Similarly, add #if TM_DEBUG around various unused debugging functions.
The #if 0 surrounded functions are good candidate for future removal
if decided they are not needed any more.
Since both the file tags and workspace tags are sorted, it is unnecessary
to completely resort the tags for the workspace - instead, remove the
tags belonging to the file from the workspace tags and merge the updated
file's tags into the workspace tags.
The merge is optimized for merging small number of tags into a large array.
For instance, the total number of tags in linux kernel is about 2300000
while the average number of tags per file is about 65 (35000 source files).
Most of the time merge won't be performed so we can avoid expensive
string comparisons and try to make bigger jumps to see if some merge is
needed or not (more details in the source file comments).
In addition, rename all functions, parameters, comments etc. mentioning
work_object and remove unnecessary parameters of various functions.
Delete dead code paths.
Also move common functions like tm_get_real_path() from tm_work_object to
tm_source_file.
In addition, this patch defines the GEANY_PRIVATE macro for the tag
manager library where it wasn't defined before, removes 2 tag manager
headers from distribution as they are not needed by plugins and
in the tag manager changes the docstrings to ordinary comments for
private definitions.
This also makes recent GCC shut up about indexing below the bounds as
it detected the code checked for a negative index yet didn't guard the
actual access. For now GCC doesn't understand the more comprehensive
check, but it might come back if GCC becomes smart enough.
Anyway, this makes the Assert() more correct, and addition of the
explicit kinds array size makes sure any future kind addition won't
miss its entry.
Rust now allows CRLF line endings in source files, which doesn't actually
change any lexing here but the comment was wrong.
Also, the hashbang initial comment has a special case where #![ doesn't count
as a comment (but instead an inner attribute that just happens to be on the
first line).
Previously, only the string contents were stored in lexerState::token_str (i.e.
not including the delimeters). Now, the delimeters are stored as well, thus
preserving them. This also simplifies the code a bit.
A new function is added to handle the character storage, which is also now
applied for normal identifiers. To that end, the MAX_STRING_LENGTH was boosted
to 256 so that all reasonably sized identifiers may fit.
Previously, things like:
struct Foo
{
#[bar]
baz: int
}
or
struct Foo
{
#![bar]
baz: int
}
would horribly confuse the parser and prevent proper parsing of the rest of the
file.
Fix a race initialization leading to incorrect handling of preprocessor
directives on the first input line.
Test case contributed by Adam Hirst, thanks.
To do this we change the internal types to only use types matched by
symbols_get_current_scope(). This is a bit of a hack, but the tag
types were already questionable and simply mapped to ones the
TagManager know.
This however merges Functions and Subroutines under the same top-level
item in the symbol list.
