Fix for #28

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

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 .

src/ctx.h

@@ -28,7 +28,7 @@ struct {
 
     pthread_mutex_t mupdf_mu;
     char * tesseract_lang;
-    char * tesseract_path;
+    const char * tesseract_path;
 } 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);
@@ -87,21 +89,15 @@ 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);
@@ -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);
@@ -136,14 +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 (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);
                 }
@@ -151,11 +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) {
@@ -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];
 

src/io/serialize.c

@@ -214,7 +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);
 
+        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);
 

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 = "1.2.9";
+static const char *const Version = "1.2.13";
 static const char *const usage[] = {
         "sist2 scan [OPTION]... PATH",
         "sist2 index [OPTION]... INDEX",

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/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);
+            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/types.h

@@ -49,7 +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];

src/web/serve.c

@@ -181,7 +181,12 @@ int chunked_response_file(const char *filename, const char *mime,
         }
     }
     onion_response_set_length(res, length);
+    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;

web/css/dark.css

@@ -423,3 +423,21 @@ 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);
 
@@ -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($("<td>").text(key))
@@ -379,6 +387,15 @@ function makeThumbnail(mimeCategory, hit, imgWrapper, small) {
         }
         thumbnail.setAttribute("src", `t/${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);
 

web/js/search.js

@@ -152,8 +152,8 @@ $.jsonPost("es", {
         target: '#mimeTree'
     });
     mimeTree.on("node.click", handleTreeClick(mimeTree));
-    mimeTree.select();
+    mimeTree.deselect();
-    mimeTree.node("any").deselect();
+    mimeTree.node("any").select();
 });
 
 function leafTag(tag) {
@@ -351,11 +351,10 @@ function search(after = null) {
         query: {
             bool: {
                 [condition]: {
-                    multi_match: {
+                    simple_query_string: {
                         query: query,
-                        type: "most_fields",
                         fields: fields,
-                        operator: "and"
+                        default_operator: "and"
                     }
                 },
                 filter: filters
@@ -406,6 +405,10 @@ function search(after = null) {
         let resultContainer = makeResultContainer();
         searchResults.appendChild(resultContainer);
 
+        window.setTimeout(() => {
+            $(".sp").SmartPhoto({animationSpeed: 0, swipeTopToClose: true, showAnimation: false, forceInterval: 50});
+        }, 100);
+
         if (!after) {
             docCount = 0;
         }
@@ -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">v1.2.9</span>
+    <span class="badge badge-pill version">v1.2.13</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>
@@ -31,7 +31,14 @@
                     </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>