about summary refs log tree commit diff stats
Commit message (Collapse)AuthorAgeFilesLines
...
* Slightly refactor commandModebptato2023-01-064-12/+27
|
* layout/engine: fix non-inheritable values for inline boxbptato2023-01-063-40/+75
| | | | | Also, a hack for inline box background-color. It doesn't work very well, but good enough for now.
* twtstr: fix oversightbptato2023-01-051-6/+1
|
* display/client: quit when pager.buffer is nullbptato2023-01-051-0/+3
|
* data/idna, utils/twtstr: clean up, fix bugsbptato2023-01-052-58/+63
| | | | | | | Looks like lazily loading the idna table doesn't work, so now it's statically loaded. Also, bugfixes in idna handling.
* socketstream: remove todobptato2023-01-051-1/+0
|
* layout/engine: fix pre-line wrap on newlinebptato2023-01-051-1/+6
| | | | Just how many white-space properties can I get wrong?
* dom: fix object conversion bug in optionsbptato2023-01-051-1/+1
|
* javascript: remove unused pointer case from getTypePtrbptato2023-01-051-1/+1
| | | | I probably removed the last call of getTypePtr with a pointer long ago.
* javascript: refactor toJSbptato2023-01-051-78/+74
| | | | Split up the huge generic toJS function.
* Get rid of eprint in place of stderr.writebptato2023-01-052-3/+3
| | | | eprint should only be used for debugging.
* Add ask()bptato2023-01-055-50/+95
|
* promise: set get, cb, next to nil after then is calledbptato2023-01-041-7/+5
| | | | | | Actually we don't even need a state variable, just set cb to nil when resolving. (I'm sure this is a great idea that will absolutely not backfire in the future.)
* pager: fix prevBuffer, nextBuffer, replace weirdnessbptato2023-01-041-8/+16
| | | | Now they should behave as expected. (I hope.)
* ua.css: fix color, css/values: fix globalsbptato2023-01-042-7/+14
|
* client, pager, dom, ...: better error handlingbptato2023-01-045-96/+121
| | | | Now the browser shouldn't completely die when a buffer crashes.
* renderdocument: add StyledNodes to backgroundsbptato2023-01-032-27/+21
| | | | | | This makes it possible to e.g. click on this: <a href=x style="display: inline-block; width: 5em; height: 5em; background-color: red"></a>
* css/values: consistently use exceptionsbptato2023-01-031-56/+41
| | | | | Options are great, but only if used consistently. This also fixes some weird bugs where exceptions were handled incorrectly.
* Update readmebptato2023-01-031-1/+1
|
* regex: fix setting start in unicode stringsbptato2023-01-031-1/+1
| | | | | We have to set it to (capture - cstr) >> 1, because libregexp later shifts it left.
* ua.css: add counter-reset to ulbptato2023-01-031-0/+1
|
* layout/engine: fix inline-table crashbptato2023-01-031-16/+19
| | | | inline-table was calling the wrong layout, which made an assertion fail.
* Update readmebptato2023-01-031-19/+27
|
* css/values: remove todobptato2023-01-031-1/+0
|
* buffer: fix infinite loop in getClickablebptato2023-01-031-2/+3
|
* buffer: fix infinite loopbptato2023-01-031-1/+2
|
* renderdocument: fix crash for image textbptato2023-01-031-1/+2
| | | | We need to check if y >= 0, or it fails.
* ua.css: only highlight a[href]bptato2023-01-021-2/+2
|
* container, pager: work around bad codegenbptato2023-01-022-10/+10
|
* dom: add better attribute reflectionbptato2023-01-026-121/+267
| | | | | Instead of creating a new function for each attribute, use a single magic function for reflected attributes.
* dom: support a few more getters/settersbptato2023-01-021-8/+47
|
* Add support for <label>bptato2023-01-022-94/+164
|
* term: hack to avoid a weird crashbptato2023-01-021-2/+8
| | | | Looks like we can't just assign canvas to pcanvas.
* Fix connection load info not being shownbptato2023-01-023-4/+7
|
* Update readmebptato2023-01-011-3/+4
|
* selectorparser: support :link, :visitedbptato2023-01-014-7/+22
| | | | I say "support", but :visited is never matched.
* convert_size: truncate resultbptato2023-01-011-13/+18
|
* loader: proper redirect handlingbptato2023-01-015-12/+24
|
* buffer: fix some search bugs & refactor regex stuffbptato2023-01-016-100/+100
| | | | | | | | cursorBytes uses twidth now. cursorNextMatch matches the byte *after* the cursor (somewhat more consistently than before). match() no longer counts capture groups. LRE_FLAG_GLOBAL now goes through the entire string.
* pager: fix bug with tab on screen shiftingbptato2023-01-011-2/+3
|
* layout/engine: fix text-align: cha-* for boxes with shrinkbptato2023-01-011-15/+22
| | | | | In shrink, we can only take up as much space as our content, so we first have to determine our content width and only then align the children.
* css/values: apply presentational hints after UA stylebptato2023-01-011-4/+5
| | | | | So that e.g. table valign overrides the user-agent vertical-align value, but still respects user/author style.
* cascade: document a bitbptato2023-01-011-2/+7
|
* rendertext: fix format position for non-asciibptato2023-01-011-1/+5
|
* makefile: separate objdir for targets, uninstallbptato2023-01-011-6/+11
|
* Update makefilebptato2023-01-012-8/+21
|
* posixstream: fix incorrect read() usagebptato2023-01-011-1/+1
| | | | It was overwriting our buffer, instead of appending to it...
* javascript: more robust toJSNumberbptato2022-12-311-20/+15
|
* layout/engine: fix crash on padding-rightbptato2022-12-312-2/+1
|
* javascript: fix incorrect error() statementbptato2022-12-311-1/+1
|