| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
|
|
| |
back to Windows-1252 and hope for the best.
Note that this only occurs when autodetecting. If the client has specified a charset, they get told about it immediately and get to decide what to do about it. If a meta charset is encountered after a successful autodetection, the client is informed in the usual way. As this requires the client to throw out the parser and start afresh, specifying the meta charset as the one to use, there's no problem as, again, the client will be informed immediately in that case.
All of this ensures that charset autodetection is pretty well transparent as far as the client application is concerned.
svn path=/trunk/hubbub/; revision=5970
|
|
|
|
|
|
|
|
|
|
| |
states.
What happened was that, given \rabc, we would advance past the \r, then read at current_offset + len (len == 1). I.E. read 'b' instead of 'a'. If the data in the inputstream's internal buffer happened to end immediately after the \r, then we'd read past the end of the buffer thanks to a bug in lpu_inputstream_peek which was fixed in r5965. In any case, we'd still be looking at the wrong character when looking for CRLF pairs.
All regression tests now pass again.
svn path=/trunk/hubbub/; revision=5967
|
|
|
|
|
|
|
| |
Drop HUBBUB_OOD and just use HUBBUB_NEEDDATA, instead.
Currently aborts in bogus comment handling if it encounters a \r at the end of the inputstream's utf-8 buffer.
svn path=/trunk/hubbub/; revision=5966
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5851
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5846
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5845
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5844
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5843
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5668
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5667
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5666
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5665
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5664
|
|
|
|
|
|
| |
Fix commentary copied from libcss
svn path=/trunk/hubbub/; revision=5663
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5656
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5575
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5431
|
|
|
|
|
|
|
|
|
| |
treebuilder (this can occur if EOF is encountered when in generic_rcdata/script_collect modes).
Note that the spec has since changed here, so we'll need to resync, anyway.
This is simply to ensure the testsuite passes once more.
svn path=/trunk/hubbub/; revision=5374
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5290
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5285
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5277
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5275
|
|
|
|
|
|
|
| |
Pedantic whitespace changes.
More paranoia surrounding entity handling.
svn path=/trunk/hubbub/; revision=5266
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5262
|
|
|
|
|
|
|
|
| |
testing whether the 4 most recently read characters in the data state are <!--.
Add a couple of assertions for paranoia.
svn path=/trunk/hubbub/; revision=5146
|
|
|
|
|
|
| |
mode, and then remove an unused variable in the treebuilder state.
svn path=/trunk/hubbub/; revision=5132
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5125
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5108
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5107
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5106
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5104
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5099
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5098
|
|
|
|
|
|
| |
store their own length and when emitting the comment, the buffer contains the whole comment and nothing else.
svn path=/trunk/hubbub/; revision=5095
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5093
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5086
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5082
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5080
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5030
|
|
|
|
|
|
|
|
|
| |
sizeof(x[0]) everywhere
- In anticipation of interning tag names, make the big map of name->type store them in lowercase
- Take advantage of the previous change to optimise away many calls to strlen
svn path=/trunk/hubbub/; revision=5010
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5009
|
|
|
|
|
|
|
| |
- Remove the no-op FINISH() macro from the tokeniser
- Fix a typo in the charset detector
svn path=/trunk/hubbub/; revision=5007
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5003
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=5000
|
|
|
|
|
|
| |
mibenum value.
svn path=/trunk/hubbub/; revision=4992
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=4991
|
|
|
|
|
|
| |
propagated properly.
svn path=/trunk/hubbub/; revision=4990
|
|
|
|
|
|
| |
allow for encoding change info to be returned more easily.
svn path=/trunk/hubbub/; revision=4989
|
|
|
|
|
|
| |
handle_in_head(). Remove prototypes for functions no longer existent.
svn path=/trunk/hubbub/; revision=4987
|
|
|
|
| |
svn path=/trunk/hubbub/; revision=4980
|