56aaae4949
Corrected example code for PHYSFS_enumerate in physfs.h
2019-08-24 21:06:54 -04:00
19ea59109d
Fixed comment typo.
2019-06-07 23:46:55 -04:00
a8180f0e4c
hog: Added support for HOG2 (Descent 3) archives.
2019-06-07 22:27:48 -04:00
f94ce0613d
Fixed compiler warning.
2019-03-18 14:28:46 -04:00
6b0839051f
windows: Workaround for WinXP systems.
...
(transplanted from c8f3bbd689d6b71b84c11db74275ea3d56fef961)
2019-03-18 13:36:16 -04:00
5cbb460bcd
Only flush file handles on close if they were opened for writing.
2019-03-18 11:27:26 -04:00
e549fe0e0f
Added some whitespace to test a fix to revision control.
2019-02-10 16:00:39 -05:00
500033f37d
Allow builds to opt-out or opt-in to specific archivers, whichever's easier.
2019-01-26 03:00:29 -05:00
b57d8960e8
Fixed some compiler warnings.
2018-11-28 00:23:08 -05:00
5786a58628
PHYSFS_flush() shouldn't call PHYSFS_Io::flush().
...
The former is meant to send PhysicsFS-buffered data to the PHYSFS_Io's
implementation, the latter is meant to tell the OS to definitely make sure the
data is safely written to disk (or at least, that's what it does in practice).
This was making PHYSFS_setBuffer()'d handles _slower_, since they would end
up blocking whenever the buffer was full until the data made the full trip to
physical media, instead of just letting the OS do its own buffering.
Now we still PHYSFS_Io::flush() on PHYSFS_close(), like this has always
worked. That might also be overkill, but that remains a historical artifact
of trying to keep the underlying file handle usable if pending writes fail
for possibly-recoverable reasons (which isn't guaranteed if you just close()
it, at least as far as I remember).
2018-11-27 23:53:33 -05:00
73d66441e3
Added PHYSFS_setRoot().
2018-10-17 23:44:02 -04:00
7d194a8a62
windows: Workaround GetUserProfileDirectory's API change in Win10 build 1809.
...
(transplanted from ece6769c0676c2d4e8a5893a1acebd0f65456817)
2018-10-03 22:40:57 -04:00
8ce294a458
PHYSFS_setWriteDir() shouldn't create an empty file if the dir doesn't exist.
2018-05-16 19:54:51 -04:00
15dd00659a
Fix up physfs.h for compilers that are sensitive about preprocessor defines.
...
(transplanted from db8f944df5c56f0244229813203fb5b24e8d9968)
2018-04-19 10:06:38 -04:00
7394c07d6b
apple: Patched to compile with older (mac 10.7) SDKs (thanks, Ken and Ryan!).
2018-03-24 00:19:59 -04:00
ea90a92016
7zip: don't forget to destroy the PHYSFS_Io when closing the archive!
2018-03-08 11:47:42 -05:00
3597a7b69a
Call the default branch "3.1.0" to avoid confusion.
...
This way, if someone builds the default branch instead of stable-3.0, we'll
be able to distinguish.
No immediate plans for a format 3.1 effort at the moment, though.
2018-03-08 11:51:46 -05:00
a45afc5d50
Catch access to paths that are just "." or ".." without any path separator.
2017-10-26 14:37:16 -04:00
62d24e228f
Fixed mounting a symlink to a real directory.
2017-10-26 14:21:36 -04:00
26db9376da
Fixed some infinite loops that a maliciously-crafted .iso can trigger.
...
These bugs exposed by American Fuzzy Lop (AFL), a powerful fuzzer.
http://lcamtuf.coredump.cx/afl/
2017-10-23 14:58:54 -04:00
1c7945461d
Don't allow NULL filenames to be mounted.
...
Regardless of what the 3.0.0 documentation says, PhysicsFS never handled this
correctly, so now we check for it so you can't get into crashy situations.
Corrected documentation to reflect reality.
2017-10-23 12:40:59 -04:00
f5e5b586c7
Fixed crash when duplicating PHYSFS_Io for zipfiles.
2017-10-23 12:16:51 -04:00
9e756312d5
Bumped version to 3.0.0!
2017-09-27 11:54:25 -04:00
bb2c38cdea
Added SLB to list of supported archives.
2017-09-27 12:33:36 -04:00
9b42fe23fd
Fixed doxygen comment.
2017-09-27 12:07:55 -04:00
3d8817346d
Fixed seeking within read buffers.
2017-09-25 16:32:52 -04:00
425131ccda
Simplified doBufferedRead().
...
No longer recurses, and deals with EOF correctly.
2017-09-25 16:19:59 -04:00
2b78f64c11
Fixed PHYSFS_flush().
2017-09-25 16:19:30 -04:00
395b2b02dd
OS/2: Patched to compile.
2017-09-15 19:52:24 -04:00
69d3df3286
OS/2: Deal with UTF-8 -> codepage conversion on older OS/2 installs.
...
(untested attempt.)
2017-09-15 19:49:40 -04:00
7a26b83733
OS/2: don't report failures on Unicode conversion on older OS/2 installs.
2017-09-14 23:36:51 -04:00
e728c65912
vdf: Don't add directories; UNPK_addEntry()/DirTree will fill in parents.
2017-09-03 13:13:51 -04:00
21b2fd39f2
Made PHYSFS_caseFold() a public API.
2017-08-20 02:02:08 -04:00
bc2ddf0a53
Added some warnings to the docs about converting bad Unicode strings.
2017-08-20 02:01:50 -04:00
9f8ecb91cb
unicode: Added UTF-16 and UCS-4 stricmp functions.
2017-08-20 01:18:41 -04:00
eb75883226
OS/2: Fixed NULL pointer dereference.
2017-08-20 01:18:05 -04:00
689333dc8f
Clarified some documentation.
2017-08-20 00:35:25 -04:00
6ad71878fe
Updated PHYSFS_Archiver with PHYSFS_EnumerateCallbackResult change.
2017-08-19 11:19:28 -04:00
4e0d3d55e9
Make PHYSFS_EnumerateCallback return an enum instead of an int.
2017-08-19 10:40:27 -04:00
5fb9119ebc
windows: Disable _CRT_SECURE_NO_WARNINGS again.
2017-08-16 21:19:16 -04:00
7607eeeb00
Cleaned up some header inclusion politics.
2017-08-16 20:03:24 -04:00
e19422cc06
qnx: Added more proper support for QNX.
2017-08-16 20:02:54 -04:00
cd0ccf64ed
Bump version to 2.1.1, since we've been at 2.1.0 for 8.5 years. :)
...
There isn't going to be an official 2.1.1 (or 2.1.0) release; we'll let this
marinate a little and jump to 3.0!
2017-08-15 02:08:24 -04:00
b311ee44ad
Fixed one more Visual Studio compiler warning.
2017-08-14 21:22:30 -04:00
735fd14ea4
Whoops, patched to compile.
2017-08-14 21:17:22 -04:00
8dfd3cff8d
Fixed some compiler warnings on Visual Studio.
...
(and maybe made a few new ones...)
2017-08-14 21:15:58 -04:00
692d5e8219
Fixed some warnings that Xcode brought up.
2017-08-14 19:58:56 -04:00
0d1a73fce4
zip: add some zip_resolve() calls, deal with DirTree-inserted parent dirs.
2017-08-14 16:24:19 -04:00
395ea71990
This appears to have been fixed at some point.
2017-08-14 14:06:43 -04:00
660171f79f
Don't fail enumeration if a directory isn't available in a given archive.
...
That shouldn't be a fatal error, that's a normal case.
2017-08-14 14:05:06 -04:00
Ryan C. Gordon
Francesco Bertolaccini