| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
[[siteconf]] now just means [siteconf.0], etc. So you can now override
parts of default siteconfs/omnirules, e.g. to change the Google search
substitute-url, etc.
To celebrate this, I've added some more default search engines:
* wk: -> Wikipedia
* wd: -> Wiktionary
* ms: -> Marginalia Search
These can be replaced by setting e.g. omnirule.wk = {}, etc.
Also, siteconf = {} can be used to clear pre-defined siteconfs.
This is an unfortunate deviation in semantics from TOML, but in practice
the way it worked before didn't match the spec either, so at least it
is now consistent.
|
| | |
|
| |
|
|
| |
It's surprisingly tricky.
|
| |
|
|
| |
also, fix a typo
|
| | |
|
| |
|
|
|
|
|
|
| |
Even if we don't do some z-ordering correctly, it's no excuse to paint
boxes incorrectly.
It's also extremely annoying when I try to use a menu and text bleeds
into the drop-down window.
|
| |
|
|
|
|
|
| |
I'm not a fan, because it hides bugs. But working around the overflow
errors is starting to get unwieldy.
On 32-bit systems, we try to use compiler intrinsics as Nim does.
|
| |
|
|
|
| |
sticky, ruby, writing-mode: lots of complexity for little gain, and the
fallback works just as well (if not better)
|
| | |
|
| |
|
|
|
| |
It's still missing a "color visited links" feature, but it's better
than nothing.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Overflow pretty much requires scrollbars, but those wouldn't work in
dump mode, plus of course they would be a pain to implement. So as a
simple alternative:
* overflow: hidden, clip works as per spec.
* overflow: auto, overlay, scroll invert the intrinsic minimum size
clamping logic instead of adding a scrollbar.
What this concretely means, is that this
<pre style="overflow: scroll; height: 1em">
test
test
test
</pre>
will, instead of creating a scroll container, just override the
specified height.
This hack works surprisingly well, because CSS pretty much requires
setting height on scroll containers, so authors aren't incentivized to
set height on the parent container too (because the contents are already
sized appropriately).
One issue left is how to deal with overflow: hidden ancestors. For now,
I've made it so that it can spill for overflow-x, and always clips
on overflow-y, because it's much less likely to bleed into other
text horizontally than vertically. But there is definitely room
for improvement, e.g. we could track space requested by scrolling
children and expand parent boxes based on that.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Necessary for flex. Previously we just used the actual height, but that
didn't account for boxes that size themselves depending on the available
height (mainly just images for now).
This also irons out intrinsic min width calculation somewhat, squashing
several bugs.
I hope it works well. It is a significant change in size calculation,
so maybe there are still new bugs lurking.
|
| |
|
|
|
|
|
|
|
|
| |
Finally it's done. It's basically w3mbookmark, but using Markdown
instead of HTML and in POSIX shell instead of C.
As a bonus, it can also (sort of) import w3mbookmark's output. Well,
at least it worked on my bookmark file, but there is a known issue with
bracket escaping... if it goes wrong, it's simple enough to edit it
manually :P
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In the past, Chawan would read global mailcap (/etc/mailcap, ...) too,
but every now and then that would run entries that I didn't even know
existed and definitely didn't intend to run. So I changed it to only
use ~/.mailcap, but this meant users now had to add mailcap entries for
every single mime type.
At some point I also changed application/octet-stream to always save to
disk, which is usually nice except when a text file is misrecognized as
binary. Often times I just want to decide myself what to do.
So now there are two layers. First, the global mailcap files (path as
per RFC) prompt before executing. Then there is ~/.chawan/auto.mailcap
(or ~/.config/chawan/auto.mailcap) which runs entries automatically.
If you press shift before selecting an option in the prompt, the
corresponding entry gets copied to auto.mailcap. It's also possible to
type a new entry on the fly. Overall I think it's quite convenient.
One unfortunate side effect is that existing users will have to migrate
their entries to auto.mailcap, or redefine external.auto-mailcap to e.g.
~/.mailcap, but this seems acceptable.
|
| |
|
|
|
| |
also, fix a bug in the chapath parser so that param expansion actually
works
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Before, the initial layout of a flex item would have inherited the
parent's sizing constraint. This almost worked, except when a descendant
of a flex item with an unspecified width would resolve its percentage
against this incorrectly inherited width - per standard, this must
resolve to auto.
Also, the shrink case was wrong, because it did not scale the unit to
the respective widths; the standard mandates this as well. Hopefully I
got it right this time.
Finally, this fixes positioned inline container blocks not being set
as the absolute container.
|
| | |
|
| |
|
|
|
|
| |
also fixes a bug in the previous commit where whitespacenum would be
reset on absolute positioning + missing absolute margin handling without
top/left/etc.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This fixes a) resolution of `position: absolute' percentage sizes with
an indefinite containing block size (wait what?), and b) positioning of
`position: absolute' inner block boxes inside inline boxes.
a) is possible because `position: absolute' does not affect its
parent's layout. I would love to have a long talk with whoever
thought that specifying it like this is a good idea. You know, just
because you can... anyways, shockingly enough, this was still the more
straightforward part.
b) forced me to change the box tree to allow blocks inside inlines,
which is ugly, but no big deal. Two questions then remained:
1. *where* to put such boxes, and
2. *how large* these boxes should be; this is hardly obvious since
an inline box does not have an unambiguous width or height.
Of course the CSS standard, never too eager to fulfill my basic
expectations, says nothing about this (other than "it's not defined").
So I first tried to reverse engineer what Gecko does, and after hours of
pain and suffering I realized... that it's broken LOL
https://bugzilla.mozilla.org/show_bug.cgi?id=489100
Therefore I ended up (partially) copying Blink behavior, which, while
equally nonsensical as Gecko (and of course divergent), at least does
not change after a partial layout. Thank you LayoutNG for saving my
sanity.
As for the practical benefits: this fixes the bug where Invidious
[video] tags wouldn't show up. Hey, it's something!
Still left to-do: `position: absolute' for `display: inline' boxes.
|
| |
|
|
|
|
|
|
|
|
| |
This finally makes it possible to use socks5 for Gemini.
Also slightly refactored the config, to make it easier to pass on the
config dir.
By the way, the known_hosts file is now stored in the config dir too.
The adapter will try to move it to there from the old location.
|
| | |
|
| |
|
|
| |
it's a waste of space, and doesn't work well with showFullAlert
|
| |
|
|
| |
ensure that images are shown in the order buffer sent them
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Sixel can only represent transparency for fully transparent (alpha
= 0) and fully opaque (alpha = 255) pixels, i.e. we would have to
do blending ourselves to do this "properly". But what do you even
blend? Background color? Images? Clearly you can't do text...
So instead of going down the blending route, we now just approximate
the 8-bit channel with Sixel's 1-bit channel and then patch it up with
dither. It does look a bit weird, but it's not *that* bad, especially
compared to the previous strategy of "blend with some color which
hopefully happens to be the background color" (it rarely was).
Note that this requires us to handle transparent images specially
in term. That is, for opaque ones, we can leave out the "clear cells
affected by image" part, but for transparent ones, we must clear the
entire image every time.
|
| |
|
|
|
|
|
|
|
|
|
| |
* align status truncating behavior with w3m (not exactly, clipping
is still different, but this should be fine for now)
* add "su" for "show last alert"
- w3m's solution here is to scroll one char at a time with
"u", but that's extremely annoying to use. We already have a
line editor that can navigate lines, so reuse that instead.
* fix peekCursor showing empty text
* update todo
|
| |
|
|
| |
+ update todo, readme
|
| |
|
|
|
|
|
| |
in fact, we consider these to be the same image, and arguably that's
a bug in and of itself. but at least we don't crash anymore.
(also, update todo)
|
| | |
|
| |
|
|
|
| |
Still not perfect, because it crashes on missing /tmp dir so you have to
manually set it...
|
| |
|
|
|
|
| |
We were not setting the invalid flag on bitmap load, so any incremental
reshape could interfere with displaying images that got loaded after
the reshape.
|
| |
|
|
| |
and enable it by default.
|
| |
|
|
|
| |
* query default ANSI colors with OSC 4
* disable queries overridden by config
|
| |
|
|
|
|
|
| |
Mainly things you could already set with [[siteconf]] but not normally.
Also, a `styling' option to disable author styles.
Also, `images' is now documented as an "experimental" option, since it's
halfway usable now.
|
| |
|
|
|
|
|
| |
* merge select into container
* avoid unnecessary redraws in draw() for parts of the screen that
haven't been updated
* various image redraw fixes
|
| |
|
|
|
|
| |
* basic repaint algorithm for sixel (instead of brute force "clear the
whole screen")
* do not re-send kitty images already on the screen
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* multi-processed and sandboxed PNG decoding & encoding (through local
CGI)
* improved request body passing (including support for output id as
response body)
* simplified & faster blob()/text() - now every request starts
suspended, and OngoingData.buf has been replaced with loader's
buffering capability
* image caching: we no longer pull bitmaps from the container after
every single getLines call
Next steps: replace our bespoke PNG decoder with something more usable,
add other decoders, and make them stream.
|
| |
|
|
|
|
|
|
|
| |
If the total specified column width is larger than the table's allowed
width, we now resize them proportionally to the specified width.
This is quite important because many tables set the width to e.g. 50%
for "give me half of the table", instead of its true meaning "give me
half of the page".
|
| |
|
|
|
|
|
|
| |
* add CSSStyleDeclaration setter
* move ident maps directly into enums
* more complete CSSComputedValue stringifier
* turn canvas into a pseudo-image in cascade
* set canvas to inline-block
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* png: add missing filters, various decoder fixes
* term: fix kitty response interpretation, add support for kitty image
detection
* buffer, pager: initial image display support
Emphasis on "initial"; it only "works" with kitty output and PNG input.
Also, it's excruciatingly slow, and repaints images way too often.
Left undocumented intentionally it for now, until it actually becomes
useful. In the meantime, adventurous users can find out themselves why:
[[siteconf]]
url = "https://.*"
images = true
|
| |
|
|
| |
(Sadly some layout tests still fail.)
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It's a bad idea for several reasons:
* it's inefficient; must allocate an environment for a closure in Nim,
even though we already have one in JS
* writing macros for automatically creating functions with variadic
arguments is suprisingly difficult (see the entire `js/javascript'
module)
* it never really worked properly, because we never freed the associated
function pointer.
We hardly used it anyway, so the easiest fix is to get rid of it
completely.
|
| | |
|
| |
|
|
|
|
|
|
| |
it's better than nothing. I suppose.
(Two-value flex syntax is encouraged even by the standard, so it gets
used a lot, and that sets 0, not flex-basis: auto, so not having
flex-basis breaks too many things.)
|
| |
|
|
|
|
|
|
|
| |
Still far from being fully standards-compliant, or even complete, but it
seems to work slightly less horribly than having no flexbox support at
all on sites that do use it.
(Also includes various refactorings in layout to make it possible at all
to add flexbox.)
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Depending on Perl just for this is silly.
Now we use libregexp for filtering basically the same things as
w3mman2html did. This required another patch to QuickJS to avoid
pulling in the entire JS engine, but in return, we can now run regexes
without a dummy JS context global variable.
Also, man.nim now tries to find a man command on the system even if it's
not in /usr/bin/man.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
The previous retrofitting of the old renderdocument error correction
usually worked, but it still had a horrible flaw in that it assumed that
all line boxes are of equal height. So if error was lower for some line
than another, it would move *all* lines by a somewhat lower error, and
that resulted in overlapping lines.
Now we do something much simpler: in flushLine, round each line's height
downwards before moving on to the next line. This gets rid of any blanks
inbetween lines, and also works much better with cleared floats.
|
| | |
|
| | |
|