gst/matroska/: Try to fix up broken matroska files containing subtitle streams with...

Original commit message from CVS:
* gst/matroska/matroska-demux.c:
(gst_matroska_demux_check_subtitle_buffer),
(gst_matroska_demux_parse_blockgroup_or_simpleblock),
(gst_matroska_demux_subtitle_caps):
* gst/matroska/matroska-ids.c:
(gst_matroska_track_init_subtitle_context):
* gst/matroska/matroska-ids.h:
Try to fix up broken matroska files containing subtitle
streams with non-UTF8 character encodings (courtesy of
mkvmerge) using either the encoding specified in the
GST_SUBTITLE_ENCODING environment variable or the
current locale's character set if it is non-UTF8.
Fixes #337076.

diff --git a/ChangeLog b/ChangeLog
index 2992f00..6580f5a 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,21 @@
 2006-06-22  Tim-Philipp Müller  <tim at centricular dot net>
 
+       * gst/matroska/matroska-demux.c:
+       (gst_matroska_demux_check_subtitle_buffer),
+       (gst_matroska_demux_parse_blockgroup_or_simpleblock),
+       (gst_matroska_demux_subtitle_caps):
+       * gst/matroska/matroska-ids.c:
+       (gst_matroska_track_init_subtitle_context):
+       * gst/matroska/matroska-ids.h:
+         Try to fix up broken matroska files containing subtitle
+         streams with non-UTF8 character encodings (courtesy of
+         mkvmerge) using either the encoding specified in the
+         GST_SUBTITLE_ENCODING environment variable or the
+         current locale's character set if it is non-UTF8.
+         Fixes #337076.
+
+2006-06-22  Tim-Philipp Müller  <tim at centricular dot net>
+
        * gst/id3demux/id3v2frames.c: (parse_picture_frame):
          Set image type from APIC frame as "image-type" field
          of GST_TAG_IMAGE buffer caps (#344605).

diff --git a/gst/matroska/matroska-demux.c b/gst/matroska/matroska-demux.c
index 00eae24..15df199 100644 (file)
--- a/gst/matroska/matroska-demux.c
+++ b/gst/matroska/matroska-demux.c
@@ -36,7 +36,7 @@
 #include "matroska-demux.h"
 #include "matroska-ids.h"
 
-GST_DEBUG_CATEGORY (matroskademux_debug);
+GST_DEBUG_CATEGORY_STATIC (matroskademux_debug);
 #define GST_CAT_DEFAULT matroskademux_debug
 
 enum
@@ -2135,6 +2135,75 @@ gst_matroska_demux_add_wvpk_header (GstMatroskaTrackContext * stream,
   return TRUE;
 }
 
+static GstBuffer *
+gst_matroska_demux_check_subtitle_buffer (GstMatroskaDemux * demux,
+    GstMatroskaTrackContext * stream, GstBuffer * buf)
+{
+  GstMatroskaTrackSubtitleContext *sub_stream;
+  const gchar *encoding, *data;
+  GError *err = NULL;
+  GstBuffer *newbuf;
+  gchar *utf8;
+  guint size;
+
+  sub_stream = (GstMatroskaTrackSubtitleContext *) stream;
+
+  if (!sub_stream->check_utf8)
+    return buf;
+
+  data = (const gchar *) GST_BUFFER_DATA (buf);
+  size = GST_BUFFER_SIZE (buf);
+
+  if (!sub_stream->invalid_utf8) {
+    if (g_utf8_validate (data, size, NULL)) {
+      return buf;
+    }
+    GST_WARNING_OBJECT (demux, "subtitle stream %d is not valid UTF-8, this "
+        "is broken according to the matroska specification", stream->num);
+    sub_stream->invalid_utf8 = TRUE;
+  }
+
+  /* file with broken non-UTF8 subtitle, do the best we can do to fix it */
+  encoding = g_getenv ("GST_SUBTITLE_ENCODING");
+  if (encoding == NULL || *encoding == '\0') {
+    /* if local encoding is UTF-8 and no encoding specified
+     * via the environment variable, assume ISO-8859-15 */
+    if (g_get_charset (&encoding)) {
+      encoding = "ISO-8859-15";
+    }
+  }
+
+  utf8 = g_convert_with_fallback (data, size, "UTF-8", encoding, "*",
+      NULL, NULL, &err);
+
+  if (err) {
+    GST_LOG_OBJECT (demux, "could not convert string from '%s' to UTF-8: %s",
+        encoding, err->message);
+    g_error_free (err);
+    g_free (utf8);
+
+    /* invalid input encoding, fall back to ISO-8859-15 (always succeeds) */
+    encoding = "ISO-8859-15";
+    utf8 = g_convert_with_fallback (data, size, "UTF-8", encoding, "*",
+        NULL, NULL, NULL);
+  }
+
+  GST_LOG_OBJECT (demux, "converted subtitle text from %s to UTF-8 %s",
+      encoding, (err) ? "(using ISO-8859-15 as fallback)" : "");
+
+  if (utf8 == NULL)
+    utf8 = g_strdup ("invalid subtitle");
+
+  newbuf = gst_buffer_new ();
+  GST_BUFFER_MALLOCDATA (newbuf) = (guint8 *) utf8;
+  GST_BUFFER_DATA (newbuf) = (guint8 *) utf8;
+  GST_BUFFER_SIZE (newbuf) = strlen (utf8);
+  gst_buffer_stamp (newbuf, buf);
+
+  gst_buffer_unref (buf);
+  return newbuf;
+}
+
 static gboolean
 gst_matroska_demux_parse_blockgroup_or_simpleblock (GstMatroskaDemux * demux,
     guint64 cluster_time, gboolean is_simpleblock)
@@ -2415,6 +2484,12 @@ gst_matroska_demux_parse_blockgroup_or_simpleblock (GstMatroskaDemux * demux,
           GST_TIME_ARGS (GST_BUFFER_DURATION (sub)));
 
       gst_buffer_set_caps (sub, GST_PAD_CAPS (stream->pad));
+
+      /* Fix up broken files with subtitles that are not UTF8 */
+      if (stream->type == GST_MATROSKA_TRACK_TYPE_SUBTITLE) {
+        sub = gst_matroska_demux_check_subtitle_buffer (demux, stream, sub);
+      }
+
       ret = gst_pad_push (stream->pad, sub);
       if (ret != GST_FLOW_OK && ret != GST_FLOW_NOT_LINKED)
         got_error = TRUE;
@@ -3448,15 +3523,20 @@ gst_matroska_demux_subtitle_caps (GstMatroskaTrackSubtitleContext *
 
   if (!strcmp (codec_id, GST_MATROSKA_CODEC_ID_SUBTITLE_UTF8)) {
     caps = gst_caps_new_simple ("text/plain", NULL);
+    subtitlecontext->check_utf8 = TRUE;
   } else if (!strcmp (codec_id, GST_MATROSKA_CODEC_ID_SUBTITLE_SSA)) {
     caps = gst_caps_new_simple ("application/x-ssa", NULL);
+    subtitlecontext->check_utf8 = TRUE;
   } else if (!strcmp (codec_id, GST_MATROSKA_CODEC_ID_SUBTITLE_ASS)) {
     caps = gst_caps_new_simple ("application/x-ass", NULL);
+    subtitlecontext->check_utf8 = TRUE;
   } else if (!strcmp (codec_id, GST_MATROSKA_CODEC_ID_SUBTITLE_USF)) {
     caps = gst_caps_new_simple ("application/x-usf", NULL);
+    subtitlecontext->check_utf8 = TRUE;
   } else {
     GST_DEBUG ("Unknown subtitle stream: codec_id='%s'", codec_id);
     caps = gst_caps_new_simple ("application/x-subtitle-unknown", NULL);
+    subtitlecontext->check_utf8 = FALSE;
   }
 
   if (data != NULL && size > 0) {

diff --git a/gst/matroska/matroska-ids.c b/gst/matroska/matroska-ids.c
index dde4676..db8261a 100644 (file)
--- a/gst/matroska/matroska-ids.c
+++ b/gst/matroska/matroska-ids.c
@@ -105,6 +105,7 @@ gst_matroska_track_init_subtitle_context (GstMatroskaTrackContext ** p_context)
   *p_context = (GstMatroskaTrackContext *) subtitle_context;
 
   (*p_context)->type = GST_MATROSKA_TRACK_TYPE_SUBTITLE;
+  subtitle_context->invalid_utf8 = FALSE;
   return TRUE;
 }
 

diff --git a/gst/matroska/matroska-ids.h b/gst/matroska/matroska-ids.h
index 0016b65..b9ba8cc 100644 (file)
--- a/gst/matroska/matroska-ids.h
+++ b/gst/matroska/matroska-ids.h
@@ -284,7 +284,8 @@ typedef struct _GstMatroskaTrackComplexContext {
 typedef struct _GstMatroskaTrackSubtitleContext {
   GstMatroskaTrackContext parent;
 
-  /* or here... */
+  gboolean    check_utf8;     /* buffers should be valid UTF-8 */
+  gboolean    invalid_utf8;   /* work around broken files      */
 } GstMatroskaTrackSubtitleContext;
 
 typedef struct _GstMatroskaIndex {