Drop the loop in mem_read() in favor of a single memcpy() call.
This greatly improves performances when nmemb > 1, for a small loss
for some values of size when nmemb == 1. Gain can theoretically be
infinite since swapping nmemb and size parameters changes almost
nothing while it had a dramatic performance impact previously. Loss
is up to about 25% in the worst case for some values of size when
nmemb is 1.
Also, now the function always copies as much data as possible, not only
whole blocks. This follows the glibc implementation of fread() and
simplifies the code. Doing so also fixes the position after a partial
read to be at the last readable character rather than the end of the
last read block.
Fix handling of scopes starting with a non-ASCII character.
Actually, just drop the check on the first byte of the scope, as it
doesn't seem to serve any purpose as it only checks the first byte (so
isn't any kind of real validation; and as it predates Geany it's
impossible to know the real reason behind this check), and breaks
support for non-ASCII scopes.
`final` is not a normal keyword, as it only have a special meaning in
some specific context. So, use a special case instead of a keyword not
to break identifiers of that name.
If there was two hashes (#) in an inline comment, only the content
between the two was considered a comment.
X-Universal-CTags-Commit-ID: ee93f5b9f393e76a850cf8c894cc748a62981156
Most of the time there's no start of a string which means all the 10
strcmp()s are done for every character of the input. This is very expensive:
before this patch this function alone takes 55% of the parser time.
When comparing by character (and avoiding further comparison if the first
character doesn't match), this function takes only 11% of the parser time
so the performance of the parser nearly doubles.
In addition check for the "rb" prefix which is possible in Python 3.
Ported from universal-ctags.
The character following an '@' was dropped if it didn't start a string
literal.
This could lead to unexpected problems if '@' was valid in other
situations.
X-Universal-CTags-Commit-ID: 2e62f475af1db08850447de46f56db14ce99d2eb
See http://www.open-std.org/jtc1/sc22/WG14/www/docs/n1256.pdf 6.4.6§3.
Note: This is not exactly the upstream Universal CTags commit because
it depends on another change for adding the `enter` label, which was
then included here.
X-Universal-CTags-Commit-ID: 3b3b60c7664a321a31ec87de336fc6bda90c405e
When tested with 200000 LOC python file (created by making many copies
of scripts/create_py_tags.py), the tm_tags_remove_file_tags() function
takes about 50% of the CPU time when only this file is open. After adding
the linear path to tm_tags_remove_file_tags() it takes just about 2%. See
the comment in the patch for more details.
Add API to truncate a vString to a certain length. This doesn't support
growing the string, only shrinking it.
X-Universal-CTags-Commit-ID: 4e3d9edf2e7a8a476ff97bc678e71c3919b960f9
The previous fix, coming from [CTags bug #1988026], was incorrect if
the parent was not a root-level element, as it checked the level name
(unqualified) against the parent name (qualified).
However, there is no need to check the level name, all what counts is
the indentation level itself: if it's smaller than an existing level,
it ends it.
This fixes [CTags bug #356].
[CTags bug #1988026]: https://sourceforge.net/p/ctags/bugs/227/
[CTags bug #356]: https://sourceforge.net/p/ctags/bugs/356/
X-Universal-CTags-Commit-ID: ab91e6e1ae84b80870a1e8712fc7f3133e4b5542
Unfortunately varType is Geany-only so this patch cannot be ported to ctags.
The removal of the extra { read is not the most elegant thing but making
skipType() aware of the argList collection complicates things too much.
The tags_lang variable is set from the first tag in the found array but
the array may actually contain tags from several languages. This may
lead to two things:
1. Goto tag definition goes to a tag from a different filetype
2. Worse, the first tag is from a different language than the current file
and we get a message that no tag was found
Get lang for every tag in the array and rename tags_lang to tag_lang.
gtk_tree_store_set() becomes very slow when the tree gets bigger
because internally it calls gtk_tree_store_get_path() which counts
all the entries in a linked list of elements at the same tree level
to get the tree path.
Avoid the call of this function when not needed.
Fixing this is however only theoretically useful, as:
* no actual code paths can currently lead to it;
* even if the code actually ended up reading the uninitialized value,
it would still have a fully defined behavior as the result of the
check is irrelevant in the only case the uninitialized read can
happen.
Anyway, fix this to avoid any possible bad surprises in the future.
This makes it easier to define it consistently to what the compiler
and platform supports, and avoids having to include a special header
everywhere, which is some kind of a problem for separate libraries
like TagManager and especially Scintilla.
As we only use these macros from the source and not the headers, it
is fine for it to be defined to a configure-time check from the build
system.
Warning: Although Waf and Windows makefiles are updated they are not
tested an will probably required tuning.
The flags in this variables are used to tune the linker behavior on the
final libgeany (currently set the version information), so should only
used on really linked libraries, not Libtool helper libraries.
Checks if the compiler supports -fvisibility and the linker supports
-dynamic-list arguments and use them instead of hardcoding. The new
geany-lib.m4 also accomodates future use of Libtool versioning.
This will allow plugins to link against the core when accessing API
functions, now that the macro/struct/funcptr stuff is gone.
Also convert the helper libraries into Libtool helper libraries as
linking a shared library against static libraries is (apparently) not
portable.
Adds a new header `pluginexport.h` to put the macros in, could be
moved into an existing header (support.h?) by I didn't want to drag
a bunch of existing stuff into the source files for this one macro.
TagManager has relative include, this could be fixed by changing the
include directories for it if it's a problem. Mark the Scintilla
functions exported by re-declaring them in sciwrappers.c with the
attribute to avoid changing upstream Scintilla code.
Regression for the triple start string issue has been introduced in SVN
revision 669 (fishman git a314e11158307db84c0dadb758846b2302fe69cd) on
2008-06-11. In ctags 5.7 it did work, in 5.8 not anymore.
See also http://sourceforge.net/p/ctags/bugs/229/ for the original bug,
which led to the old fix.
The other issue with normal strings in skipEverything() is even older.
Anything at the module/file level:
* _... is protected
* rest is public
Anything at the class level:
* __...__ is public (magic method)
* __... is private
* _... is protected
* rest is public
Anything at the function/method level:
* everything is private
Closes https://github.com/fishman/ctags/issues/216.
