(disposition), "Content-Disposition: inline; filename=\"%s\"", file);
 
     char full_path[PATH_MAX];
@@ -248,7 +256,7 @@ int serve_file_from_url(cJSON *json, index_t *idx, struct mg_connection *nc) {
 
     const char *ext = cJSON_GetObjectItem(json, "extension")->valuestring;
 
-    char url[8196];
+    char url[8192];
     snprintf(url, sizeof(url),
              "%s%s/%s%s%s",
              idx->desc.rewrite_url, path_unescaped, name_unescaped, strlen(ext) == 0 ? "" : ".", ext);
@@ -283,7 +291,7 @@ void serve_file_from_disk(cJSON *json, index_t *idx, struct mg_connection *nc, s
 
     LOG_DEBUGF("serve.c", "Serving file from disk: %s", full_path)
 
-    char disposition[8196];
+    char disposition[8192];
     snprintf(disposition, sizeof(disposition), "Content-Disposition: inline; filename=\"%s%s%s\"",
              name, strlen(ext) == 0 ? "" : ".", ext);
 
@@ -422,6 +430,177 @@ void status(struct mg_connection *nc) {
     nc->flags |= MG_F_SEND_AND_CLOSE;
 }
 
+typedef struct {
+    char *name;
+    int delete;
+    char *relpath;
+    char *doc_id;
+} tag_req_t;
+
+tag_req_t *parse_tag_request(cJSON *json) {
+
+    if (!cJSON_IsObject(json)) {
+        return NULL;
+    }
+
+    cJSON *arg_name = cJSON_GetObjectItem(json, "name");
+    if (arg_name == NULL || !cJSON_IsString(arg_name)) {
+        return NULL;
+    }
+
+    cJSON *arg_delete = cJSON_GetObjectItem(json, "delete");
+    if (arg_delete == NULL || !cJSON_IsBool(arg_delete)) {
+        return NULL;
+    }
+
+    cJSON *arg_relpath = cJSON_GetObjectItem(json, "relpath");
+    if (arg_relpath == NULL || !cJSON_IsString(arg_relpath)) {
+        return NULL;
+    }
+
+    cJSON *arg_doc_id = cJSON_GetObjectItem(json, "doc_id");
+    if (arg_doc_id == NULL || !cJSON_IsString(arg_doc_id)) {
+        return NULL;
+    }
+
+    tag_req_t *req = malloc(sizeof(tag_req_t));
+    req->delete = arg_delete->valueint;
+    req->name = arg_name->valuestring;
+    req->relpath = arg_relpath->valuestring;
+    req->doc_id = arg_doc_id->valuestring;
+
+    return req;
+}
+
+void tag(struct mg_connection *nc, struct http_message *hm, struct mg_str *path) {
+    if (path->len != UUID_STR_LEN + 4) {
+        LOG_DEBUGF("serve.c", "Invalid tag path: %.*s", (int) path->len, path->p)
+        mg_http_send_error(nc, 404, NULL);
+        nc->flags |= MG_F_SEND_AND_CLOSE;
+        return;
+    }
+
+    char arg_index[UUID_STR_LEN];
+    memcpy(arg_index, hm->uri.p + 5, UUID_STR_LEN);
+    *(arg_index + UUID_STR_LEN - 1) = '\0';
+
+    if (hm->body.len < 2 || hm->method.len != 4 || memcmp(&hm->method, "POST", 4) == 0) {
+        LOG_DEBUG("serve.c", "Invalid tag request")
+        mg_http_send_error(nc, 400, NULL);
+        nc->flags |= MG_F_SEND_AND_CLOSE;
+        return;
+    }
+
+    store_t *store = get_tag_store(arg_index);
+    if (store == NULL) {
+        LOG_DEBUGF("serve.c", "Could not get tag store for index: %s", arg_index)
+        mg_http_send_error(nc, 404, NULL);
+        nc->flags |= MG_F_SEND_AND_CLOSE;
+        return;
+    }
+
+    char *body = malloc(hm->body.len + 1);
+    memcpy(body, hm->body.p, hm->body.len);
+    *(body + hm->body.len) = '\0';
+    cJSON *json = cJSON_Parse(body);
+
+    tag_req_t *arg_req = parse_tag_request(json);
+    if (arg_req == NULL) {
+        LOG_DEBUGF("serve.c", "Could not parse tag request", arg_index)
+        cJSON_Delete(json);
+        free(body);
+        mg_http_send_error(nc, 400, NULL);
+        nc->flags |= MG_F_SEND_AND_CLOSE;
+        return;
+    }
+
+    cJSON *arr = NULL;
+
+    size_t data_len = 0;
+    const char *data = store_read(store, arg_req->relpath, strlen(arg_req->relpath), &data_len);
+    if (data_len == 0) {
+        arr = cJSON_CreateArray();
+    } else {
+        arr = cJSON_Parse(data);
+    }
+
+    if (arg_req->delete) {
+
+        if (data_len > 0) {
+            cJSON *element = NULL;
+            int i = 0;
+            cJSON_ArrayForEach(element, arr) {
+                if (strcmp(element->valuestring, arg_req->name) == 0) {
+                    cJSON_DeleteItemFromArray(arr, i);
+                    break;
+                }
+                i++;
+            }
+        }
+
+        char buf[8192];
+        snprintf(buf, sizeof(buf),
+                 "{"
+                 "    \"script\" : {"
+                 "        \"source\": \"if (ctx._source.tag.contains(params.tag)) { ctx._source.tag.remove(ctx._source.tag.indexOf(params.tag)) }\","
+                 "        \"lang\": \"painless\","
+                 "        \"params\" : {"
+                 "            \"tag\" : \"%s\""
+                 "        }"
+                 "    }"
+                 "}", arg_req->name
+        );
+
+        char url[4096];
+        snprintf(url, sizeof(url), "%s/sist2/_update/%s", WebCtx.es_url, arg_req->doc_id);
+        nc->user_data = web_post_async(url, buf);
+
+    } else {
+        cJSON_AddItemToArray(arr, cJSON_CreateString(arg_req->name));
+
+        char buf[8192];
+        snprintf(buf, sizeof(buf),
+                 "{"
+                 "    \"script\" : {"
+                 "        \"source\": \"if(ctx._source.tag == null) {ctx._source.tag = new ArrayList()} ctx._source.tag.add(params.tag)\","
+                 "        \"lang\": \"painless\","
+                 "        \"params\" : {"
+                 "            \"tag\" : \"%s\""
+                 "        }"
+                 "    }"
+                 "}", arg_req->name
+        );
+
+        char url[4096];
+        snprintf(url, sizeof(url), "%s/sist2/_update/%s", WebCtx.es_url, arg_req->doc_id);
+        nc->user_data = web_post_async(url, buf);
+    }
+
+    char *json_str = cJSON_PrintUnformatted(arr);
+    store_write(store, arg_req->relpath, strlen(arg_req->relpath) + 1, json_str, strlen(json_str) + 1);
+
+    free(arg_req);
+    free(json_str);
+    cJSON_Delete(json);
+    cJSON_Delete(arr);
+    free(body);
+}
+
+int validate_auth(struct mg_connection *nc, struct http_message *hm) {
+    char user[256] = {0,};
+    char pass[256] = {0,};
+
+    int ret = mg_get_http_basic_auth(hm, user, sizeof(user), pass, sizeof(pass));
+    if (ret == -1 || strcmp(user, WebCtx.auth_user) != 0 || strcmp(pass, WebCtx.auth_pass) != 0) {
+        mg_printf(nc, "HTTP/1.1 401 Unauthorized\r\n"
+                      "WWW-Authenticate: Basic realm=\"sist2\"\r\n"
+                      "Content-Length: 0\r\n\r\n");
+        nc->flags |= MG_F_SEND_AND_CLOSE;
+        return FALSE;
+    }
+    return TRUE;
+}
+
 static void ev_router(struct mg_connection *nc, int ev, void *p) {
     struct mg_str scheme;
     struct mg_str user_info;
@@ -442,15 +621,7 @@ static void ev_router(struct mg_connection *nc, int ev, void *p) {
 
 
         if (WebCtx.auth_enabled == TRUE) {
-            char user[256] = {0,};
+            if (!validate_auth(nc, hm)) {
-            char pass[256] = {0,};
-
-            int ret = mg_get_http_basic_auth(hm, user, sizeof(user), pass, sizeof(pass));
-            if (ret == -1 || strcmp(user, WebCtx.auth_user) != 0 || strcmp(pass, WebCtx.auth_pass) != 0) {
-                mg_printf(nc, "HTTP/1.1 401 Unauthorized\r\n"
-                              "WWW-Authenticate: Basic realm=\"sist2\"\r\n"
-                              "Content-Length: 0\r\n\r\n");
-                nc->flags |= MG_F_SEND_AND_CLOSE;
                 return;
             }
         }
@@ -479,6 +650,13 @@ static void ev_router(struct mg_connection *nc, int ev, void *p) {
             thumbnail(nc, hm, &path);
         } else if (has_prefix(&path, &((struct mg_str) MG_MK_STR("/s/")))) {
             stats_files(nc, hm, &path);
+        } else if (has_prefix(&path, &((struct mg_str) MG_MK_STR("/tag/")))) {
+            if (WebCtx.tag_auth_enabled == TRUE) {
+                if (!validate_auth(nc, hm)) {
+                    return;
+                }
+            }
+            tag(nc, hm, &path);
         } else if (has_prefix(&path, &((struct mg_str) MG_MK_STR("/d/")))) {
             document_info(nc, hm, &path);
         } else {