We need to distribute the GtkDoc header generation script.
We need not to distribute the generated header, otherwise the VPATH
lookup would find the distributed one in $(srcdir) and try to update
it; but touching $(srcdir) is a bad thing.
- do not write out scintilla_object_* functions
- remove scioutput file parameter
- do not open output file until after xml parsing to avoid stale files
- fixed simplesect handling
- handle output being a pipe to head or tail
- split doxygen make recipe up
- rename geany_includedir to geany_gtkdocincludedir
- removed all references to geany-scintilla-gtkdoc.h
Also, remove geany-scintilla-gtkdoc.h from .gitignore
This prevents a make-time error later if it can't be imported. lxml is
considered optional so there's no build failure if Python doesn't exist or if
the lxml package can't be imported.
The script reads the doxygen xml output and generates two headers (optionally
a single header) that contains all of the plugin API in gtk-doc format.
Two headers because it's preferrable to group Scintilla related stuff
into its own namespace. This is a lot easier if g-ir-scanner can
work with a separate header file. If we change minds later on the script is
prepared to generate only one header.
A script will use the xml to generate a gtkdoc'ized header of the plugin API.
The xml files are also installed so that external users can use the xml
that corresponds to the installed version of Geany.
For now a separet doxyfile is used because the gtkdoc'ized header needs
a few types to be documented which not desired to be documented generally.
This script was used in the nightly builds to verify plugins will load
and have no undefined symbol references. Since the new way plugins
are built and linked, this is no longer necessary.
Additionally, this script won't work with new style geany_load_module()
plugins.
Because the stash_group_new() is an exported API, it has to be at least a boxed
type to be usable for gobject introspection. The boxed type uses reference
counting as opposed to memory duplication.
The obligatory stash_group_dup() is not exported (doesn't have to).
Because the tm_source_file_new() is an exported API, it has to be at least a
boxed type to be usable for gobject introspection. The boxed type uses
reference counting as opposed to memory duplication.
The obligatory tm_source_file_dup() is not exported (doesn't have to).
This helps g-ir-scanner recognizing the data parameter as context storage,
allowing object methods to be used as callback (via wrappers). It goes even
so far that g_object_unref is propery passed as destroy func to
keybindings_set_item_full() and plugin_set_key_group_full().
See also https://sourceforge.net/p/ctags/bugs/194/
I didn't use the exact upstream patch only altering the C++ code path,
because as far as I know no c.c language recognize two consecutive
colons separated by whitespace as a single token, so there's no point
in carrying on mistakes from the past.
Both sorting by name and appearance makes sense for most languages. Some
users may prefer sorting by appearance so make it configurable in
preferences (the possibility to override the settings for specific
filetypes is preserved).
Thanks to Colomban Wendling for lots of improvements of this patch.
Fixes#313.
