Fix memory leak with virtual files in parse.c

src/io/serialize.c

@@ -193,6 +193,8 @@ void write_document(document_t *doc) {
 
 void thread_cleanup() {
     close(index_fd);
+    cleanup_parse();
+    cleanup_font();
 }
 
 

src/parsing/media.c

@@ -113,10 +113,13 @@ static AVFrame *read_frame(AVFormatContext *pFormatCtx, AVCodecContext *decoder,
         // Feed it to decoder
         int decode_ret = avcodec_send_packet(decoder, &avPacket);
         if (decode_ret != 0) {
-            LOG_WARNINGF(doc->filepath,
+            LOG_ERRORF(doc->filepath,
                          "(media.c) avcodec_send_packet() returned error code [%d] %s",
                          decode_ret, av_err2str(decode_ret)
             )
+            av_frame_free(&frame);
+            av_packet_unref(&avPacket);
+            return NULL;
         }
         av_packet_unref(&avPacket);
         receive_ret = avcodec_receive_frame(decoder, frame);
@@ -139,7 +142,7 @@ static void append_audio_meta(AVFormatContext *pFormatCtx, document_t *doc) {
 
     AVDictionaryEntry *tag = NULL;
     while ((tag = av_dict_get(pFormatCtx->metadata, "", tag, AV_DICT_IGNORE_SUFFIX))) {
-        char key[32];
+        char key[256];
         strncpy(key, tag->key, sizeof(key));
 
         char *ptr = key;
@@ -160,7 +163,8 @@ static void append_audio_meta(AVFormatContext *pFormatCtx, document_t *doc) {
 }
 
 __always_inline
-static void append_video_meta(AVFormatContext *pFormatCtx, AVFrame *frame, document_t *doc, int include_audio_tags, int is_video) {
+static void
+append_video_meta(AVFormatContext *pFormatCtx, AVFrame *frame, document_t *doc, int include_audio_tags, int is_video) {
 
     if (is_video) {
         meta_line_t *meta_duration = malloc(sizeof(meta_line_t));

src/parsing/parse.c

@@ -37,7 +37,13 @@ void *read_all(parse_job_t *job, const char *buf, int bytes_read) {
 
         int ret = job->vfile.read(&job->vfile, full_buf + bytes_read, job->info.st_size - bytes_read);
         if (ret < 0) {
-            LOG_ERRORF(job->filepath, "read(): [%d] %s", errno, strerror(errno))
+            free(full_buf);
+
+            if (job->vfile.is_fs_file) {
+                LOG_ERRORF(job->filepath, "read(): [%d] %s", errno, strerror(errno))
+            } else {
+                LOG_ERRORF(job->filepath, "(virtual) read(): [%d] %s", ret, archive_error_string(job->vfile.arc))
+            }
             return NULL;
         }
     }
@@ -92,7 +98,13 @@ void parse(void *arg) {
         // Get mime type with libmagic
         bytes_read = job->vfile.read(&job->vfile, buf, PARSE_BUF_SIZE);
         if (bytes_read < 0) {
-            LOG_WARNINGF(job->filepath, "read() Error: %s", strerror(errno))
+
+            if (job->vfile.is_fs_file) {
+                LOG_ERRORF(job->filepath, "read(): [%d] %s", errno, strerror(errno))
+            } else {
+                LOG_ERRORF(job->filepath, "(virtual) read(): [%d] %s", bytes_read, archive_error_string(job->vfile.arc))
+            }
+
             CLOSE_FILE(job->vfile)
             return;
         }
@@ -180,3 +192,9 @@ void parse(void *arg) {
 
     CLOSE_FILE(job->vfile)
 }
+
+void cleanup_parse() {
+    if (Magic != NULL) {
+        magic_close(Magic);
+    }
+}

src/parsing/parse.h

@@ -10,4 +10,6 @@ void fs_close(struct vfile *f);
 
 void parse(void *arg);
 
+void cleanup_parse();
+
 #endif

src/parsing/pdf.c

@@ -168,7 +168,7 @@ void fill_image(fz_context *ctx, UNUSED(fz_device *dev),
     }
 }
 
-void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
+void parse_pdf(const void *buf, size_t buf_len, document_t *doc) {
 
     if (buf == NULL) {
         return;

src/parsing/pdf.h

@@ -4,6 +4,6 @@
 #include "src/sist.h"
 
 
-void parse_pdf(void *buf, size_t buf_len, document_t *doc);
+void parse_pdf(const void *buf, size_t buf_len, document_t *doc);
 
 #endif