Author: muggenhor
Date: Mon May  5 15:07:49 2008
New Revision: 4936

URL: http://svn.gna.org/viewcvs/warzone?rev=4936&view=rev
Log:
Merged revisions r4878:4879, r4881:4884, r4886:4889, r4891:4897, r4899, r4901, 
r4903, r4905:4908, r4910, r4912:4918, r4920, r4923, r4925, r4929 and r4933:4935 
into the 2.1 branch via svnmerge from trunk

........
  r4878 | muggenhor | 2008-05-02 14:38:44 +0200 (vr, 02 mei 2008) | 1 line
  
  Assert that the pointers to strlcpy & strlcat are not NULL
........
  r4879 | muggenhor | 2008-05-02 19:59:39 +0200 (vr, 02 mei 2008) | 1 line
  
  Don't ASSERT on a parsing error, use debug(LOG_ERROR, ... instead (the 
parser's return-value will be enough to determine that an error occurred)
........
  r4882 | muggenhor | 2008-05-02 20:22:59 +0200 (vr, 02 mei 2008) | 1 line
  
  Allow level files (*.lev) to be empty, though give a warning when they are 
(LOG_WARNING)
........
  r4883 | muggenhor | 2008-05-02 20:44:01 +0200 (vr, 02 mei 2008) | 5 lines
  
  Add the beginnings of an external unit test framework (i.e. it doesn't run in 
Warzone itself, it just links to pieces of Warzone's code to test it):
   * Add a new directory with unit tests (/tests/)
    * Run `make test` in this directory to execute all these unit tests and get 
reports from them if they fail
   * Add a single unit test to the directory which tests level_parser.y and the 
input it should & shouldn't accept
........
  r4884 | muggenhor | 2008-05-02 20:58:43 +0200 (vr, 02 mei 2008) | 1 line
  
  Less leaking of memory from level_parser.y
........
  r4886 | muggenhor | 2008-05-02 21:51:23 +0200 (vr, 02 mei 2008) | 1 line
  
  clean up levShutDown()
........
  r4887 | muggenhor | 2008-05-02 22:30:45 +0200 (vr, 02 mei 2008) | 1 line
  
  Call debug_exit() on exit from the parser unit test so that valgrind testing 
becomes easier
........
  r4888 | muggenhor | 2008-05-02 22:37:23 +0200 (vr, 02 mei 2008) | 5 lines
  
  Fix several memory leaks in level_parser.y that occur when parsing errors 
occur:
   * In case of a semantic error (i.e. we detect something is wrong in a 
parsing rule) deallocate the resources of our current tokens __before__ calling 
YYABORT
   * Only add a LEVEL_DATASET* to the psLevels list when we've fully and 
succesfully finished parsing of the corresponding level_entry
   * Add a bison "destructor" that handles destruction of tokens on the parsing 
stack when a syntax error occurs (i.e. the current tokens cannot be matched to 
any parsing rule)
........
  r4889 | muggenhor | 2008-05-02 22:59:35 +0200 (vr, 02 mei 2008) | 1 line
  
  #include "frame.h" in strlfuncs.h; to hopefully fix bug #11609
........
  r4891 | per | 2008-05-02 23:10:08 +0200 (vr, 02 mei 2008) | 2 lines
  
  Remove unused aMapLinePoints global.
........
  r4892 | muggenhor | 2008-05-02 23:19:15 +0200 (vr, 02 mei 2008) | 3 lines
  
   * Utilise ARRAY_SIZE to determine the amount of elements in a statically 
sized array
   * Use sizeof(*ptr) instead of sizeof(type) in malloc & memset calls
........
  r4893 | muggenhor | 2008-05-02 23:38:50 +0200 (vr, 02 mei 2008) | 1 line
  
  Update svn:ignore and bugtraq:* properties on /tests/
........
  r4894 | muggenhor | 2008-05-02 23:55:24 +0200 (vr, 02 mei 2008) | 1 line
  
  Perform the actual function call of the parsing function (levParse) in 
doParseTest() in t1000.c rather than parse_test.c
........
  r4895 | muggenhor | 2008-05-03 12:17:06 +0200 (za, 03 mei 2008) | 3 lines
  
  Fix bug #11480:
   * Don't call PHYSFS_close if our PHYSFS_file* pointer is NULL
........
  r4896 | muggenhor | 2008-05-03 12:24:36 +0200 (za, 03 mei 2008) | 1 line
  
  Don't call function pointers when they're NULL (fixes a segfault bug that 
occurs when a resource type without release function couldn't be loaded)
........
  r4897 | muggenhor | 2008-05-03 13:16:03 +0200 (za, 03 mei 2008) | 1 line
  
  Move psActionTarget into a more local scope, to ease debugging of bug #11609
........
  r4899 | per | 2008-05-03 19:04:15 +0200 (za, 03 mei 2008) | 2 lines
  
  Remove references to old bitmap font images.
........
  r4906 | per | 2008-05-04 00:04:05 +0200 (zo, 04 mei 2008) | 2 lines
  
  Remove unused code
........
  r4907 | muggenhor | 2008-05-04 01:13:20 +0200 (zo, 04 mei 2008) | 4 lines
  
  Remove function declarations for non-implemented functions receiveWholeDroid 
and sendWholeDroid
  
  PS Thanks Troman for granting me the privilege of removing these 
declarations. I just love cleaning up code ;-)
........
  r4908 | per | 2008-05-04 13:43:28 +0200 (zo, 04 mei 2008) | 6 lines
  
  WZ_DECL_CONST now implies WZ_DECL_WARN_UNUSED_RESULT and also works on MSVC.
  WZ_DECL_RESTRICT now works on MSVC as well.
  WZ_DECL_THREAD added, declares variables thread local, support for it made 
mandatory.
  WZ_ASSERT_STATIC_STRING (gcc only) to assert that a string is static added.
  The sstr functions now support pointer-length strings again.
........
  r4913 | per | 2008-05-04 14:34:20 +0200 (zo, 04 mei 2008) | 2 lines
  
  Remove MSVC support for WZ_DECL_CONST since it didn't work.
........
  r4914 | troman | 2008-05-04 14:55:05 +0200 (zo, 04 mei 2008) | 1 line
  
  Fix MSVC compilation.
........
  r4915 | muggenhor | 2008-05-04 15:56:27 +0200 (zo, 04 mei 2008) | 1 line
  
  Don't check for strlcpy and strlcat in the configure script as simply 
#defining _XOPEN_SOURCE (what config.h does) is enough to prevent declarations 
of these functions from being included (thus the #inclusion of "frame.h" in 
strlfuncs.h should be enough (this should fix bug #11609)
........
  r4916 | muggenhor | 2008-05-04 16:56:33 +0200 (zo, 04 mei 2008) | 3 lines
  
   * Turn macros TILE_OCCUPIED, TILE_HAS_STRUCTURE, TILE_HAS_FEATURE, 
TILE_HAS_WALL, TILE_HIGHLIGHT, TILE_HAS_TALLSTRUCTURE and 
TILE_HAS_SMALLSTRUCTURE into static inline functions TileIsOccupied, 
TileHasStructure, TileHasFeature, TileHasWall, TileIsHighlighted, 
TileHasTallStructure and TileHasSmallStructure respectively
   * Move the definition of MAPTILE above these ^^ macros and functions to 
prevent declarations problems
........
  r4917 | muggenhor | 2008-05-04 17:04:07 +0200 (zo, 04 mei 2008) | 1 line
  
  sDisplay is a member variable from BASE_OBJECT, so don't bother casting 
tile->psObject to a STRUCTURE* or FEATURE* before using this member variable
........
  r4918 | muggenhor | 2008-05-04 17:12:37 +0200 (zo, 04 mei 2008) | 3 lines
  
   * Remove macro TERRAIN_TYPES which was just an alias for TER_MAX (and 
replace all instances of it with TER_MAX)
   * Move some variables into a more local scope
........
  r4933 | per | 2008-05-04 22:58:20 +0200 (zo, 04 mei 2008) | 2 lines
  
  Various cosmetic cleanups in hci.c
........
  r4935 | per | 2008-05-04 23:31:44 +0200 (zo, 04 mei 2008) | 2 lines
  
  More cosmetic cleanups. Add comment on previous hack fix.
........

Added:
    branches/2.1/tests/
      - copied from r4935, trunk/tests/
Modified:
    branches/2.1/   (props changed)
    branches/2.1/Makefile.am
    branches/2.1/configure.ac
    branches/2.1/data/images/frontend.img
    branches/2.1/lib/framework/frameresource.c
    branches/2.1/lib/framework/strlfuncs.h
    branches/2.1/lib/framework/wzglobal.h
    branches/2.1/lib/ivis_common/pievector.h
    branches/2.1/lib/netplay/netplay.c
    branches/2.1/src/action.c
    branches/2.1/src/display3d.c
    branches/2.1/src/droid.c
    branches/2.1/src/droiddef.h
    branches/2.1/src/feature.c
    branches/2.1/src/fpath.c
    branches/2.1/src/frend.h
    branches/2.1/src/game.c
    branches/2.1/src/hci.c
    branches/2.1/src/level_parser.y
    branches/2.1/src/levels.c
    branches/2.1/src/map.c
    branches/2.1/src/map.h
    branches/2.1/src/message.c
    branches/2.1/src/mission.c
    branches/2.1/src/missiondef.h
    branches/2.1/src/multirecv.h
    branches/2.1/src/multistruct.c
    branches/2.1/src/multisync.c
    branches/2.1/src/raycast.c
    branches/2.1/src/scriptai.c
    branches/2.1/src/scriptfuncs.c
    branches/2.1/src/stats.c
    branches/2.1/src/structure.c
    branches/2.1/src/visibility.c


_______________________________________________
Warzone-commits mailing list
[email protected]
https://mail.gna.org/listinfo/warzone-commits

Reply via email to