Author: muggenhor
Date: Mon May 12 23:03:30 2008
New Revision: 5098
URL: http://svn.gna.org/viewcvs/warzone?rev=5098&view=rev
Log:
Merged revisions r5070:5075, r5077 and r5079:5097 into the 2.1 branch via
svnmerge from trunk
........
r5070 | muggenhor | 2008-05-12 19:04:32 +0200 (ma, 12 mei 2008) | 1 line
Don't bother to check whether we're in a comment when reaching the end of a
WRF file
........
r5071 | muggenhor | 2008-05-12 19:13:11 +0200 (ma, 12 mei 2008) | 1 line
Reduce code duplication that occurs in both branches of an if-statement
........
r5072 | muggenhor | 2008-05-12 19:23:33 +0200 (ma, 12 mei 2008) | 3 lines
* Use the lexer_input framework instead of a custom YY_INPUT implementation
for resource_lexer.l
* Don't add the function name to the debug message in openLoadFile as
debug() does that itself already
........
r5073 | muggenhor | 2008-05-12 19:34:56 +0200 (ma, 12 mei 2008) | 3 lines
* Get rid of lexer function resGetErrorData and use res_get_lineno() and
res_get_text() instead
* Don't call abort() in function res_error as the parser will return an
error anyway (causing resLoad to return false)
........
r5074 | muggenhor | 2008-05-12 19:38:45 +0200 (ma, 12 mei 2008) | 1 line
Attempt to match an empty string __before__ a non-empty one (otherwise we
won't ever match it at all)
........
r5075 | muggenhor | 2008-05-12 19:41:00 +0200 (ma, 12 mei 2008) | 1 line
Use yytext instead of res_text
........
r5077 | muggenhor | 2008-05-12 19:50:49 +0200 (ma, 12 mei 2008) | 4 lines
From resource_parser.y: "No longer static ... lets use this area globally":
* Lets __not__ use very bad practices just because we can!!! Make it static
again
* Give strres_lexer.l such an array of its own (static as well)
........
r5079 | muggenhor | 2008-05-12 20:00:09 +0200 (ma, 12 mei 2008) | 3 lines
* Put textual tokens as found by resource_lexer.l in heap memory rather than
some crazy stack-like static string buffer
* Make sure that this heap memory __will__ be released regardless of syntax
errors during scanning
........
r5080 | muggenhor | 2008-05-12 20:03:26 +0200 (ma, 12 mei 2008) | 1 line
Cleanup (behaviour should be unchanged)
........
r5081 | muggenhor | 2008-05-12 20:09:25 +0200 (ma, 12 mei 2008) | 1 line
Minor cleanup and add Doxygen comments
........
r5082 | muggenhor | 2008-05-12 20:16:30 +0200 (ma, 12 mei 2008) | 1 line
Don't duplicate code that appears in both branches of an if-statement
........
r5083 | muggenhor | 2008-05-12 20:22:05 +0200 (ma, 12 mei 2008) | 3 lines
* Use the lexer_input framework instead of a custom YY_INPUT implementation
for strres_lexer.l
* Don't bother to check whether we're in a comment when reaching the end of
a string resource file
........
r5084 | muggenhor | 2008-05-12 20:35:36 +0200 (ma, 12 mei 2008) | 4 lines
* Use yytext instead of strres_text
* Get rid of lexer function strresGetErrorData and use strres_get_lineno()
and strres_get_text() instead
* Don't call abort() in function strres_error as the parser will return an
error anyway (causing strresLoad to return false)
........
r5085 | muggenhor | 2008-05-12 20:47:25 +0200 (ma, 12 mei 2008) | 4 lines
* Put textual tokens as found by strres_lexer.l in heap memory rather than
some crazy stack-like static string buffer
* Make sure that this heap memory __will__ be released regardless of syntax
errors during scanning
* Reduce code duplication in strres_parser.y by reducing a single parsing
rule with duplicated code up into two rules
........
r5086 | muggenhor | 2008-05-12 20:56:42 +0200 (ma, 12 mei 2008) | 5 lines
Cleanup:
* Remove our own strcpy implementation: stringCpy
* Don't use "d = malloc(strlen(s) + 1); strcpy(d, s);" to copy a string into
fresh heap memory, use "d = strdup(s);" instead
* Fix indentation
........
r5087 | muggenhor | 2008-05-12 20:58:41 +0200 (ma, 12 mei 2008) | 1 line
Fix a syntax error introduced in r5086
........
r5088 | muggenhor | 2008-05-12 21:09:19 +0200 (ma, 12 mei 2008) | 1 line
Call fflush(stdout) after fprintf(stdout, ... in parsetest.c
........
r5089 | muggenhor | 2008-05-12 21:40:53 +0200 (ma, 12 mei 2008) | 1 line
Fix an error causing us to use the wrong names for yyget_extra(),
yyset_extra(), yyget_lineno() and yyget_text()
........
r5090 | muggenhor | 2008-05-12 21:52:35 +0200 (ma, 12 mei 2008) | 1 line
Fix LDADD in tests/Makefile.am so that crosscompiling (or rather
crosslinking) with mingw32 is possible
........
r5091 | muggenhor | 2008-05-12 21:55:55 +0200 (ma, 12 mei 2008) | 1 line
Remove our own custom definition of YYLMAX (give Flex the liberty of
determining this itself)
........
r5092 | muggenhor | 2008-05-12 22:06:19 +0200 (ma, 12 mei 2008) | 1 line
Rename some audp_(.*) stuff to yy$1
........
r5093 | muggenhor | 2008-05-12 22:09:10 +0200 (ma, 12 mei 2008) | 3 lines
* Get rid of never initialised variables g_bParsingSubFile and g_fpOld (the
compiler probably set them to false & NULL respectively everytime, if not then
this code should have crashed consistently)
* Remove now useless yywrap() function
........
r5094 | muggenhor | 2008-05-12 22:29:13 +0200 (ma, 12 mei 2008) | 5 lines
* Use the lexer_input framework instead of a custom YY_INPUT implementation
for audp_lexer.l
* Get rid of lexer function parseGetErrorData and use audp_get_lineno() and
audp_get_text() instead
* Add a WZ_DECL_FORMAT declaration to audp_error to enable GCC's printf
format checking
* In audp_error allocate the buffer on the stack using alloca() instead of
an auto 1kB text buffer
........
r5095 | muggenhor | 2008-05-12 22:42:55 +0200 (ma, 12 mei 2008) | 3 lines
* Parse quoted text in audp_lexer.l similar to most other parsers/lexers we
use (is also less error prone)
* Add single line comments (using '//') to audp_lexer.l
........
r5096 | muggenhor | 2008-05-12 22:51:46 +0200 (ma, 12 mei 2008) | 3 lines
* ONESHOT and LOOP are __not__ value tokens (they're just regular tokens),
so don't mark them as such
* Add a new subrule `looping` and use that in rule audio_track instead of
duplicating two code paths for audio_track
........
r5097 | muggenhor | 2008-05-12 22:55:29 +0200 (ma, 12 mei 2008) | 3 lines
* Put textual tokens as found by audp_lexer.l in heap memory rather than a
static string buffer
* Make sure that this heap memory __will__ be released regardless of syntax
errors during scanning
........
Modified:
branches/2.1/ (props changed)
branches/2.1/lib/framework/frame.c
branches/2.1/lib/framework/frameresource.c
branches/2.1/lib/framework/resly.h
branches/2.1/lib/framework/resource_lexer.l
branches/2.1/lib/framework/resource_parser.y
branches/2.1/lib/framework/strres.c
branches/2.1/lib/framework/strres_lexer.l
branches/2.1/lib/framework/strres_parser.y
branches/2.1/lib/framework/strresly.h
branches/2.1/lib/gamelib/audp_lexer.l
branches/2.1/lib/gamelib/audp_parser.y
branches/2.1/lib/gamelib/parser.h
branches/2.1/lib/script/chat_parser.y
branches/2.1/src/parsetest.c
branches/2.1/tests/Makefile.am
_______________________________________________
Warzone-commits mailing list
[email protected]
https://mail.gna.org/listinfo/warzone-commits
