#libwebvtt The library for interpreting and authoring conformant WebVTT content WebVTT is a caption and subtitle format designed for use with HTML5 audio and video elements.

See: W3C WebVTT Draft

##Build Instructions:

In Unix-like environments, we use autotools:

./bootstrap.sh && ./configure && make

On Windows, we use a Visual Studio Project, see files in build/msvc2010

##Running Tests:

All tests are written using Google Test, and run using make check. You can configure the tests to run with our without valgrind, for memory checking.

Without valgrind:

./configure
make
make check

With valgrind:

./configure --enable-valgrind-testing
make
make check

When running tests with valgrind, any test that fails valgrind (even if it passes Google Test) will fail. See test/unit/Makefile.am for info on known test failures, and how to add/remove them.

##Routines available to application:

webvtt_status webvtt_create_parser( webvtt_cue_fn_ptr, webvtt_error_fn_ptr,
	void *, webvtt_parser * );
void webvtt_delete_parser( webvtt_parser );
webvtt_status webvtt_parse_chunk( webvtt_parser, const void *, webvtt_uint );

webvtt_status webvtt_string_new( webvtt_uint32, webvtt_string * );
void webvtt_string_delete( webvtt_string );
webvtt_status webvtt_string_append_utf8( webvtt_string *, const webvtt_byte *,
	webvtt_uint *, webvtt_uint, webvtt_utf8_reader );

void *webvtt_alloc( webvtt_uint );
void *webvtt_alloc0( webvtt_uint );
void webvtt_free( void * );
void webvtt_set_allocator( webvtt_alloc_fn_ptr, webvtt_free_fn_ptr, void * );

const char *webvtt_strerror( webvtt_error );

##TODO:

###General

• Change parser to convert to UTF16 as it goes through the buffer and convert parser to work with UTF16 instead of UTF8.
• node-ffi bindings for libwebvtt structures and public API, to assist with unit testing
• New unit tests, and convert some existing conformance tests to unit tests
• Integration with TravisCI
• Bug fixes
• Thread-safety where needed (Lock before setting global function pointers)
• Reference counting objects (may be a hassle, but would help to prevent leaking dropped cue objects)
• Improvements to lexer/parser state machines, in terms of speed and conformance
• Default error handlers when not provided by user
• Internationalized error messages
• Shared-library compilation
• Routines for generating conformant WebVTT text
• Language bindings (Rust, C++, D, etc)
• GNU Autotools build system to enhance portability (With MPL2'd M4 from Mozilla-Central to support Windows environments)

###Cue Text Payload Parsing

• Get shared UTF16 time stamp parser function working in order to create a time stamp node/token.
• Error reporting
• Line and character tracking in the cue_text in order to report errors correctly
• Possible change of token and node data structs to allow easier readability
• MANY bugs