--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(

README.md

@@ -92,7 +92,7 @@ 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` | [Common EXIF tags](https://github.com/simon987/sist2/blob/efdde2734eca9b14a54f84568863b7ffd59bdba3/src/parsing/media.c#L190) |
@@ -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).
 

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

@@ -29,6 +29,9 @@ struct {
     pthread_mutex_t mupdf_mu;
     char * tesseract_lang;
     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() {
+void *create_bulk_buffer(int max, int *count, size_t *buf_len) {
-
-    if (Indexer == NULL) {
-        Indexer = create_indexer(IndexCtx.es_url);
-    }
-
     es_bulk_line_t *line = Indexer->line_head;
-
+    *count = 0;
-    int 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);
@@ -116,17 +112,20 @@ 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;
+    *buf_len = buf_cur;
-    Indexer->line_tail = NULL;
+    return buf;
-    Indexer->queued = 0;
+}
+
+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?pipeline=tie", Indexer->es_url);
@@ -136,15 +135,33 @@ void elastic_flush() {
         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) {
 
-    if (r->status_code != 200 && r->status_code != 413) {
+        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);
+                    char *str = cJSON_Print(err);
                     LOG_ERRORF("elastic.c", "%s\n", str);
                     cJSON_free(str);
                 }
@@ -152,12 +169,44 @@ void elastic_flush() {
         }
 
         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();
+        }
     }
 
     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) {
@@ -194,7 +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];
 
@@ -285,7 +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/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 = "1.2.12";
+static const char *const Version = "1.2.15";
 static const char *const usage[] = {
         "sist2 scan [OPTION]... PATH",
         "sist2 index [OPTION]... INDEX",
@@ -53,12 +53,15 @@ void sist2_scan(scan_args_t *args) {
     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);
@@ -238,6 +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"),
@@ -284,9 +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) {
@@ -302,8 +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/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/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/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

@@ -95,7 +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 +91,7 @@ text_buffer_t text_buffer_create(int max_size) {
 }
 
 void text_buffer_terminate_string(text_buffer_t *buf) {
-    if (*(buf->dyn_buffer.buf + buf->dyn_buffer.cur - 1) == ' ') {
+    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');

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;
 }

web/js/dom.js

@@ -183,7 +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) || key.startsWith("exif_"))
+                .filter(key => key.startsWith("_keyword.") || key.startsWith("_text.") || displayFields.has(key) || key.startsWith("exif_"))
                 .forEach(key => {
                     tbody.append($("<tr>")
                         .append($("<td>").text(key))
@@ -377,6 +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");

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">v1.2.12</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>