Our previous assumption that all paths in Windows are encoded in UTF-8
is rather weak, specially when considering that Git is
encoding-agnostic.
These set of functions allow the user to change the library's active
codepage globally, so it is possible to access paths and files on all
international versions of Windows.
Note that the default encoding here is UTF-8 because we assume that 99%
of all Git repositories will be in UTF-8.
Also, if you use non-ascii characters in paths, anywhere, please burn on
a fire.
If ExpandEnvironmentStringsW is successful, it returns the amount of
characters written, including the NUL terminator.
Thanks to Emeric for reading the MSDN documentation correctly.
Signed-off-by: Carlos Martín Nieto <carlos@cmartin.tk>
There were quite a few places were spaces were being used instead of
tabs. Try to catch them all. This should hopefully not break anything.
Except for `git blame`. Oh well.
1. The license header is technically not valid if it doesn't have a
copyright signature.
2. The COPYING file has been updated with the different licenses used in
the project.
3. The full GPLv2 header in each file annoys me.
Drop the GLibc implementation of Merge Sort and replace it with Timsort.
The algorithm has been tuned to work on arrays of pointers (void **),
so there's no longer a need to abstract the byte-width of each element
in the array.
All the comparison callbacks now take pointers-to-elements, not
pointers-to-pointers, so there's now one less level of dereferencing.
E.g.
int index_cmp(const void *a, const void *b)
{
- const git_index_entry *entry_a = *(const git_index_entry **)(a);
+ const git_index_entry *entry_a = (const git_index_entry *)(a);
The result is up to a 40% speed-up when sorting vectors. Memory usage
remains lineal.
A new `bsearch` implementation has been added, whose callback also
supplies pointer-to-elements, to uniform the Vector API again.
Cleaned up the structure of the whole OS-abstraction layer.
fileops.c now contains a set of utility methods for file management used
by the library. These are abstractions on top of the original POSIX
calls.
There's a new file called `posix.c` that contains
emulations/reimplementations of all the POSIX calls the library uses.
These are prefixed with `p_`. There's a specific posix file for each
platform (win32 and unix).
All the path-related methods have been moved from `utils.c` to `path.c`
and have their own prefix.
If a config has several files, we need to check all of them before we
can say that a variable doesn't exist.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
"git_config_backend" have been renamed to "git_config_file", which
implements a generic interface to access a configuration file -- be it
either on disk, from a DB or whatever mumbojumbo.
I think this makes more sense.
Configuration options can come from different sources. Currently,
there is only support for reading them from a flat file, but it might
make sense to read it from a database at some point.
Move the parsing code into src/config_file.c and create an include
file include/git2/config_backend.h to allow for other backends to be
developed.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
The section and variable names use different rules, so store them as
two different variables internally.
This will simplify the configuration-writing code as well later on,
but even with parsing, the code is simpler.
Take this opportunity to add a variable to the list directly when
parsing instead of passing through config_set.
long int is a safer type than int unless the user knows that the
variable is going to be quite small.
The code has been reworked to use strtol instead of the more
complicated sscanf.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
Many error paths freed their local data althought it is freed later on
when the end of the function notices that there was an error. This can
cause double frees and invalid memory access.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
Make cvar_free return void instad of the next element, as it was
mostly a hack to make cvar_list_free shorter but it's now using the
list macros.
Also check if the input is NULL and return immediately in that case.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
There is no need to keep config file in memory until the the
configuration is freed. Free the buffer immediately after the
configuration has been parsed.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
If a variable name appears on its own in a line, it's assumed the
value is true. Store the variable name as NULL in that case.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
If a line ends at EOF there is no need to check for the newline
character and doing so will cause us to read memory beyond the
allocatd memory as we check for the Windows-style new-line, which is
two bytes long.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
If a variable value has the traditional continuation character (\) as
the last non-space character in the line, then we continue reading the
value on the next line.
Using more than two lines is also supported.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
Make header and variable parse functions use their own buffers instead
of giving them the line they need to read as a parameter which they
mostly ignore.
This is in preparation for multiline configuration variables.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
A variable name is stored internally with its section the way it
appeared in the configuration file in order to have the information
about what parts are case-sensitive inline.
Really implement parse_section_header_ext and move the assignment of
variables to config_parse.
The variable name matching is now done in a case-away way by
cvar_name_match and cvar_section_match. Before the user sees it, it's
normalized to the two- or three-dot version.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
Such a list preserves the order the variables were first read in which
will be useful later for merging different data-sets. Furthermore,
reading and writing out the same configuration should not reorganize
the variables, which could happen when iterating through all the items
in a hash table.
A hash table is overkill for this small a data-set anyway.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
git_config_open shouldn't have to initialise variables that are only
used inside config_parse and its callees.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
Config variables should be interpreted at run-time, as we don't know if a
zero means false or zero, or if yes means true or "yes".
As a variable has no intrinsic type, git_cvtype is gone and the public
API takes care of enforcing a few rules.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
These functions can be used to query or modify the variables in a
given configuration. No sanity checking is done on the variable names.
This is mostly meant as an API preview.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
If cfg_readline consumes the line, then parse_section_header will read
past it and if we read a character, parse_variable won't have the full
name.
This solution is a bit hackish, but it's the simplest way to get the
code to parse correctly.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
Save the location of the name in section_out instead of returning it
as an int. Use the return code to signal success or failure.
Signed-off-by: Carlos Martín Nieto <cmn@elego.de>
