Update doc extraction tool to handle GObjectIntrospection annotations.

	* codegen/docextract.py (FunctionDoc): Renamed class to GtkDoc.
	(GtkDoc::annotations): Added a list field to store annotations which
	are 2-tuples of (name, value).
	(GtkDoc::ret): Modified field to store the return description along
	with a list of annotations as described above.
	(GtkDoc::params): Now holds a list of 3-tupples: name, description and
	annotations (as described above).
	(GtkDoc::block_type): Add a field to tell if the comment block is a
	function block, signal block or property block.
	(GtkDoc::set_type):
	(GtkDoc::get_type): Add methods for setting/getting the block type.
	(GtkDoc::add_param): Modified to also accept a list of annotations to
	be added with the parameter.
	(GtkDoc::add_annotation):
	(GtkDoc::get_annotations): Added methods to add/get annotations for
	the comment block.
	(GtkDoc::append_description): Renamed to append_to_description().
	(GtkDoc::get_param_description): Removed unused method.
	(GtkDoc::get_description): Added method to get block description.
	(GtkDoc::add_return): Added method to add a return accepting the first
	line of the description and its annotations.
	(GtkDoc::append_return): Renamed to append_to_return().
	(Regular expressions):
	 - Made the names of the variables un-abbreviated.

	 - Added 'since', 'deprecated' and 'rename to' regular expressions.

	 - Modified the return matching regular expression so that it doesn't
	   match descriptions that begin with 'Returns ...'.  This improves
	   the docs of many function.

	 - Added signal and property comment block identifier matching regular
	   expressions in case those are useful.

	- Modified existing identifier matching regular expressions (function,
	  signal, and property regular expressions) to properly parse
	  annotations.  Also added a regular expression for extracting
	  annotations from the parameter and return descriptions.

	- Refined the function name matching regular expression to only accept
	  identifiers that begin with a lowercase letter.  This eliminates
	  'SECTION:' matches.

	- Finally, grouped commonly related expressions like return_pattern,
	  since_pattern, etc.  into groups (in lists) so that matching those
	  sections can be done using loops.

	(Parsing algorithm): Modified the algorithm to use a functional
	approach to parsing.  Extra methods like skip_to_comment() and
	processs_params() have been added and used in the parse_file()
	function to now process the comment blocks.
	(parse_dir): Added file processing output to stderr.
	* codegen/docextract_to_xml.py (usage): Added function to print out
	the usage.
	(print_annotations): Added function to print the given list of
	annotations.
	(options): Added --with-signals (-i), with-properties (-p) and
	--with-annotation (-a) to the existing --source-dir (-s) option.

	(algorithm): Now prints annotations, if specified.  Also, prints
	signals and properties correctly (using names like Class::signal-one
	for signals and Classs:property) with xml such as <signal
	name="...">...</signal>.  The return xml is slightly modified with
	annotations but this would only be exhibited if annotation xml is
	requested.

0 files changed, 0 insertions, 0 deletions