summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
* Just to be on the safe side about avoiding other portability hazardsSimon Tatham2005-11-13
| | | | | | | | | | | | in future, add `-ansi -pedantic' to the Halibut default compile options and fix the few resulting warnings (mostly signed/unsigned char mismatches and commas at the ends of enums). The one remaining warning I'm still seeing is `missing initializer' for the big table in charset/iso2022.c, but I think the code genuinely is more readable this way, and I haven't found a gcc option to disable that specific warning. [originally from svn r6458]
* `version' needs to be declared `extern'.Simon Tatham2005-11-13
| | | | [originally from svn r6457]
* Revert to "C" locale for LC_NUMERIC, so that PDFs won't be corruptedSimon Tatham2005-11-13
| | | | | | by the use of a comma as a decimal separator. [originally from svn r6456]
* Fix various compiler warnings and errors. In particular, my cunningSimon Tatham2005-11-13
| | | | | | | | auto-type-checking DEPLANARISE and REPLANARISE macros have turned out to only work in gcc, which is a shame. [originally from svn r6455] [this svn revision also touched charset,filter,timber]
* `style.c' appears to have been around since 1999 and never hadSimon Tatham2005-11-12
| | | | | | | | | | anything in it! In its current form it presents the portability hazards of an empty structure and an empty source file. Therefore, I'm removing it; if I ever have a clear idea of what a user style mechanism ought to look like, it might make a reappearance, but don't hold your breath. [originally from svn r6453]
* write_utf8() is used in iso2022.c as of r6378; declare it.Jacob Nevins2005-10-23
| | | | | | | | (Fixes a warning in iso2022.c. There are lots more.) [originally from svn r6424] [r6378 == 41e50e9f2e3e67da805c5d9037cc650f363e5279] [this svn revision also touched charset,filter,timber]
* Working ISO 2022 output function. Outputs full ISO 2022 (not sureSimon Tatham2005-10-07
| | | | | | | | | | | | | | | what that's useful for but it seemed a pity not to do it) and compound text. I've completely removed the compound text implementation from iso2022s.c in favour of using the more flexible iso2022.c, meaning we can cope with nastiness such as DOCS. This is largely untested: I've checked it on small examples as I went along, but it lacks anything resembling a proper test suite. [originally from svn r6378] [this svn revision also touched charset,filter,timber]
* PostScript StandardEncoding might occasionally come in handy. WhileSimon Tatham2005-10-06
| | | | | | | I'm here, I've updated the URL to the Adobe Glyph List. [originally from svn r6376] [this svn revision also touched charset,filter,timber]
* Remove the error message `no text found in paragraph'. Aaron BrownSimon Tatham2005-09-29
| | | | | | | | | | | | | points out that it's perfectly possible to generate an empty paragraph using legal Halibut syntax: a paragraph containing nothing but a \#{...} comment will do the job, and is quite likely to happen if you've commented out a load of Halibut code. Therefore, an empty paragraph is now silently ignored rather than being an error condition in itself; if you create an empty paragraph due to it containing an unrecognised directive, then you'll get an error for _that_ and only that. [originally from svn r6361]
* Correct the URL of the X Registry to the one given in the Registry, whichBen Harris2005-09-26
| | | | | | | works (unlike our old one). [originally from svn r6358] [this svn revision also touched charset,filter,timber]
* Never loop up to _and including_ lenof(array).Simon Tatham2005-09-26
| | | | | [originally from svn r6357] [this svn revision also touched charset,filter,timber]
* Correct copy and paste error.Simon Tatham2005-09-24
| | | | | [originally from svn r6354] [this svn revision also touched charset,filter,timber]
* EUC-TW implementation, plus an explanation of why ISO-2022-CN is difficult.Ben Harris2005-09-24
| | | | | [originally from svn r6353] [this svn revision also touched charset,filter,timber]
* Space-saving restructure of the CNS 11643 data tables. Reduces theSimon Tatham2005-09-24
| | | | | | | | | | | | | | | RO data size in cns11643.o from 400k to 240k. Relies on there being at most seven planes (7*94*94 <= 64k) and on the character set not encoding any Unicode code point above U+40000; if either of these becomes untrue later on we can always fall back to the previous approach, or to somewhere between that and here. The new version passes all the same tests as the old one did, and generates the same output under the new `cstable -v'. I'm confident that I haven't broken it. [originally from svn r6351] [this svn revision also touched charset,filter,timber]
* Fix a couple of warnings.Ben Harris2005-09-24
| | | | | [originally from svn r6350] [this svn revision also touched charset,filter,timber]
* Introduce the -v flag which outputs the actual index of each codeSimon Tatham2005-09-24
| | | | | | | point in every charset. [originally from svn r6349] [this svn revision also touched charset,filter,timber]
* Add support for CNS 11643.Ben Harris2005-09-24
| | | | | [originally from svn r6348] [this svn revision also touched charset,filter,timber]
* Include CNS 11643 in the cstable diagnostic utility.Simon Tatham2005-09-24
| | | | | [originally from svn r6347] [this svn revision also touched charset,filter,timber]
* IRG source T3 includes not only plane 3 of CNS 11643, but also "some additionalBen Harris2005-09-24
| | | | | | | characters". We now filter out the latter from our mapping table. [originally from svn r6345] [this svn revision also touched charset,filter,timber]
* CNS 11643 goes above the BMP, so the test code should take that intoSimon Tatham2005-09-24
| | | | | | | | account when checking the reverse mapping for every potentially relevant Unicode character. [originally from svn r6343] [this svn revision also touched charset,filter,timber]
* Add a mapping table for CNS 11643-1992. It's a bit big, and nothingBen Harris2005-09-24
| | | | | | | uses it yet. [originally from svn r6342] [this svn revision also touched charset,filter,timber]
* Support for the ESC $ ( 0 and ESC $ ( 1 sets that Emacs uses to embedBen Harris2005-09-21
| | | | | | | | Big5 in COMPOUND_TEXT. Emacs does lots of other rude things to COMPOUND_TEXT, but this one is supported by XLib as well. [originally from svn r6336] [this svn revision also touched charset,filter,timber]
* Add support for COMPOUND_TEXT extended segments encoding ISO 98859-14,Ben Harris2005-09-21
| | | | | | | ISO 8859-15, and BIG5. [originally from svn r6335] [this svn revision also touched charset,filter,timber]
* Add two new SBCSes: BS 4730 (alias UK-ASCII) and DEC graphics (alias VT100Ben Harris2005-09-18
| | | | | | | | line-drawing). I think this means that libcharset supports all the character sets that PuTTY supports, which is nice. [originally from svn r6330] [this svn revision also touched charset,filter,timber]
* When documenting s0 and s1, get then the right way around.Ben Harris2005-09-18
| | | | | [originally from svn r6329] [this svn revision also touched charset,filter,timber]
* 1: Better documentation of how read_iso2022() stores its state.Ben Harris2005-09-18
| | | | | | | | 2: Minimal write_iso2022(): it can't encode anything, but promises not to segfault. [originally from svn r6328] [this svn revision also touched charset,filter,timber]
* Ben points out that ESC ( J in ISO-2022-JP should encode theSimon Tatham2005-09-18
| | | | | | | | | _bottom_ half of JIS X 0201 (the one that's almost identical to ASCII, equivalent to the bottom half of Shift-JIS), not the top half. [originally from svn r6327] [this svn revision also touched charset,filter,timber]
* Make read_utf8(), like read_sbcs(), accessible to the rest of the library,Ben Harris2005-09-18
| | | | | | | so it can be used directly in iso2022.c. [originally from svn r6326] [this svn revision also touched charset,filter,timber]
* Undo another change that leaked through with the ISO-2022 commit.Ben Harris2005-09-18
| | | | | [originally from svn r6325] [this svn revision also touched charset,filter,timber]
* Update comment to reflect state of DOCS support.Ben Harris2005-09-18
| | | | | [originally from svn r6324] [this svn revision also touched charset,filter,timber]
* Undo accidental change in previous commit.Ben Harris2005-09-18
| | | | | [originally from svn r6323] [this svn revision also touched charset,filter,timber]
* Support for using DOCS to switch to and from UTF-8 mode.Ben Harris2005-09-17
| | | | | [originally from svn r6321] [this svn revision also touched charset,filter,timber]
* Reasonably complete ISO 2022 support. Huge and hairy, but it seems toBen Harris2005-09-17
| | | | | | | largely work. It might even be useful for something. [originally from svn r6320] [this svn revision also touched charset,filter,timber]
* Use standard "WILD" markers for unregistered Big 5 aliases.Ben Harris2005-09-17
| | | | | [originally from svn r6319] [this svn revision also touched charset,filter,timber]
* Fix stupid typo.Ben Harris2005-09-17
| | | | | [originally from svn r6318] [this svn revision also touched charset,filter,timber]
* Names for ASCII and JIS X 0201 that appear both in the X registry and inBen Harris2005-09-17
| | | | | | | the usual X fonts. [originally from svn r6317] [this svn revision also touched charset,filter,timber]
* TIS-620 is equivalent to ISO 8859-11, so map the MIME name for the former toBen Harris2005-09-17
| | | | | | | the latter. [originally from svn r6316] [this svn revision also touched charset,filter,timber]
* Substantial overhaul of the UTF-8 decoder. It now uses 26 bits of stateBen Harris2005-09-16
| | | | | | | | rather than 32, which might make it possible to use it inside another decoder. All the tests still pass. [originally from svn r6315] [this svn revision also touched charset,filter,timber]
* Bring utf8.c's internal tests up to date in the (somewhat belated)Simon Tatham2005-09-16
| | | | | | | | wake of r3713. [originally from svn r6314] [r3713 == 9392c7ecb6da984e144ce00dfb8c7cb725ad0a61 in timber repository] [this svn revision also touched charset,filter,timber]
* I've apparently had this todo-list comment sitting on stormhawk forSimon Tatham2005-09-13
| | | | | | | nearly a year and not checked it in. [originally from svn r6309] [this svn revision also touched charset,filter,timber]
* Ben Hutchings points out that `.UC', which I think I must haveSimon Tatham2005-07-29
| | | | | | | | | copied mindlessly from whatever man page I read when working out the format of man pages, has the function of putting `3rd Berkely Distribution' in one of the footer slots and therefore is probably inaccurate... [originally from svn r6146]
* The heading depth of a para_Subsect is p->aux+1, not p->aux+2.Ben Harris2005-05-18
| | | | | | Spotted (indirectly) by Damien Miller. [originally from svn r5802]
* input.c was capable of generating a paragraph structure with no textSimon Tatham2005-04-12
| | | | | | | | | | in it, if the input paragraph contained (say) an unrecognised control command and nothing else. Such paragraphs can confuse back ends later on, so input.c should refrain from generating them. Added a check and a polite error message (just in case the user manages to generate an empty paragraph using otherwise legal syntax). [originally from svn r5629]
* Also, while I'm wandering past here, I've just noticed thatSimon Tatham2005-03-10
| | | | | | | `anonfrag' should be a char[] rather than a char *, or else lenof() will give the wrong answer. [originally from svn r5477]
* Explicitly constify a bunch of static data declarations which wereSimon Tatham2005-03-10
| | | | | | | | | conceptually const but not declared as such. Halibut is now back to the practically-speaking-pointless but rather satisfying status of having no global writable data whatsoever :-) [originally from svn r5476] [this svn revision also touched charset,filter,timber]
* Ability to specify multiple arguments to \cfg{html-template-fragment};Jacob Nevins2005-03-08
| | | | | | | | | | | | | | | Halibut will output fragment names in all specified formats. (I forget now precisely why I thought this was necessary, but it seems potentially useful.) Also ensure that legal fragment names are generated even if none of the characters from the original turn out to be legal (e.g., %k with an entirely numeric keyword), and correct an untruth I inserted in the documentation of this. (This commit hits more than just the HTML backend as I've generalised an error message, and fixed a fault in the info backend's error handling while there.) [originally from svn r5457]
* Move contact details inside <address> to make validator.w3.org happy.Jacob Nevins2005-03-07
| | | | [originally from svn r5454]
* In contents lists, put <ul> inside <li> to make validator.w3.org happier (forJacob Nevins2005-03-07
| | | | | | | | | all HTML flavours). Also add newlines to the contents output so that the source is vaguely legible. Couple of extra comments in the code. [originally from svn r5453]
* Formatting fix if no indexJacob Nevins2005-02-18
| | | | [originally from svn r5350]
* Oops, overzealous search-and-replaceJacob Nevins2005-02-18
| | | | [originally from svn r5349]
generated by cgit v1.1 at 2026-05-21 09:49:17 +0000