Commit Graph

639 Commits (e3c884a2498ab81705c3782ca19d5f9573a8f5d5)

Author SHA1 Message Date
Per Inge Mathisen 704117f85c Fix various doxygen errors and warnings
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5992 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-11 18:25:34 +00:00
Per Inge Mathisen be263e0b8c Fix missing define of WZ_DECL_THREAD on windows
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5991 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-11 18:16:06 +00:00
Buginator 21fdab0117 Modify debug output to specify the 'real' directory that physfs is using for the file in question.
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5958 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-07 20:00:54 +00:00
Giel van Schijndel 01c6ba831c Allow compiling of utf.c outside of framework's debug facilities (i.e. allow it to be used in betawidget)
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5942 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-06 16:20:06 +00:00
Giel van Schijndel bdae8ce9ef Only #include types.h in utf.h as we don't depend on the entire framework
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5940 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-06 16:02:26 +00:00
Giel van Schijndel 906f766b8a * When using a va_list multiple times make sure to va_copy it first
* Make sure to call va_end on all instances of va_list when we're done with it

Similar to r5930 (which fixed ticket:50)

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5931 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-05 20:10:47 +00:00
Giel van Schijndel 37c6b31651 * When using a va_list multiple times make sure to va_copy it first
* Because va_copy is C99 provide a fallback for it on non-C99 systems (which assumes that va_list is just a pointer to the stack)

This fixes ticket:50, thanks to Buginator for finding the cause (though this solution is different from his suggestion)


git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5930 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-05 19:38:22 +00:00
Freddie Witherden bef604763f Add an nbytes parameter to UTF8toUTF16 and UTF16toUTF8 which allow for fetching the number of bytes allocated.
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5917 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-02 14:37:01 +00:00
Freddie Witherden a8709983e0 Commit the final patch of patch #45; this cleans up utf.[ch].
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5916 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-02 11:21:12 +00:00
Giel van Schijndel 99fe22e0a4 Don't accept out-of-range Unicode codepoints in function unicode_utf8_char_length either
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5915 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-01 19:16:00 +00:00
Giel van Schijndel aabf08d6fa Change the UTF-8 encoder and decoder to only accept valid Unicode codepoints, not just every representable 32 bit unsigned integer
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5914 4a71c877-e1ca-e34f-864e-861f7616d084
2008-09-01 19:10:03 +00:00
Giel van Schijndel 0e88a86b8e Add two new functions: utf(8|16)_char_at_offset, which return a pointer to the index-th character in the given stream
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5911 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-31 20:39:17 +00:00
Giel van Schijndel f0f68e7386 Rename utf8.[ch] to utf.[ch] because it deals with different UTFs (Unicode Transformation Format) but not just specifically with UTF-8
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5910 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-31 20:06:40 +00:00
Giel van Schijndel d3eaf3bbd1 Add several new functions for converting between UTF-8 and UTF-16 to utf8.[ch]
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5902 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-30 15:15:16 +00:00
Giel van Schijndel a0cdfd44ab Rename utf8_(encode|decode) to utf8_(encode|decode)_utf32 because they convert between UTF-8 and UTF-32
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5901 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-30 15:15:09 +00:00
Giel van Schijndel dfa9c2bc01 Rename function unicode_strlen to utf32_strlen because it really determines the length of a UTF-32 encoded string
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5900 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-30 13:49:11 +00:00
Giel van Schijndel 2cb4c71883 * Split out the code that encodes a single Unicode character as UTF-8 into a separate function: encode_utf8_char
* This will allow encoding of Unicode strings to UTF-8 regardless of its current encoding

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5899 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-30 13:47:15 +00:00
Giel van Schijndel ac08ca70b0 * Seperate out the part that determines a single Unicode character's (codepoint actually) length when encoded as UTF-8 into a separate function: unicode_utf8_char_length
* This will allow determining of the length of a string when encoded as UTF-8 regardless of its current encoding

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5898 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-30 13:47:06 +00:00
Giel van Schijndel 5222037799 Move the duplicated UTF-8 decoding code out of functions utf8_character_count and utf8_decode and into new function decode_utf8_char, which only decodes a single UTF-8 character sequence
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5897 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-30 13:24:27 +00:00
Giel van Schijndel b10871a0ab Add utf8.c to the win32 makefile system (which I forgot in r5895)
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5896 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-30 12:38:26 +00:00
Giel van Schijndel d45e64bf55 Reintroduce utf8.{c,h} for encoding and decoding UTF-8 strings
This is meant to be used with betawidget

This is a revert of r2477 (which removed the UTF-8 module)


git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5895 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-30 12:34:46 +00:00
Ari Johnson d5b341334d More changes from #ifdef __APPLE__ to #if defined(WZ_OS_MAC)
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5880 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-27 15:24:21 +00:00
Freddie Witherden 55be99a19d Add sync-to-vblank support to Warzone; this closes ticket #38.
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5871 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-25 18:01:49 +00:00
Giel van Schijndel 173ee31214 * Split ASSERT out into ASSERT_HELPER:
* ASSERT_HELPER can be used by debug functions that wish to alter the location information attached and printed with ASSERTs
 * Add some missing semi-colons (the previous ASSERT macro had a terminating semi-colon)
 * Change an ASSERT that checks calloc's return value into an if() block (we can also run out of memory on non-debug builds...)


git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5865 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-24 15:42:59 +00:00
Giel van Schijndel 5e593ad615 Add platform detection code for GNU/kFreeBSD (OS with FreeBSD kernel and GNU utilities and libraries) to allow compiling on this system
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5794 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-11 15:29:44 +00:00
Per Inge Mathisen 6f89f25b80 Attempt fix for bug #12118: Could not find string for ID. Has high risk for creating
more asserts - please report them to me.


git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5781 4a71c877-e1ca-e34f-864e-861f7616d084
2008-08-07 15:40:21 +00:00
Giel van Schijndel 3eabf084ad Get rid of the ID number cruft in the string resource system:
* At the only place where these ID numbers where used (scrValDefSave and scrValDefLoad), use the key-string instead (e.g. "ZNULLBODY" instead of 17)
 * Rename function strresGetStringByID to strresGetString
 * Increment event save file version to 4 (version of savegame.es files)
 * Reject loading of savegames that need to load non-NULL ST_TEXTSTRING variables from version < 4

NOTE: This may break current savegames of games that use scripts with variables of type TEXTSTRING in them (AFAIK currently only ''some'' campaign and tutorial savegames have these).

This closes ticket:19

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5685 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-27 23:58:49 +00:00
Giel van Schijndel 5f496b5b33 Allow (warning-less) compilation with Flex 2.5.33
Patch #1089 by Bertrand Coconnier <bcoconni>

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5679 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-27 13:34:58 +00:00
Giel van Schijndel f2bc3c056f Change the debug message about a string that couldn't be found from an error message into a warning message because it seems to occur quite often
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5667 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-25 22:14:47 +00:00
Giel van Schijndel b40e7a6a90 Further hardcode the treap management code to the string resource system:
* Don't maintain a secondary linked list of strings in strres.c
 * Add the string that's tracked and it's ID number to the TREAP_NODE structure
 * Let the treap code manage its own memory
 * Have the treap code perform reverse lookups using either the resource string (i.e. not the "key" string) or associated ID number

NOTE: Searching a treap for anything else than the key is O(n), but that's no worse than a linked list, which is O(n) as well. Walking a treap has O(log n) as memory requirement, whereas walking a linked list has O(1) as memory requirement (stack size). So the only disadvantage is that searching the treap has a (highly remote!) possibility of stack overflows.

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5666 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-25 20:27:54 +00:00
Giel van Schijndel e8e664f85a * Get rid of function strresGetIDString who's only purpose is to determine whether the given ID string has a string resource associated with it, and if so return a string that's equal (according to strcmp) to the passed in ID string
* Replace the only use of strresGetIDString with a call to function strresGetStringByID, which will also return NULL on failure to find the given ID string

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5663 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 21:12:43 +00:00
Giel van Schijndel 8a41fb89d3 Make function strresStoreString const correct
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5662 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 21:12:39 +00:00
Giel van Schijndel a0e779ceb8 * Add a new function strresGetStringByID which combines the functionality of strresGetIDNum and strresGetString
* Get rid of function strresGetIDNum with its inconvenient API and use strresGetStringByID instead

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5661 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 20:10:16 +00:00
Giel van Schijndel 12e5120e69 Replace the linked list of arrays of string pointers (premature) optimisation with just a linked list of strings
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5660 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 18:57:09 +00:00
Giel van Schijndel 47d02a31da * Change strresGetString to return its string as const (to prevent potential double-free problems)
* Instead use strdup() where a non-const reference to these strings is used
  * NOTE: This may introduce memory leaks, but I'd rather have memory leaks than double-frees, of which I'm frankly surprised that they don't occur ''much'' more often
 * Change FindDroidTemplate to take its parameter as const
 * Don't multi-purpose the parameters of FindDroidTemplate as local variables
 * Store string variables in getName as const
 * scrvGetString:
  * Directly return the string it retrieved and use NULL to indicate failure
  * Return the string as `const'

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5658 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 18:18:07 +00:00
Giel van Schijndel da3c28efa2 * Move the allocation and initialisation of STR_ID structures to a function of its own: strresAllocIDStr
* Allocate STR_ID structures in such a way that a single free() call on it will suffice as cleanup

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5657 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 16:13:14 +00:00
Giel van Schijndel 8101f3bf5b * Change strresGetIDString to return its ID string const
* Change allocateName to do as the name suggests and actually ''allocate'' a string!

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5656 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 16:13:08 +00:00
Giel van Schijndel e3a7f0cab8 Return the retrieved string from strresGetIDString and allocateName and use NULL to indicate failure.
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5655 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 16:13:03 +00:00
Giel van Schijndel e8efe132f8 * Get rid of global variable *psCurrRes, which is used by the strres_parser to operate on
* Instead pass the current STR_RES* pointer to work on to the parser function

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5654 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 14:16:27 +00:00
Giel van Schijndel 16e5fd8764 Get rid of the ID_ALLOC bitflag (marks strings allocated by the string resource system), as ''all'' strings are allocated by the string resource system anyway.
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5652 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-24 13:59:16 +00:00
Giel van Schijndel 8eb2ed2827 Add some ASSERTs which I think should always hold true for the purpose of ''understanding'' the string resource code. As it's difficult to understand, due to it multi-purposing one variable to hold two semantically different values (at the same time!).
In any case I ''think'' these ASSERTs always hold true, (and for the testing I've done so far they do). But I'd like some more certainty and confirmation, hence why I'm committing this (before I make changes that rely on that assertion holding true), so other people get a chance to test this and prove my assertion wrong.

Thus, ''if'', these ASSERTs are triggered, please open a bugreport or otherwise notify me of it (e-mail, IRC, text message, postal mail, phone call, etc.)

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5649 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-23 23:42:18 +00:00
Giel van Schijndel cf374ee5e1 * Move the definitions of the datatypes used by the string resource system from strres.h to strres.c
* Add a forward declaration for STR_RES to strres.h (so that we can still declare pointers to this type)

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5645 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-22 22:36:23 +00:00
Giel van Schijndel d430c4b44d * Initialize global psStringRes to NULL
* Return the STR_RES* pointer from strresCreate and use NULL to indicate failure
 * Use size_t to store memory sizes in instead of UDWORD

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5644 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-22 22:33:10 +00:00
Giel van Schijndel 4f3c6d39a4 * Return the resulting STR_BLOCK* from strresAllocBlock and return NULL on failure
* Use calloc instead of a malloc followed by memset(0)
 * Check the return value from malloc&calloc for ''all'' invocations of it

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5643 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-22 22:14:48 +00:00
Giel van Schijndel 38f5b64c91 Add a new command line option: --flush-debug-stderr which causes the debug output to be flushed when used
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5636 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-21 19:49:36 +00:00
Giel van Schijndel ef6fa22424 Fix a wrong call to calloc (parameter list was malloc style)
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5631 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-21 01:00:37 +00:00
Giel van Schijndel f6a6179569 * Use calloc instead of malloc followed by memset(0)
* Use an unsigned int instead of UDWORD as we don't depend on the size guarantee
 * Move the definition of aUsage into #ifdef DEBUG_CHECK_FOR_UNUSED_STRINGS, because that's the only place where it's used

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5630 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-21 00:57:32 +00:00
Giel van Schijndel 7d42778f80 Don't #define DEBUG_GROUP0 in places where it cannot possibly be used
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5629 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-21 00:52:33 +00:00
Giel van Schijndel a3ecb8a325 Get rid of the custom TREAP* structure and instead use a pointer to a pointer to a TREAP_NODE** structure
git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5628 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-21 00:43:33 +00:00
Giel van Schijndel 7c8789f108 * Don't associate line number and file information with calls to treapAdd, as it is only called once
* Replace use of the forwarding macro TREAP_ADD with a direct call to treapAdd
 * Remove treap debugging code (actually that's what the above changes effectively accomplished)

git-svn-id: svn+ssh://svn.gna.org/svn/warzone/trunk@5627 4a71c877-e1ca-e34f-864e-861f7616d084
2008-07-21 00:30:55 +00:00