--fast option #27

CMakeLists.txt

@@ -26,6 +26,7 @@ add_executable(
         src/parsing/arc.c src/parsing/arc.h
         src/parsing/doc.c src/parsing/doc.h
         src/log.c src/log.h
+        src/parsing/cbr.h src/parsing/cbr.c
 
         # argparse
         argparse/argparse.h argparse/argparse.c
@@ -136,6 +137,8 @@ TARGET_LINK_LIBRARIES(
         ${PROJECT_SOURCE_DIR}/lib/libcrypto.a
         ${PROJECT_SOURCE_DIR}/lib/libssl.a
         dl
+
+        pcre
 )
 
 add_custom_target(

Docker/build.sh

@@ -5,8 +5,11 @@ strip sist2
 version=$(./sist2 --version)
 
 echo "Version ${version}"
-docker build . -t simon987/sist2:${version} -t simon987/sist2:latest
+docker build . -t simon987/sist2:${version} -t simon987/sist2:latest \
+  -t docker.pkg.github.com/simon987/sist2/sist2:latest -t docker.pkg.github.com/simon987/sist2/sist2:${version}
 docker push simon987/sist2:${version}
 docker push simon987/sist2:latest
+docker push docker.pkg.github.com/simon987/sist2/sist2:latest
+docker push docker.pkg.github.com/simon987/sist2/sist2:${version}
 
 docker run --rm -it simon987/sist2 -v

README.md

@@ -92,10 +92,10 @@ docker stop sist2
 
 File type | Library | Content | Thumbnail | Metadata
 :---|:---|:---|:---|:---
-pdf,xps,cbz,fb2,epub | MuPDF | text+ocr | yes, `png` | title |
+pdf,xps,cbz,cbr,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` |
+`image/*` | ffmpeg | - | yes, `jpeg` | [Common EXIF tags](https://github.com/simon987/sist2/blob/efdde2734eca9b14a54f84568863b7ffd59bdba3/src/parsing/media.c#L190) |
 ttf,ttc,cff,woff,fnt,otf | Freetype2 | - | yes, `bmp` | Name & style |
 `text/plain` | *(none)* | yes | no | - |
 tar, zip, rar, 7z, ar ...  | Libarchive | yes\* | - | no |
@@ -120,7 +120,7 @@ To check if a media file can be parsed without *seek*, execute `cat file.mp4 | f
  
 ### OCR
 
-You can enable OCR support for pdf,xps,cbz,fb2,epub file types with the
+You can enable OCR support for pdf,xps,cbz,cbr,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).
 

mime.csv

@@ -320,7 +320,7 @@ video/x-dv, dif|dv
 video/x-fli, fli
 video/x-isvideo, isu
 video/x-motion-jpeg, mjpg
-video/x-ms-asf, asf|asx
+video/x-ms-asf, asf|asx|wmv
 video/x-qtc, qtc
 video/x-sgi-movie, movie|mv
 application/x-7z-compressed, 7z

schema/mappings.json

@@ -1,5 +1,9 @@
 {
   "properties": {
+    "_tie": {
+      "type": "keyword",
+      "doc_values": true
+    },
     "path": {
       "type": "text",
       "analyzer": "path_analyzer",
@@ -105,6 +109,30 @@
     },
     "tag": {
       "type": "keyword"
+    },
+    "exif_make": {
+      "type": "text"
+    },
+    "exif_model": {
+      "type": "text"
+    },
+    "exif:software": {
+      "type": "text"
+    },
+    "exif_exposure_time": {
+      "type": "keyword"
+    },
+    "exif_fnumber": {
+      "type": "keyword"
+    },
+    "exif_iso_speed_ratings": {
+      "type": "keyword"
+    },
+    "exif_focal_length": {
+      "type": "keyword"
+    },
+    "exif_user_comment": {
+      "type": "text"
     }
   }
 }

schema/pipeline.json

@@ -0,0 +1,10 @@
+{
+  "description": "Copy _id to _tie",
+  "processors": [
+    {
+      "script": {
+        "source": "ctx._tie = ctx._id;"
+      }
+    }
+  ]
+}

scripts/get_static_libs.sh

@@ -13,7 +13,7 @@ mv mupdf/build/release/libmupdf-third.a .
 
 # openjp2
 cd openjpeg
-cmake . -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS="-O3 -march=native -DNDEBUG -fPIC"
+cmake . -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_FLAGS="-O3 -DNDEBUG -fPIC"
 make -j $THREADS
 cd ..
 mv openjpeg/bin/libopenjp2.a .
@@ -75,7 +75,7 @@ cd tesseract
 mkdir build
 cd build
 cmake -DSTATIC=on -DBUILD_TRAINING_TOOLS=off -DBUILD_TESTS=off -DCMAKE_BUILD_TYPE=Release \
-  -DCMAKE_CXX_FLAGS="-fPIC"  ..
+  -DCMAKE_CXX_FLAGS="-fPIC" -DAUTO_OPTIMIZE=off ..
 make -j $THREADS
 cd ../..
 mv tesseract/build/libtesseract.a .

scripts/index_static.py

@@ -1,6 +1,9 @@
+import json
+
 files = [
     "schema/mappings.json",
     "schema/settings.json",
+    "schema/pipeline.json",
 ]
 
 
@@ -9,6 +12,6 @@ def clean(filepath):
 
 
 for file in files:
-    with open(file, "rb") as f:
-        data = f.read()
+    with open(file, "r") as f:
+        data = json.dumps(json.load(f), separators=(",", ":")).encode()
     print("char %s[%d] = {%s};" % (clean(file), len(data), ",".join(str(int(b)) for b in data)))

src/cli.c

@@ -162,6 +162,26 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
         args->tesseract_path = path;
     }
 
+    if (args->exclude_regex != NULL) {
+        const char *error;
+        int error_offset;
+
+        pcre *re = pcre_compile(args->exclude_regex, 0, &error, &error_offset, 0);
+        if (error != NULL) {
+            LOG_FATALF("cli.c", "pcre_compile returned error: %s (offset:%d)", error, error_offset)
+        }
+
+        pcre_extra *re_extra = pcre_study(re, 0, &error);
+        if (error != NULL) {
+            LOG_FATALF("cli.c", "pcre_study returned error: %s", error)
+        }
+
+        ScanCtx.exclude = re;
+        ScanCtx.exclude_extra = re_extra;
+    } else {
+        ScanCtx.exclude = NULL;
+    }
+
     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)
@@ -175,6 +195,8 @@ int scan_args_validate(scan_args_t *args, int argc, const char **argv) {
     LOG_DEBUGF("cli.c", "arg archive=%s", args->archive)
     LOG_DEBUGF("cli.c", "arg tesseract_lang=%s", args->tesseract_lang)
     LOG_DEBUGF("cli.c", "arg tesseract_path=%s", args->tesseract_path)
+    LOG_DEBUGF("cli.c", "arg exclude=%s", args->exclude_regex)
+    LOG_DEBUGF("cli.c", "arg fast=%d", args->fast)
 
     return 0;
 }

src/cli.h

@@ -18,10 +18,14 @@ typedef struct scan_args {
     archive_mode_t archive_mode;
     char *tesseract_lang;
     const char *tesseract_path;
+    char *exclude_regex;
+    int fast;
 } 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);
 
 typedef struct index_args {
@@ -45,12 +49,15 @@ typedef struct web_args {
 } web_args_t;
 
 index_args_t *index_args_create();
+
 void index_args_destroy(index_args_t *args);
 
 web_args_t *web_args_create();
+
 void web_args_destroy(web_args_t *args);
 
 int index_args_validate(index_args_t *args, int argc, const char **argv);
+
 int web_args_validate(web_args_t *args, int argc, const char **argv);
 
 #endif

src/ctx.h

@@ -28,7 +28,10 @@ struct {
 
     pthread_mutex_t mupdf_mu;
     char * tesseract_lang;
-    char * tesseract_path;
+    const char * tesseract_path;
+    pcre *exclude;
+    pcre_extra *exclude_extra;
+    int fast;
 } ScanCtx;
 
 struct {

src/index/elastic.c

@@ -20,6 +20,8 @@ typedef struct es_indexer {
 
 static es_indexer_t *Indexer;
 
+void delete_queue(int max);
+
 void print_json(cJSON *document, const char uuid_str[UUID_STR_LEN]) {
 
     cJSON *line = cJSON_CreateObject();
@@ -64,7 +66,7 @@ void execute_update_script(const char *script, const char index_id[UUID_STR_LEN]
     cJSON *term_obj = cJSON_AddObjectToObject(query, "term");
     cJSON_AddStringToObject(term_obj, "index", index_id);
 
-    char * str = cJSON_Print(body);
+    char *str = cJSON_Print(body);
 
     char bulk_url[4096];
     snprintf(bulk_url, 4096, "%s/sist2/_update_by_query?pretty", Indexer->es_url);
@@ -87,24 +89,18 @@ void execute_update_script(const char *script, const char index_id[UUID_STR_LEN]
     cJSON_Delete(resp);
 }
 
-void elastic_flush() {
-
-    if (Indexer == NULL) {
-        Indexer = create_indexer(IndexCtx.es_url);
-    }
-
+void *create_bulk_buffer(int max, int *count, size_t *buf_len) {
     es_bulk_line_t *line = Indexer->line_head;
-
-    int count = 0;
+    *count = 0;
 
     size_t buf_size = 0;
     size_t buf_cur = 0;
     char *buf = malloc(1);
 
-    while (line != NULL) {
+    while (line != NULL && *count < max) {
         char action_str[512];
         snprintf(action_str, 512,
-                "{\"index\":{\"_id\":\"%s\", \"_type\":\"_doc\", \"_index\":\"sist2\"}}\n", line->uuid_str);
+                 "{\"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);
@@ -116,46 +112,101 @@ void elastic_flush() {
         memcpy(buf + buf_cur, line->line, line_len);
         buf_cur += line_len;
 
-        es_bulk_line_t *tmp = line;
         line = line->next;
-        free(tmp);
-        count++;
+        (*count)++;
     }
     buf = realloc(buf, buf_size + 1);
-    *(buf+buf_cur) = '\0';
+    *(buf + buf_cur) = '\0';
 
-    Indexer->line_head = NULL;
-    Indexer->line_tail = NULL;
-    Indexer->queued = 0;
+    *buf_len = buf_cur;
+    return buf;
+}
+
+void _elastic_flush(int max) {
+    size_t buf_len;
+    int count;
+    void *buf = create_bulk_buffer(max, &count, &buf_len);
 
     char bulk_url[4096];
-    snprintf(bulk_url, 4096, "%s/sist2/_bulk", Indexer->es_url);
+    snprintf(bulk_url, 4096, "%s/sist2/_bulk?pipeline=tie", Indexer->es_url);
     response_t *r = web_post(bulk_url, buf, "Content-Type: application/x-ndjson");
 
     if (r->status_code == 0) {
         LOG_FATALF("elastic.c", "Could not connect to %s, make sure that elasticsearch is running!\n", IndexCtx.es_url)
     }
 
-    LOG_INFOF("elastic.c", "Indexed %d documents (%zukB) <%d>", count, buf_cur / 1024, r->status_code);
+    if (r->status_code == 413) {
 
-    cJSON *ret_json = cJSON_Parse(r->body);
-    if (cJSON_GetObjectItem(ret_json, "errors")->valueint != 0) {
-        cJSON *err;
-        cJSON_ArrayForEach(err, cJSON_GetObjectItem(ret_json, "items")) {
-            if (cJSON_GetObjectItem(cJSON_GetObjectItem(err, "index"), "status")->valueint != 201) {
-                char* str = cJSON_Print(err);
-                LOG_ERRORF("elastic.c", "%s\n", str);
-                cJSON_free(str);
+        if (max <= 1) {
+            LOG_ERRORF("elastic.c", "Single document too large, giving up: {%s}", Indexer->line_head->uuid_str)
+            free_response(r);
+            free(buf);
+            delete_queue(1);
+            if (Indexer->queued != 0) {
+                elastic_flush();
             }
+            return;
+        }
+
+        LOG_WARNINGF("elastic.c", "Payload too large, retrying (%d documents)", count);
+
+        free_response(r);
+        free(buf);
+        _elastic_flush(max / 2);
+        return;
+
+    } else if (r->status_code != 200) {
+        cJSON *ret_json = cJSON_Parse(r->body);
+        if (cJSON_GetObjectItem(ret_json, "errors")->valueint != 0) {
+            cJSON *err;
+            cJSON_ArrayForEach(err, cJSON_GetObjectItem(ret_json, "items")) {
+                if (cJSON_GetObjectItem(cJSON_GetObjectItem(err, "index"), "status")->valueint != 201) {
+                    char *str = cJSON_Print(err);
+                    LOG_ERRORF("elastic.c", "%s\n", str);
+                    cJSON_free(str);
+                }
+            }
+        }
+
+        cJSON_Delete(ret_json);
+        delete_queue(Indexer->queued);
+
+    } else {
+        LOG_INFOF("elastic.c", "Indexed %d documents (%zukB) <%d>", count, buf_len / 1024, r->status_code);
+
+        delete_queue(max);
+
+        if (Indexer->queued != 0) {
+            elastic_flush();
         }
     }
 
-    cJSON_Delete(ret_json);
-
     free_response(r);
     free(buf);
 }
 
+void delete_queue(int max) {
+    for (int i = 0; i < max; i++) {
+        es_bulk_line_t *tmp = Indexer->line_head;
+        Indexer->line_head = tmp->next;
+        if (Indexer->line_head == NULL) {
+            Indexer->line_tail = NULL;
+        } else {
+            free(tmp);
+        }
+        Indexer->queued -= 1;
+    }
+}
+
+void elastic_flush() {
+
+    if (Indexer == NULL) {
+        Indexer = create_indexer(IndexCtx.es_url);
+    }
+
+    _elastic_flush(Indexer->queued);
+}
+
 void elastic_index_line(es_bulk_line_t *line) {
 
     if (Indexer == NULL) {
@@ -192,7 +243,7 @@ es_indexer_t *create_indexer(const char *url) {
     return indexer;
 }
 
-void destroy_indexer(char * script, char index_id[UUID_STR_LEN]) {
+void destroy_indexer(char *script, char index_id[UUID_STR_LEN]) {
 
     char url[4096];
 
@@ -245,6 +296,11 @@ void elastic_init(int force_reset) {
         LOG_INFOF("elastic.c", "Close index <%d>", r->status_code);
         free_response(r);
 
+        snprintf(url, 4096, "%s/_ingest/pipeline/tie", IndexCtx.es_url);
+        r = web_put(url, pipeline_json, "Content-Type: application/json");
+        LOG_INFOF("elastic.c", "Create pipeline <%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");
         LOG_INFOF("elastic.c", "Update settings <%d>", r->status_code);
@@ -278,7 +334,7 @@ cJSON *elastic_get_document(const char *uuid_str) {
 char *elastic_get_status() {
     char url[4096];
     snprintf(url, 4096,
-            "%s/_cluster/state/metadata/sist2?filter_path=metadata.indices.*.state", WebCtx.es_url);
+             "%s/_cluster/state/metadata/sist2?filter_path=metadata.indices.*.state", WebCtx.es_url);
 
     response_t *r = web_get(url);
     cJSON *json = NULL;

src/io/serialize.c

@@ -118,6 +118,24 @@ char *get_meta_key_text(enum metakey meta_key) {
             return "font_name";
         case MetaParent:
             return "parent";
+        case MetaExifMake:
+            return "exif_make";
+        case MetaExifSoftware:
+            return "exif_software";
+        case MetaExifExposureTime:
+            return "exif_exposure_time";
+        case MetaExifFNumber:
+            return "exif_fnumber";
+        case MetaExifFocalLength:
+            return "exif_focal_length";
+        case MetaExifUserComment:
+            return "exif_user_comment";
+        case MetaExifIsoSpeedRatings:
+            return "exif_iso_speed_ratings";
+        case MetaExifModel:
+            return "exif_model";
+        case MetaExifDateTime:
+            return "exif_datetime";
         default:
             return NULL;
     }
@@ -196,7 +214,12 @@ 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);
 
-        cJSON_AddStringToObject(document, "mime", 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 {
+            cJSON_AddStringToObject(document, "mime", mime_get_mime_text(line.mime));
+        }
         cJSON_AddNumberToObject(document, "size", (double) line.size);
         cJSON_AddNumberToObject(document, "mtime", line.mtime);
 
@@ -255,6 +278,15 @@ void read_index_bin(const char *path, const char *index_id, index_func func) {
                 case MetaGenre:
                 case MetaFontName:
                 case MetaParent:
+                case MetaExifMake:
+                case MetaExifSoftware:
+                case MetaExifExposureTime:
+                case MetaExifFNumber:
+                case MetaExifFocalLength:
+                case MetaExifUserComment:
+                case MetaExifIsoSpeedRatings:
+                case MetaExifDateTime:
+                case MetaExifModel:
                 case MetaTitle: {
                     buf.cur = 0;
                     while ((c = getc(file)) != 0) {

src/io/walk.c

@@ -28,8 +28,18 @@ parse_job_t *create_fs_parse_job(const char *filepath, const struct stat *info,
     return job;
 }
 
+int sub_strings[30];
+#define EXCLUDED(str) (pcre_exec(ScanCtx.exclude, ScanCtx.exclude_extra, filepath, strlen(filepath), 0, 0, sub_strings, sizeof(sub_strings)) >= 0)
+
 int handle_entry(const char *filepath, const struct stat *info, int typeflag, struct FTW *ftw) {
-    if (ftw->level <= ScanCtx.depth && typeflag == FTW_F && S_ISREG(info->st_mode)) {
+
+    if (typeflag == FTW_F && S_ISREG(info->st_mode) && ftw->level <= ScanCtx.depth) {
+
+        if (ScanCtx.exclude != NULL && EXCLUDED(filepath)) {
+            LOG_DEBUGF("walk.c", "Excluded: %s", filepath)
+            return 0;
+        }
+
         parse_job_t *job = create_fs_parse_job(filepath, info, ftw->base);
         tpool_add_work(ScanCtx.pool, parse, job);
     }

src/main.c

@@ -6,7 +6,7 @@
 #define EPILOG "Made by simon987 <me@simon987.net>. Released under GPL-3.0"
 
 
-static const char *const Version = "1.2.4";
+static const char *const Version = "1.2.15";
 static const char *const usage[] = {
         "sist2 scan [OPTION]... PATH",
         "sist2 index [OPTION]... INDEX",
@@ -49,15 +49,19 @@ void sist2_scan(scan_args_t *args) {
     strncpy(ScanCtx.index.path, args->output, sizeof(ScanCtx.index.path));
     strncpy(ScanCtx.index.desc.name, args->name, sizeof(ScanCtx.index.desc.name));
     strncpy(ScanCtx.index.desc.root, args->path, sizeof(ScanCtx.index.desc.root));
+    strncpy(ScanCtx.index.desc.rewrite_url, args->rewrite_url, sizeof(ScanCtx.index.desc.rewrite_url));
     ScanCtx.index.desc.root_len = (short) strlen(ScanCtx.index.desc.root);
     ScanCtx.tesseract_lang = args->tesseract_lang;
     ScanCtx.tesseract_path = args->tesseract_path;
+    ScanCtx.fast = args->fast;
 
     init_dir(ScanCtx.index.path);
 
     ScanCtx.mime_table = mime_get_mime_table();
     ScanCtx.ext_table = mime_get_ext_table();
 
+    cbr_init();
+
     char store_path[PATH_MAX];
     snprintf(store_path, PATH_MAX, "%sthumbs", ScanCtx.index.path);
     mkdir(store_path, S_IWUSR | S_IRUSR | S_IXUSR);
@@ -237,6 +241,8 @@ int main(int argc, const char *argv[]) {
                                                           "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)"),
+            OPT_STRING('e', "exclude", &scan_args->exclude_regex, "Files that match this regex will not be scanned"),
+            OPT_BOOLEAN(0, "fast", &scan_args->fast, "Only index file names & mime type"),
 
             OPT_GROUP("Index options"),
             OPT_STRING(0, "es-url", &common_es_url, "Elasticsearch url with port. DEFAULT=http://localhost:9200"),
@@ -283,9 +289,7 @@ int main(int argc, const char *argv[]) {
         }
         sist2_scan(scan_args);
 
-    }
-
-    else if (strcmp(argv[0], "index") == 0) {
+    } else if (strcmp(argv[0], "index") == 0) {
 
         int err = index_args_validate(index_args, argc, argv);
         if (err != 0) {
@@ -301,8 +305,7 @@ int main(int argc, const char *argv[]) {
         }
         sist2_web(web_args);
 
-    }
-    else {
+    } else {
         fprintf(stderr, "Invalid command: '%s'\n", argv[0]);
         argparse_usage(&argparse);
         return 1;

src/parsing/arc.c

@@ -1,8 +1,6 @@
 #include "arc.h"
 #include "src/ctx.h"
 
-#define ARC_BUF_SIZE 8192
-
 int should_parse_filtered_file(const char *filepath, int ext) {
     char tmp[PATH_MAX * 2];
 

src/parsing/arc.h

@@ -2,6 +2,7 @@
 #define SIST2_ARC_H
 
 #include "src/sist.h"
+#define ARC_BUF_SIZE 8192
 
 int should_parse_filtered_file(const char *filepath, int ext);
 

src/parsing/cbr.c

@@ -0,0 +1,52 @@
+#import "cbr.h"
+#import "src/ctx.h"
+
+unsigned int cbr_mime;
+unsigned int cbz_mime;
+
+void cbr_init() {
+    cbr_mime = mime_get_mime_by_string(ScanCtx.mime_table, "application/x-cbr");
+    cbz_mime = mime_get_mime_by_string(ScanCtx.mime_table, "application/x-cbz");
+}
+
+int is_cbr(unsigned int mime) {
+    return mime == cbr_mime;
+}
+
+void parse_cbr(void *buf, size_t buf_len, document_t *doc) {
+    char *out_buf = malloc(buf_len * 2);
+    size_t out_buf_used = 0;
+
+    struct archive *rar_in = archive_read_new();
+    archive_read_support_filter_none(rar_in);
+    archive_read_support_format_rar(rar_in);
+
+    archive_read_open_memory(rar_in, buf, buf_len);
+
+    struct archive *zip_out = archive_write_new();
+    archive_write_set_format_zip(zip_out);
+    archive_write_open_memory(zip_out, out_buf, buf_len * 2, &out_buf_used);
+
+    struct archive_entry *entry;
+    while (archive_read_next_header(rar_in, &entry) == ARCHIVE_OK) {
+        archive_write_header(zip_out, entry);
+
+        char arc_buf[ARC_BUF_SIZE];
+        int len = archive_read_data(rar_in, arc_buf, ARC_BUF_SIZE);
+        while (len > 0) {
+            archive_write_data(zip_out, arc_buf, len);
+            len = archive_read_data(rar_in, arc_buf, ARC_BUF_SIZE);
+        }
+    }
+
+    archive_write_close(zip_out);
+    archive_write_free(zip_out);
+
+    archive_read_close(rar_in);
+    archive_read_free(rar_in);
+
+    doc->mime = cbz_mime;
+    parse_pdf(out_buf, out_buf_used, doc);
+    doc->mime = cbr_mime;
+    free(out_buf);
+}

src/parsing/cbr.h

@@ -0,0 +1,12 @@
+#ifndef SIST2_CBR_H
+#define SIST2_CBR_H
+
+#include "src/sist.h"
+
+void cbr_init();
+
+int is_cbr(unsigned int mime);
+
+void parse_cbr(void *buf, size_t buf_len, document_t *doc);
+
+#endif

src/parsing/doc.c

@@ -1,10 +1,20 @@
 #include "doc.h"
 #include "src/ctx.h"
 
-void dump_text(mceTextReader_t *reader, dyn_buffer_t *buf) {
+int dump_text(mceTextReader_t *reader, dyn_buffer_t *buf) {
 
     mce_skip_attributes(reader);
 
+    xmlErrorPtr err = xmlGetLastError();
+    if (err != NULL) {
+        if (err->level == XML_ERR_FATAL) {
+            LOG_ERRORF("doc.c", "Got fatal XML error while parsing document: %s", err->message)
+            return -1;
+        } else {
+            LOG_ERRORF("doc.c", "Got recoverable XML error while parsing document: %s", err->message)
+        }
+    }
+
     mce_start_children(reader) {
         mce_start_element(reader, NULL, _X("t")) {
             mce_skip_attributes(reader);
@@ -18,10 +28,14 @@ void dump_text(mceTextReader_t *reader, dyn_buffer_t *buf) {
         } mce_end_element(reader);
 
         mce_start_element(reader, NULL, NULL) {
-            dump_text(reader, buf);
+            int ret = dump_text(reader, buf);
+            if (ret != 0) {
+                return ret;
+            }
         } mce_end_element(reader);
 
     } mce_end_children(reader)
+    return 0;
 }
 
 __always_inline
@@ -52,30 +66,28 @@ int should_read_part(opcPart part) {
 }
 
 __always_inline
-void read_part(opcContainer *c, dyn_buffer_t *buf, opcPart part, document_t *doc) {
+int read_part(opcContainer *c, dyn_buffer_t *buf, opcPart part, document_t *doc) {
 
     mceTextReader_t reader;
-    int options;
-    if (LogCtx.very_verbose) {
-        options = XML_PARSE_NONET;
-    } else {
-        options = XML_PARSE_NOWARNING | XML_PARSE_NOERROR | XML_PARSE_NONET;
-    }
-
-    int ret = opcXmlReaderOpen(c, &reader, part, NULL, "UTF-8", options);
+    int ret = opcXmlReaderOpen(c, &reader, part, NULL, "UTF-8", XML_PARSE_NOWARNING | XML_PARSE_NOERROR | XML_PARSE_NONET);
 
     if (ret != OPC_ERROR_NONE) {
         LOG_ERRORF(doc->filepath, "(doc.c) opcXmlReaderOpen() returned error code %d", ret);
-        return;
+        return -1;
     }
 
     mce_start_document(&reader) {
         mce_start_element(&reader, NULL, NULL) {
-                dump_text(&reader, buf);
+            ret = dump_text(&reader, buf);
+            if (ret != 0) {
+                mceTextReaderCleanup(&reader);
+                return -1;
+            }
         } mce_end_element(&reader);
     } mce_end_document(&reader);
 
     mceTextReaderCleanup(&reader);
+    return 0;
 }
 
 void parse_doc(void *mem, size_t mem_len, document_t *doc) {
@@ -95,7 +107,10 @@ void parse_doc(void *mem, size_t mem_len, document_t *doc) {
     opcPart part = opcPartGetFirst(c);
     do {
         if (should_read_part(part)) {
-            read_part(c, &buf, part, doc);
+            int ret = read_part(c, &buf, part, doc);
+            if (ret != 0) {
+                break;
+            }
         }
     } while ((part = opcPartGetNext(c, part)));
 

src/parsing/media.c

@@ -193,6 +193,24 @@ append_video_meta(AVFormatContext *pFormatCtx, AVFrame *frame, document_t *doc,
                 APPEND_TAG_META(doc, tag, MetaArtist)
             } else if (strcmp(tag->key, "ImageDescription") == 0) {
                 APPEND_TAG_META(doc, tag, MetaContent)
+            } else if (strcmp(tag->key, "Make") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifMake)
+            } else if (strcmp(tag->key, "Model") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifModel)
+            } else if (strcmp(tag->key, "Software") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifSoftware)
+            } else if (strcmp(tag->key, "FNumber") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifFNumber)
+            } else if (strcmp(tag->key, "FocalLength") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifFocalLength)
+            } else if (strcmp(tag->key, "UserComment") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifUserComment)
+            } else if (strcmp(tag->key, "ISOSpeedRatings") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifIsoSpeedRatings)
+            } else if (strcmp(tag->key, "ExposureTime") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifExposureTime)
+            } else if (strcmp(tag->key, "DateTime") == 0) {
+                APPEND_TAG_META(doc, tag, MetaExifDateTime)
             }
         }
     }

src/parsing/mime.h

@@ -8,7 +8,7 @@
 #define MIME_EMPTY 1
 
 #define DONT_PARSE 0x80000000
-#define SHOULD_PARSE(mime_id) (mime_id & DONT_PARSE) != DONT_PARSE && mime_id != 0
+#define SHOULD_PARSE(mime_id) (ScanCtx.fast == 0 && (mime_id & DONT_PARSE) != DONT_PARSE && mime_id != 0)
 
 #define PDF_MASK 0x40000000
 #define IS_PDF(mime_id) (mime_id & PDF_MASK) == PDF_MASK

src/parsing/mime_generated.c

@@ -1293,6 +1293,7 @@ g_hash_table_insert(ext_table, "isu", (gpointer)video_x_isvideo);
 g_hash_table_insert(ext_table, "mjpg", (gpointer)video_x_motion_jpeg);
 g_hash_table_insert(ext_table, "asf", (gpointer)video_x_ms_asf);
 g_hash_table_insert(ext_table, "asx", (gpointer)video_x_ms_asf);
+g_hash_table_insert(ext_table, "wmv", (gpointer)video_x_ms_asf);
 g_hash_table_insert(ext_table, "qtc", (gpointer)video_x_qtc);
 g_hash_table_insert(ext_table, "movie", (gpointer)video_x_sgi_movie);
 g_hash_table_insert(ext_table, "mv", (gpointer)video_x_sgi_movie);

src/parsing/parse.c

@@ -149,6 +149,13 @@ void parse(void *arg) {
         if (doc_buf != buf && doc_buf != NULL) {
             free(doc_buf);
         }
+    } else if (is_cbr(doc.mime)) {
+        void *cbr_buf = read_all(job, (char *) buf, bytes_read);
+        parse_cbr(cbr_buf, doc.size, &doc);
+
+        if (cbr_buf != buf && cbr_buf != NULL) {
+            free(cbr_buf);
+        }
     }
 
     //Parent meta

src/parsing/pdf.c

@@ -1,7 +1,8 @@
 #include "pdf.h"
 #include "src/ctx.h"
 
-#define MIN_OCR_SIZE 128
+#define MIN_OCR_SIZE 350
+#define MIN_OCR_LEN 10
 __thread text_buffer_t thread_buffer;
 
 
@@ -128,6 +129,7 @@ int read_stext_block(fz_stext_block *block, text_buffer_t *tex) {
     return 0;
 }
 
+#define IS_VALID_BPP(d) (d==1 || d==2 || d==4 || d==8 || d==16 || d==24 || d==32)
 
 void fill_image(fz_context *ctx, UNUSED(fz_device *dev),
                 fz_image *img, UNUSED(fz_matrix ctm), UNUSED(float alpha),
@@ -135,7 +137,7 @@ void fill_image(fz_context *ctx, UNUSED(fz_device *dev),
 
     int l2factor = 0;
 
-    if (img->w > MIN_OCR_SIZE && img->h > MIN_OCR_SIZE) {
+    if (img->w > MIN_OCR_SIZE && img->h > MIN_OCR_SIZE && IS_VALID_BPP(img->n)) {
 
         fz_pixmap *pix = img->get_pixmap(ctx, img, NULL, img->w, img->h, &l2factor);
 
@@ -148,12 +150,14 @@ void fill_image(fz_context *ctx, UNUSED(fz_device *dev),
 
             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
-            )
+            if (len >= MIN_OCR_LEN) {
+                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);

src/sist.h

@@ -35,6 +35,7 @@
 #include <libxml/xmlstring.h>
 #define BOOL int
 #include <tesseract/capi.h>
+#include <pcre.h>
 
 #include <onion/onion.h>
 #include <onion/handler.h>
@@ -59,6 +60,7 @@
 #include "parsing/font.h"
 #include "parsing/arc.h"
 #include "parsing/doc.h"
+#include "parsing/cbr.h"
 #include "cli.h"
 #include "log.h"
 #include "utf8.h/utf8.h"

src/types.h

@@ -2,9 +2,9 @@
 #define SIST2_TYPES_H
 
 
-#define META_INT_MASK 0xF0
-#define META_STR_MASK 0xE0
-#define META_LONG_MASK 0xD0
+#define META_INT_MASK 0x80
+#define META_STR_MASK 0x40
+#define META_LONG_MASK 0x20
 #define IS_META_INT(key) (key & META_INT_MASK) == META_INT_MASK
 #define IS_META_LONG(key) (key & META_LONG_MASK) == META_LONG_MASK
 #define IS_META_STR(meta) (meta->key & META_STR_MASK) == META_STR_MASK
@@ -31,6 +31,16 @@ enum metakey {
     MetaTitle = 12 | META_STR_MASK,
     MetaFontName = 13 | META_STR_MASK,
     MetaParent = 14 | META_STR_MASK,
+    MetaExifMake = 15 | META_STR_MASK,
+    MetaExifSoftware = 16 | META_STR_MASK,
+    MetaExifExposureTime = 17 | META_STR_MASK,
+    MetaExifFNumber = 18 | META_STR_MASK,
+    MetaExifFocalLength = 19 | META_STR_MASK,
+    MetaExifUserComment = 20 | META_STR_MASK,
+    MetaExifModel = 21 | META_STR_MASK,
+    MetaExifIsoSpeedRatings = 22 | META_STR_MASK,
+    MetaExifDateTime = 23 | META_STR_MASK,
+    //Note to self: this will break after 31 entries
 };
 
 #define INDEX_TYPE_BIN "binary"
@@ -39,7 +49,7 @@ enum metakey {
 
 typedef struct index_descriptor {
     char uuid[UUID_STR_LEN];
-    char version[6];
+    char version[64];
     long timestamp;
     char root[PATH_MAX];
     char rewrite_url[8196];
@@ -85,7 +95,6 @@ typedef int (*read_func_t)(struct vfile *, void *buf, size_t size);
 typedef void (*close_func_t)(struct vfile *);
 
 typedef struct vfile {
-
     union {
         int fd;
         struct archive *arc;

src/util.c

@@ -91,7 +91,11 @@ text_buffer_t text_buffer_create(int max_size) {
 }
 
 void text_buffer_terminate_string(text_buffer_t *buf) {
-    dyn_buffer_write_char(&buf->dyn_buffer, '\0');
+    if (buf->dyn_buffer.cur > 0 && *(buf->dyn_buffer.buf + buf->dyn_buffer.cur - 1) == ' ') {
+        *(buf->dyn_buffer.buf + buf->dyn_buffer.cur - 1) = '\0';
+    } else {
+        dyn_buffer_write_char(&buf->dyn_buffer, '\0');
+    }
 }
 
 __always_inline
@@ -172,8 +176,8 @@ int text_buffer_append_string0(text_buffer_t *buf, char *str) {
 
 int text_buffer_append_char(text_buffer_t *buf, int c) {
 
-    if (SHOULD_IGNORE_CHAR(c)) {
-        if (!buf->last_char_was_whitespace) {
+    if (SHOULD_IGNORE_CHAR(c) || c == ' ') {
+        if (!buf->last_char_was_whitespace && buf->dyn_buffer.cur != 0) {
             dyn_buffer_write_char(&buf->dyn_buffer, ' ');
             buf->last_char_was_whitespace = TRUE;
 

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 >= '0' && c <= '9') || (c >= 'A' && c <= 'z') || (c > 127))
+#define SHOULD_KEEP_CHAR(c) ((c >= '\'' && c <= ';') || (c >= 'A' && c <= 'z') || (c > 127))
 
 
 typedef struct dyn_buffer {

src/web/serve.c

@@ -181,7 +181,12 @@ int chunked_response_file(const char *filename, const char *mime,
         }
     }
     onion_response_set_length(res, length);
-    onion_response_set_header(res, "Content-Type", mime);
+    if (mime != NULL) {
+        onion_response_set_header(res, "Content-Type", mime);
+    } else {
+        onion_response_set_header(res, "Content-Type", "application/octet-stream");
+    }
+
     onion_response_write_headers(res);
     if ((onion_request_get_flags(request) & OR_HEAD) == OR_HEAD) {
         length = 0;
@@ -221,14 +226,6 @@ int search(UNUSED(void *p), onion_request *req, onion_response *res) {
         return OCS_NOT_PROCESSED;
     }
 
-    char *scroll_param;
-    const char *scroll = onion_request_get_query(req, "scroll");
-    if (scroll != NULL) {
-        scroll_param = "?scroll=3m";
-    } else {
-        scroll_param = "";
-    }
-
     const struct onion_block_t *block = onion_request_get_data(req);
 
     if (block == NULL) {
@@ -236,7 +233,7 @@ int search(UNUSED(void *p), onion_request *req, onion_response *res) {
     }
 
     char url[4096];
-    snprintf(url, 4096, "%s/sist2/_search%s", WebCtx.es_url, scroll_param);
+    snprintf(url, 4096, "%s/sist2/_search", WebCtx.es_url);
     response_t *r = web_post(url, onion_block_data(block), "Content-Type: application/json");
 
     set_default_headers(res);
@@ -254,43 +251,6 @@ int search(UNUSED(void *p), onion_request *req, onion_response *res) {
     return OCS_PROCESSED;
 }
 
-int scroll(UNUSED(void *p), onion_request *req, onion_response *res) {
-
-    int flags = onion_request_get_flags(req);
-    if ((flags & OR_METHODS) != OR_GET) {
-        return OCS_NOT_PROCESSED;
-    }
-
-    char url[4096];
-    snprintf(url, 4096, "%s/_search/scroll", WebCtx.es_url);
-
-    const char *scroll_id = onion_request_get_query(req, "scroll_id");
-
-    cJSON *json = cJSON_CreateObject();
-    cJSON_AddStringToObject(json, "scroll_id", scroll_id);
-    cJSON_AddStringToObject(json, "scroll", "3m");
-
-    char *json_str = cJSON_PrintUnformatted(json);
-    response_t *r = web_post(url, json_str, "Content-Type: application/json");
-
-    cJSON_Delete(json);
-    cJSON_free(json_str);
-
-    if (r->status_code != 200) {
-        free_response(r);
-        return OCS_NOT_PROCESSED;
-    }
-
-    set_default_headers(res);
-    onion_response_set_header(res, "Content-Type", "application/json");
-    onion_response_set_header(res, "Content-Disposition", "application/json");
-    onion_response_set_length(res, r->size);
-    onion_response_write(res, r->body, r->size);
-    free_response(r);
-
-    return OCS_PROCESSED;
-}
-
 int serve_file_from_url(cJSON *json, index_t *idx, onion_request *req, onion_response *res) {
 
     const char *path = cJSON_GetObjectItem(json, "path")->valuestring;
@@ -466,7 +426,6 @@ void serve(const char *hostname, const char *port) {
     onion_url_add(urls, "img/sprite-skin-flat.png", img_sprite_skin_flag);
 
     onion_url_add(urls, "es", search);
-    onion_url_add(urls, "scroll", scroll);
     onion_url_add(urls, "status", status);
     onion_url_add(
             urls,

web/css/dark.css

@@ -102,6 +102,7 @@ body {
     border-bottom: none;
     border-left: none;
     border-right: none;
+    padding: .25rem 0.5rem;
 }
 
 .list-group-item:first-child {
@@ -199,7 +200,7 @@ body {
     max-width: 100%;
     max-height: 175px;
     margin: 0 auto 0;
-    padding: 3px 3px 0 3px;
+    padding: 3px 3px 0;
     width: auto;
     height: auto;
 }
@@ -208,7 +209,7 @@ body {
     display: block;
     max-width: 64px;
     max-height: 64px;
-    margin: 0 auto 0;
+    margin: 0 auto;
     width: auto;
     height: auto;
 }
@@ -391,10 +392,6 @@ option {
     margin-top: 1em;
 }
 
-.list-group-item {
-    padding: .25rem 0.5rem;
-}
-
 .wrapper-sm {
     min-width: 64px;
 }
@@ -423,3 +420,21 @@ option {
     margin-top: -14px;
     font-size: 11px;
 }
+
+@media (min-width: 800px) {
+    .small-btn {
+        display: none;
+    }
+    .large-btn {
+        display: inherit;
+    }
+}
+
+@media (max-width: 801px) {
+    .small-btn {
+        display: inherit;
+    }
+    .large-btn {
+        display: none;
+    }
+}

web/css/light.css

@@ -287,3 +287,21 @@ mark {
     margin-top: -14px;
     font-size: 11px;
 }
+
+@media (min-width: 800px) {
+    .small-btn {
+        display: none;
+    }
+    .large-btn {
+        display: inherit;
+    }
+}
+
+@media (max-width: 801px) {
+    .small-btn {
+        display: inherit;
+    }
+    .large-btn {
+        display: none;
+    }
+}

web/js/dom.js

@@ -75,6 +75,10 @@ function shouldPlayVideo(hit) {
     return videoc !== "hevc" && videoc !== "mpeg2video" && videoc !== "wmv3";
 }
 
+function shouldDisplayRawImage(hit) {
+    return hit["_source"]["mime"] && hit["_source"]["mime"].startsWith("image/") && hit["_source"]["videoc"] !== "tiff";
+}
+
 function makePlaceholder(w, h, small) {
     let calc;
     if (small) {
@@ -96,10 +100,14 @@ function makePlaceholder(w, h, small) {
     return el;
 }
 
+function ext(hit) {
+    return hit["_source"].hasOwnProperty("extension") && hit["_source"]["extension"] !== "" ? "." + hit["_source"]["extension"] : "";
+}
+
 function makeTitle(hit) {
     let title = document.createElement("div");
     title.setAttribute("class", "file-title");
-    let extension = hit["_source"].hasOwnProperty("extension") && hit["_source"]["extension"] !== "" ? "." + hit["_source"]["extension"] : "";
+    let extension = ext(hit);
 
     applyNameToTitle(hit, title, extension);
 
@@ -113,7 +121,7 @@ function getTags(hit, mimeCategory) {
     switch (mimeCategory) {
         case "video":
         case "image":
-            if (hit["_source"].hasOwnProperty("videoc")) {
+            if (hit["_source"].hasOwnProperty("videoc") && hit["_source"]["videoc"]) {
                 const formatTag = document.createElement("span");
                 formatTag.setAttribute("class", "badge badge-pill badge-video");
                 formatTag.appendChild(document.createTextNode(hit["_source"]["videoc"].replace(" ", "")));
@@ -121,7 +129,7 @@ function getTags(hit, mimeCategory) {
             }
             break;
         case "audio": {
-            if (hit["_source"].hasOwnProperty("audioc")) {
+            if (hit["_source"].hasOwnProperty("audioc") && hit["_source"]["audioc"]) {
                 let formatTag = document.createElement("span");
                 formatTag.setAttribute("class", "badge badge-pill badge-audio");
                 formatTag.appendChild(document.createTextNode(hit["_source"]["audioc"]));
@@ -156,7 +164,7 @@ function getTags(hit, mimeCategory) {
 function infoButtonCb(hit) {
     return () => {
         getDocumentInfo(hit["_id"]).then(doc => {
-            $("#modal-title").text(doc["name"] + (doc["extension"] ? "." + doc["extension"] : ""));
+            $("#modal-title").text(doc["name"] + ext(hit));
 
             const tbody = $("<tbody>");
             $("#modal-body").empty()
@@ -175,7 +183,7 @@ function infoButtonCb(hit) {
                 "bitrate", "artist", "album", "album_artist", "genre", "title", "font_name", "tag"
             ]);
             Object.keys(doc)
-                .filter(key => key.startsWith("_keyword.") || key.startsWith("_text.") || displayFields.has(key))
+                .filter(key => key.startsWith("_keyword.") || key.startsWith("_text.") || displayFields.has(key) || key.startsWith("exif_"))
                 .forEach(key => {
                     tbody.append($("<tr>")
                         .append($("<td>").text(key))
@@ -369,6 +377,7 @@ function makeThumbnail(mimeCategory, hit, imgWrapper, small) {
         || hit["_source"]["mime"] === "application/pdf"
         || hit["_source"]["mime"] === "application/epub+zip"
         || hit["_source"]["mime"] === "application/x-cbz"
+        || hit["_source"]["mime"] === "application/x-cbr"
         || hit["_source"].hasOwnProperty("font_name")
     ) {
         thumbnail = document.createElement("img");
@@ -379,6 +388,15 @@ function makeThumbnail(mimeCategory, hit, imgWrapper, small) {
         }
         thumbnail.setAttribute("src", `t/${hit["_source"]["index"]}/${hit["_id"]}`);
 
+        if (!hit["_source"]["parent"] && shouldDisplayRawImage(hit)) {
+            imgWrapper.setAttribute("id", "sp" + hit["_id"]);
+            imgWrapper.setAttribute("data-src", `t/${hit["_source"]["index"]}/${hit["_id"]}`);
+            imgWrapper.setAttribute("href", `f/${hit["_id"]}`);
+            imgWrapper.setAttribute("data-caption", hit["_source"]["path"] + "/" + hit["_source"]["name"] + ext(hit));
+            imgWrapper.setAttribute("data-group", "p" + Math.floor(docCount / SIZE));
+            imgWrapper.classList.add("sp");
+        }
+
         const placeholder = makePlaceholder(hit["_source"]["width"], hit["_source"]["height"], small);
         imgWrapper.appendChild(placeholder);
 
@@ -499,8 +517,7 @@ function makePreloader() {
 function makePageIndicator(searchResult) {
     let pageIndicator = document.createElement("div");
     pageIndicator.setAttribute("class", "page-indicator font-weight-light");
-    const totalHits = searchResult["hits"]["total"].hasOwnProperty("value")
-        ? searchResult["hits"]["total"]["value"] : searchResult["hits"]["total"];
+    const totalHits = searchResult["aggregations"]["total_count"]["value"];
     pageIndicator.appendChild(document.createTextNode(docCount + " / " + totalHits));
     return pageIndicator;
 }
@@ -547,8 +564,7 @@ function makeStatsCard(searchResult) {
     });
 
     let stat = document.createElement("span");
-    const totalHits = searchResult["hits"]["total"].hasOwnProperty("value")
-        ? searchResult["hits"]["total"]["value"] : searchResult["hits"]["total"];
+    const totalHits = searchResult["aggregations"]["total_count"]["value"];
     stat.appendChild(document.createTextNode(totalHits + " results in " + searchResult["took"] + "ms"));
 
     statsCardBody.appendChild(stat);

web/js/search.js

@@ -6,7 +6,8 @@ let tagTree;
 
 let searchBar = document.getElementById("searchBar");
 let pathBar = document.getElementById("pathBar");
-let scroll_id = null;
+let lastDoc = null;
+let reachedEnd = false;
 let docCount = 0;
 let coolingDown = false;
 let searchBusy = true;
@@ -151,8 +152,8 @@ $.jsonPost("es", {
         target: '#mimeTree'
     });
     mimeTree.on("node.click", handleTreeClick(mimeTree));
-    mimeTree.select();
-    mimeTree.node("any").deselect();
+    mimeTree.deselect();
+    mimeTree.node("any").select();
 });
 
 function leafTag(tag) {
@@ -259,41 +260,18 @@ function insertHits(resultContainer, hits) {
 }
 
 window.addEventListener("scroll", function () {
-    if (!coolingDown && !searchBusy) {
+    if (!searchBusy) {
         let threshold = 400;
 
         if ((window.innerHeight + window.scrollY) >= document.body.offsetHeight - threshold) {
-            coolingDown = true;
-            doScroll();
+            if (!reachedEnd) {
+                coolingDown = true;
+                search(lastDoc);
+            }
         }
     }
 });
 
-function doScroll() {
-    $.get("scroll", {scroll_id: scroll_id})
-        .then(searchResult => {
-            let searchResults = document.getElementById("searchResults");
-            let hits = searchResult["hits"]["hits"];
-
-            //Page indicator
-            let pageIndicator = makePageIndicator(searchResult);
-            searchResults.appendChild(pageIndicator);
-
-            //Result container
-            let resultContainer = makeResultContainer();
-            searchResults.appendChild(resultContainer);
-
-            insertHits(resultContainer, hits);
-
-            if (hits.length === SIZE) {
-                coolingDown = false;
-            }
-        })
-        .fail(() => {
-            window.location.reload();
-        })
-}
-
 function getSelectedNodes(tree) {
     let selectedNodes = [];
 
@@ -314,21 +292,25 @@ function getSelectedNodes(tree) {
     return selectedNodes
 }
 
-function search() {
+function search(after = null) {
+    lastDoc = null;
+
     if (searchBusy) {
         return;
     }
     searchBusy = true;
 
-    //Clear old search results
     let searchResults = document.getElementById("searchResults");
-    while (searchResults.firstChild) {
-        searchResults.removeChild(searchResults.firstChild);
+    //Clear old search results
+    let preload;
+    if (!after) {
+        while (searchResults.firstChild) {
+            searchResults.removeChild(searchResults.firstChild);
+        }
+        preload = makePreloader();
+        searchResults.appendChild(preload);
     }
 
-    const preload = makePreloader();
-    searchResults.appendChild(preload);
-
     let query = searchBar.value;
     let empty = query === "";
     let condition = empty ? "should" : "must";
@@ -362,25 +344,25 @@ function search() {
         filters.push([{terms: {"tag": tags}}]);
     }
 
-    $.jsonPost("es?scroll=1", {
+    let q = {
         "_source": {
-            excludes: ["content"]
+            excludes: ["content", "_tie"]
         },
         query: {
             bool: {
                 [condition]: {
-                    multi_match: {
+                    simple_query_string: {
                         query: query,
-                        type: "most_fields",
                         fields: fields,
-                        operator: "and"
+                        default_operator: "and"
                     }
                 },
                 filter: filters
             }
         },
-        sort: [
-            "_score"
+        "sort": [
+            {"_score": {"order": "desc"}},
+            {"_tie": {"order": "asc"}}
         ],
         highlight: {
             pre_tags: ["<mark>"],
@@ -393,24 +375,45 @@ function search() {
                 font_name: {},
             }
         },
-        aggs: {
-            total_size: {"sum": {"field": "size"}}
-        },
+        aggs:
+            {
+                total_size: {"sum": {"field": "size"}},
+                total_count: {"value_count": {"field": "size"}}
+            },
         size: SIZE,
-    }).then(searchResult => {
-        scroll_id = searchResult["_scroll_id"];
+    };
 
-        preload.remove();
-        //Search stats
-        searchResults.appendChild(makeStatsCard(searchResult));
+    if (after) {
+        q.search_after = [after["_score"], after["_id"]];
+    }
+
+    $.jsonPost("es", q).then(searchResult => {
+        let hits = searchResult["hits"]["hits"];
+        if (hits) {
+            lastDoc = hits[hits.length - 1];
+        }
+
+        if (!after) {
+            preload.remove();
+            searchResults.appendChild(makeStatsCard(searchResult));
+        } else {
+            let pageIndicator = makePageIndicator(searchResult);
+            searchResults.appendChild(pageIndicator);
+        }
 
         //Setup page
         let resultContainer = makeResultContainer();
         searchResults.appendChild(resultContainer);
 
-        docCount = 0;
-        insertHits(resultContainer, searchResult["hits"]["hits"]);
+        window.setTimeout(() => {
+            $(".sp").SmartPhoto({animationSpeed: 0, swipeTopToClose: true, showAnimation: false, forceInterval: 50});
+        }, 100);
 
+        if (!after) {
+            docCount = 0;
+        }
+        reachedEnd = hits.length !== SIZE;
+        insertHits(resultContainer, hits);
         searchBusy = false;
     });
 }
@@ -471,6 +474,15 @@ function updateIndices() {
 document.getElementById("indices").addEventListener("change", updateIndices);
 updateIndices();
 
+window.onkeyup = function(e) {
+    if (e.key === "/" || e.key === "Escape") {
+        const bar = document.getElementById("searchBar");
+        bar.scrollIntoView();
+        bar.focus();
+    }
+    console.log(e)
+};
+
 //Suggest
 function getPathChoices() {
     return new Promise(getPaths => {

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">v1.2.4</span>
+    <span class="badge badge-pill version">v1.2.15</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>
@@ -25,13 +25,20 @@
             <div class="input-group">
                 <div class="input-group-prepend">
                     <div class="input-group-text">
-                        <span title="Toggle fuzzy searching"  onclick="document.getElementById('fuzzyToggle').click()">Fuzzy&nbsp</span>
+                        <span title="Toggle fuzzy searching" onclick="document.getElementById('fuzzyToggle').click()">Fuzzy&nbsp</span>
                         <input title="Toggle fuzzy searching" type="checkbox" id="fuzzyToggle"
                                onclick="toggleFuzzy()" checked>
                     </div>
                 </div>
                 <input id="searchBar" type="search" class="form-control" placeholder="Search">
-
+                <div class="input-group-append">
+                    <button class="btn btn-outline-secondary small-btn" type="button" data-toggle="modal"
+                            data-target="#help">?
+                    </button>
+                    <button class="btn btn-outline-secondary large-btn" type="button" data-toggle="modal"
+                            data-target="#help">Help
+                    </button>
+                </div>
             </div>
 
             <input title="File size" id="sizeSlider" name="size">
@@ -45,10 +52,12 @@
                 <div class="col" id="treeTabs">
                     <ul class="nav nav-tabs" role="tablist">
                         <li class="nav-item">
-                            <a class="nav-link active" data-toggle="tab" href="#mime" role="tab" aria-controls="home" aria-selected="true">Mime Types</a>
+                            <a class="nav-link active" data-toggle="tab" href="#mime" role="tab" aria-controls="home"
+                               aria-selected="true">Mime Types</a>
                         </li>
                         <li class="nav-item">
-                            <a class="nav-link" data-toggle="tab" href="#tag" role="tab" aria-controls="profile" aria-selected="false" title="User-defined tags">Tags</a>
+                            <a class="nav-link" data-toggle="tab" href="#tag" role="tab" aria-controls="profile"
+                               aria-selected="false" title="User-defined tags">Tags</a>
                         </li>
                     </ul>
                     <div class="tab-content" id="myTabContent">
@@ -79,6 +88,69 @@
         </div>
     </div>
 
+    <div class="modal" id="help" tabindex="-1" role="dialog" aria-labelledby="modal-title" aria-hidden="true">
+        <div class="modal-dialog modal-lg modal-dialog-centered" role="document">
+            <div class="modal-content">
+                <div class="modal-header">
+                    <h5 class="modal-title">Search help</h5>
+                    <button type="button" class="close" data-dismiss="modal" aria-label="Close">
+                        <span aria-hidden="true">&times;</span>
+                    </button>
+                </div>
+                <div class="modal-body">
+
+                    <table class="table">
+                        <tbody>
+                        <tr>
+                            <td><code>+</code></td>
+                            <td>signifies AND operation</td>
+                        </tr>
+                        <tr>
+                            <td><code>|</code></td>
+                            <td>signifies OR operation</td>
+                        </tr>
+                        <tr>
+                            <td><code>-</code></td>
+                            <td>negates a single token</td>
+                        </tr>
+                        <tr>
+                            <td><code>""</code></td>
+                            <td>wraps a number of tokens to signify a phrase for searching</td>
+                        </tr>
+                        <tr>
+                            <td><code>*</code></td>
+                            <td>at the end of a term signifies a prefix query</td>
+                        </tr>
+                        <tr>
+                            <td><code>(</code> and <code>)</code></td>
+                            <td>signify precedence</td>
+                        </tr>
+                        <tr>
+                            <td><code>~N</code></td>
+                            <td>after a word signifies edit distance (fuzziness)</td>
+                        </tr>
+                        <tr>
+                            <td><code>~N</code></td>
+                            <td>after a phrase signifies slop amount</td>
+                        </tr>
+                        </tbody>
+                    </table>
+
+                    <p>For example: <code>"fried eggs" +(eggplant | potato) -frittata</code> will match the phrase
+                        <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 the <b>Fuzzy</b> option is checked, partial matches are also returned.</p>
+                    <br>
+                    <p>For more information, see <a target="_blank"
+                                                    href="//www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-simple-query-string-query.html">Elasticsearch
+                        documentation</a></p>
+                </div>
+            </div>
+        </div>
+    </div>
+
     <div id="searchResults"></div>
 </div>