matroska: add new parsing code

Add a new EBML parser implementation that should allow significant
improvements to the Matroska demuxer. The new parsing code is not
actually used yet by the demuxer. The only changes to existing code in
this commit are to generate the MATROSKA_ID_* / EBML_ID_* macro
definitions from the new implementation and to rename some of them
(the new implementation uses names matching the official Matroska spec).

The main parser implementation is added in ebml.c. There are two new
generated files, ebml_defs.c and ebml_types.h, that contain
definitions of EBML elements. Those are generated by the new script
TOOLS/matroska.py. There's a new Makefile target "generated_ebml" that
run the script to refresh the content of the generated files.

1 files changed, 46 insertions, 145 deletions

diff --git a/libmpdemux/ebml.h b/libmpdemux/ebml.h
index c3a785099c..fc6a0ddbac 100644
--- a/libmpdemux/ebml.h
+++ b/libmpdemux/ebml.h
@@ -20,158 +20,56 @@
 #define MPLAYER_EBML_H
 
 #include <inttypes.h>
+#include <stddef.h>
+#include <stdbool.h>
+
 #include "stream/stream.h"
 
 
 /* EBML version supported */
 #define EBML_VERSION 1
 
-/*
- * EBML element IDs. max. 32-bit.
- */
-
-/* top-level master-IDs */
-#define EBML_ID_HEADER                   0x1A45DFA3
-
-/* IDs in the HEADER master */
-#define EBML_ID_EBMLVERSION              0x4286
-#define EBML_ID_EBMLREADVERSION          0x42F7
-#define EBML_ID_EBMLMAXIDLENGTH          0x42F2
-#define EBML_ID_EBMLMAXSIZELENGTH        0x42F3
-#define EBML_ID_DOCTYPE                  0x4282
-#define EBML_ID_DOCTYPEVERSION           0x4287
-#define EBML_ID_DOCTYPEREADVERSION       0x4285
-
-/* general EBML types */
-#define EBML_ID_VOID                     0xEC
+enum ebml_elemtype {
+    EBML_TYPE_SUBELEMENTS,
+    EBML_TYPE_UINT,
+    EBML_TYPE_SINT,
+    EBML_TYPE_FLOAT,
+    EBML_TYPE_STR,
+    EBML_TYPE_BINARY,
+    EBML_TYPE_EBML_ID,
+};
+
+struct ebml_field_desc {
+    uint32_t id;
+    bool multiple;
+    int offset;
+    int count_offset;
+    const struct ebml_elem_desc *desc;
+};
+
+struct ebml_elem_desc {
+    char *name;
+    enum ebml_elemtype type;
+    int size;
+    int field_count;
+    const struct ebml_field_desc *fields;
+};
+
+struct ebml_parse_ctx {
+    void *talloc_ctx;
+    bool has_errors;
+    bool no_error_messages;
+};
+
+struct bstr {
+    uint8_t *start;
+    int len;
+};
+
+#include "ebml_types.h"
+
+#define EBML_ID_INVALID 0xffffffff
 
-/* ID returned in error cases */
-#define EBML_ID_INVALID                  0xFFFFFFFF
-
-
-/*
- * Matroska element IDs. max. 32-bit.
- */
-
-/* toplevel segment */
-#define MATROSKA_ID_SEGMENT              0x18538067
-
-/* matroska top-level master IDs */
-#define MATROSKA_ID_INFO                 0x1549A966
-#define MATROSKA_ID_TRACKS               0x1654AE6B
-#define MATROSKA_ID_CUES                 0x1C53BB6B
-#define MATROSKA_ID_TAGS                 0x1254C367
-#define MATROSKA_ID_SEEKHEAD             0x114D9B74
-#define MATROSKA_ID_ATTACHMENTS          0x1941A469
-#define MATROSKA_ID_CHAPTERS             0x1043A770
-#define MATROSKA_ID_CLUSTER              0x1F43B675
-
-/* IDs in the info master */
-#define MATROSKA_ID_TIMECODESCALE        0x2AD7B1
-#define MATROSKA_ID_DURATION             0x4489
-#define MATROSKA_ID_WRITINGAPP           0x5741
-#define MATROSKA_ID_MUXINGAPP            0x4D80
-#define MATROSKA_ID_DATEUTC              0x4461
-#define MATROSKA_ID_SEGMENTUID           0x73A4
-
-/* ID in the tracks master */
-#define MATROSKA_ID_TRACKENTRY           0xAE
-
-/* IDs in the trackentry master */
-#define MATROSKA_ID_TRACKNUMBER          0xD7
-#define MATROSKA_ID_TRACKUID             0x73C5
-#define MATROSKA_ID_TRACKTYPE            0x83
-#define MATROSKA_ID_TRACKAUDIO           0xE1
-#define MATROSKA_ID_TRACKVIDEO           0xE0
-#define MATROSKA_ID_CODECID              0x86
-#define MATROSKA_ID_CODECPRIVATE         0x63A2
-#define MATROSKA_ID_CODECNAME            0x258688
-#define MATROSKA_ID_CODECINFOURL         0x3B4040
-#define MATROSKA_ID_CODECDOWNLOADURL     0x26B240
-#define MATROSKA_ID_TRACKNAME            0x536E
-#define MATROSKA_ID_TRACKLANGUAGE        0x22B59C
-#define MATROSKA_ID_TRACKFLAGENABLED     0xB9
-#define MATROSKA_ID_TRACKFLAGDEFAULT     0x88
-#define MATROSKA_ID_TRACKFLAGLACING      0x9C
-#define MATROSKA_ID_TRACKMINCACHE        0x6DE7
-#define MATROSKA_ID_TRACKMAXCACHE        0x6DF8
-#define MATROSKA_ID_TRACKDEFAULTDURATION 0x23E383
-#define MATROSKA_ID_TRACKENCODINGS       0x6D80
-
-/* IDs in the trackaudio master */
-#define MATROSKA_ID_AUDIOSAMPLINGFREQ    0xB5
-#define MATROSKA_ID_AUDIOBITDEPTH        0x6264
-#define MATROSKA_ID_AUDIOCHANNELS        0x9F
-
-/* IDs in the trackvideo master */
-#define MATROSKA_ID_VIDEOFRAMERATE       0x2383E3
-#define MATROSKA_ID_VIDEODISPLAYWIDTH    0x54B0
-#define MATROSKA_ID_VIDEODISPLAYHEIGHT   0x54BA
-#define MATROSKA_ID_VIDEOPIXELWIDTH      0xB0
-#define MATROSKA_ID_VIDEOPIXELHEIGHT     0xBA
-#define MATROSKA_ID_VIDEOFLAGINTERLACED  0x9A
-#define MATROSKA_ID_VIDEOSTEREOMODE      0x53B9
-#define MATROSKA_ID_VIDEODISPLAYUNIT     0x54B2
-#define MATROSKA_ID_VIDEOASPECTRATIO     0x54B3
-#define MATROSKA_ID_VIDEOCOLOURSPACE     0x2EB524
-#define MATROSKA_ID_VIDEOGAMMA           0x2FB523
-
-/* IDs in the trackencodings master */
-#define MATROSKA_ID_CONTENTENCODING      0x6240
-#define MATROSKA_ID_CONTENTENCODINGORDER 0x5031
-#define MATROSKA_ID_CONTENTENCODINGSCOPE 0x5032
-#define MATROSKA_ID_CONTENTENCODINGTYPE  0x5033
-#define MATROSKA_ID_CONTENTCOMPRESSION   0x5034
-#define MATROSKA_ID_CONTENTCOMPALGO      0x4254
-#define MATROSKA_ID_CONTENTCOMPSETTINGS  0x4255
-
-/* ID in the cues master */
-#define MATROSKA_ID_POINTENTRY           0xBB
-
-/* IDs in the pointentry master */
-#define MATROSKA_ID_CUETIME              0xB3
-#define MATROSKA_ID_CUETRACKPOSITION     0xB7
-
-/* IDs in the cuetrackposition master */
-#define MATROSKA_ID_CUETRACK             0xF7
-#define MATROSKA_ID_CUECLUSTERPOSITION   0xF1
-
-/* IDs in the seekhead master */
-#define MATROSKA_ID_SEEKENTRY            0x4DBB
-
-/* IDs in the seekpoint master */
-#define MATROSKA_ID_SEEKID               0x53AB
-#define MATROSKA_ID_SEEKPOSITION         0x53AC
-
-/* IDs in the chapters master */
-#define MATROSKA_ID_EDITIONENTRY         0x45B9
-#define MATROSKA_ID_EDITIONFLAGDEFAULT   0x45DB
-#define MATROSKA_ID_EDITIONFLAGORDERED   0x45DD
-#define MATROSKA_ID_CHAPTERATOM          0xB6
-#define MATROSKA_ID_CHAPTERTIMESTART     0x91
-#define MATROSKA_ID_CHAPTERTIMEEND       0x92
-#define MATROSKA_ID_CHAPTERDISPLAY       0x80
-#define MATROSKA_ID_CHAPSTRING           0x85
-#define MATROSKA_ID_CHAPTERSEGMENTUID    0x6E67
-#define MATROSKA_ID_CHAPTERSEGMENTEDITIONUID\
-                                         0x6EBC
-
-/* IDs in the cluster master */
-#define MATROSKA_ID_CLUSTERTIMECODE      0xE7
-#define MATROSKA_ID_BLOCKGROUP           0xA0
-
-/* IDs in the blockgroup master */
-#define MATROSKA_ID_BLOCKDURATION        0x9B
-#define MATROSKA_ID_BLOCK                0xA1
-#define MATROSKA_ID_SIMPLEBLOCK          0xA3
-#define MATROSKA_ID_REFERENCEBLOCK       0xFB
-
-/* IDs in the attachments master */
-#define MATROSKA_ID_ATTACHEDFILE	 0x61A7
-#define MATROSKA_ID_FILENAME		 0x466E
-#define MATROSKA_ID_FILEMIMETYPE	 0x4660
-#define MATROSKA_ID_FILEDATA		 0x465C
-#define MATROSKA_ID_FILEUID		 0x46AE
 
 /* matroska track types */
 #define MATROSKA_TRACK_VIDEO    0x01 /* rectangle-shaped pictures aka video */
@@ -213,4 +111,7 @@ int ebml_read_skip (stream_t *s, uint64_t *length);
 uint32_t ebml_read_master (stream_t *s, uint64_t *length);
 char *ebml_read_header (stream_t *s, int *version);
 
+int ebml_read_element(struct stream *s, struct ebml_parse_ctx *ctx,
+                      void *target, const struct ebml_elem_desc *desc);
+
 #endif /* MPLAYER_EBML_H */