about summary refs log tree commit diff stats
path: root/adapter
Commit message (Collapse)AuthorAgeFilesLines
* lcgi: support HTTP proxybptato2024-10-131-15/+52
| | | | | | | not quite sure this approach is right, because it still won't work with HTTPS without linking everything with OpenSSL... maybe a socks5 bridge through a UNIX socket would work better?
* lcgi_ssl: fix libressl compatbptato2024-10-132-6/+8
| | | | | * use correct headers for types (libressl doesn't have types.h) * use SSL_get_peer_certificate (libressl doesn't have a get0 variant)
* Add `nc' tool & adjust finger/spartan to use itbptato2024-10-124-57/+82
| | | | | | | | | Simple netcat clone, useful for portable scripts. Especially because some netcats will close the connection as soon as I close stdin... this one only quits when either stdout or the socket refuses new data. Also, it uses our standard TCP connection routine, meaning it respects ALL_PROXY. (i.e. now spartan works with socks5 too)
* dirlist: separate out into dirlist2htmlbptato2024-10-126-275/+224
| | | | | Now the dirlist parser lives inside dirlist2html, and the file protocol just emulates FTP LIST.
* gmifetch: rewrite in Nimbptato2024-10-124-658/+491
| | | | | | | | | | 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.
* layout, pager: preserve tabs on display, selection & outputbptato2024-10-101-0/+8
| | | | | | | | | | | | Substitute tabs with one of eight PUA characters based on their width, and convert them back in the pager: * TUI mode always prints spaces, but now handles tabs appropriately on cursor movement * dump mode tries to preserve hard tabs, but uses soft tabs when that is not possible (e.g. tabs after a margin, tab with background color, etc) * selection mode always outputs hard tabs.
* Fix compilation on Nim 2.2.0bptato2024-10-101-1/+1
|
* resize: use srgbbptato2024-10-091-5/+5
| | | | | | | | since that's our input color space (Actually, mainstream browsers seem to use linear scaling nevertheless. But apparently that produces surprising results in some cases, and stbir can do srgb, so might as well use it.)
* color: reduce CellColor size, misc color refactoringbptato2024-10-063-12/+9
| | | | | | * split out CSSColor from CellColor; now CellColor is just 4 bytes (which helps reduce FormatCell size) * unify color function naming (still not perfect)
* man: check exit code after reading first linebptato2024-10-061-8/+14
|
* sixel: simplify compressSixelbptato2024-10-051-12/+5
|
* md2html: support + lists, ~~~ blocksbptato2024-10-041-5/+7
|
* dynstream: fix memory leakbptato2024-10-043-4/+4
| | | | now I know why overloading dealloc felt wrong
* sixel, term: reduce half-dump special casingbptato2024-10-041-62/+46
| | | | | | | Makes it slightly easier to debug image output. Also, we stop sending dimension headers, and no longer check for the scheme env var to make CLI invocation a bit less annoying.
* sixel: fix some weird artifactsbptato2024-10-041-0/+13
| | | | | | | | | | | | Since getColor's color assignment eagerly fills in holes with the nearest neighbor of the first such observed node, it is prone to falling into a local minimum where dithering ends up switching between a few very wrong colors. Ensure this doesn't happen by always allocating at least 4 nodes of the octree branch. (Allocating all 7 seems to result in a heavy performance hit; just 4 means at worst 4096 extra nodes, but likely a lot less, and I don't see a difference in either output or performance.)
* sixel: transparency improvementsbptato2024-10-031-5/+5
| | | | | | | * don't set transparency when raster attributes suffice - it seems terminals don't background-fill in that case either. * fix transparency in encoder standalone mode * update comments
* http: remove twtstr dependencybptato2024-10-011-4/+10
|
* ftp: remove libcurl dependencybptato2024-10-013-248/+464
| | | | | | | | | | | | | This splits out sftp into a separate binary that *does* depend on libcurl. However, ftp now uses the same socket code as gopher. ftps is dropped, because I've never even tested it. Maybe I'll add it back when we have working OpenSSL bindings. This is still "doing the easy part first", now I have no clue how to handle sftp because my initial plan ("just use the sftp binary") doesn't work - sftp batch mode doesn't accept passwords. libssh2 remains the sole candidate, but that's what libcurl wraps anyway.
* lcgi: fix broken error messagebptato2024-09-281-1/+1
|
* loader: send headers for tocache requests toobptato2024-09-281-2/+4
| | | | | This lets us send the transparency bit as a header, and also halves the number of header parsers in loader.
* gopher: do not depend on libcurlbptato2024-09-282-74/+202
| | | | | | | | I'm thinking of making libcurl entirely optional; let's start with the easiest part. I've added a SOCKS5 client for ALL_PROXY support; I know curl supported others too, but whatever.
* loader: clean up connecterrorbptato2024-09-285-33/+22
| | | | | | * allow string values for public errors * remove unused errors * update naming
* sixel: reset chunk on transparencybptato2024-09-271-0/+1
| | | | this was causing weird artifacts
* term: prevent negative line damage, fix off by onebptato2024-09-251-2/+6
|
* sixel: support transparencybptato2024-09-241-33/+38
| | | | | | | | | | | | | | | | | | 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.
* Replace std/selectors with pollbptato2024-09-231-19/+16
| | | | | | | | | | | | std/selectors uses OS-specific selector APIs, which sounds good in theory (faster than poll!), but sucks for portability in practice. Sure, you can fix portability bugs, but who knows how many there are on untested platforms... poll is standard, so if it works on one computer it should work on all other ones. (I hope.) As a bonus, I rewrote the timeout API for poll, which incidentally fixes setTimeout across forks. Also, SIGWINCH should now work on all platforms (as we self-pipe instead of signalfd/kqueue magic).
* loader: mmap intermediate image files, misc refactoringbptato2024-09-227-74/+88
| | | | | | | | | | | | | | | | | | | | | | | | | | | * refactor parseHeader * optimize response blob() * add direct "to cache" mode for loader requests which sets stdout to a file, and use it for image processing * move image resizing into a separate process * mmap cache files in between processing steps when possible At last, resize is no longer a part of image decoding. Also, it feels much nicer to keep encoded image data in the same cache as everything else. The mmap operations *should* be more efficient than copying the whole RGBA data through a pipe. In practice, it only makes a difference for loading (well, now just mmapping) the encoded image into the pager, where it singlehandedly speeds up image display by 10x on my test image. For the other steps, the unfortunate fact that "tocache" must delay the next fork/exec in the pipeline until the entire image is processed seems to equal out any wins we might have gotten from skipping a single raw RGBA copy. I have tried moving the delay before the exec (it's possible with yet another pipe), but it didn't help much and made the code much uglier. (Not that tocache didn't, but I can live with this...)
* Refactor img/*bptato2024-09-151-5/+225
| | | | | I've moved most image logic to adapter, so it doesn't really make sense to have this subdir anymore.
* sixel: do not reserve palette entry for transparencybptato2024-09-121-8/+11
| | | | | | | | | Turns out this isn't actually needed. Which makes sense, as transparency doesn't have a color register at all - it's just the default state of pixels. Also, skip octree-based quantization with palette <= 2; unsurprisingly, monochrome gives much better results.
* sixel: factor out root node, fix a bugbptato2024-09-121-30/+31
|
* sixel: small optimizationsbptato2024-09-111-31/+48
| | | | | | | | * Reduce unnecessary branch allocation even more * Use linked lists to represent sixel bands Also inlined octree-based getColor for clarity, although the compiler was already doing that.
* sixel: faster quantizationbptato2024-09-111-80/+92
| | | | | | | | | | | Reduce eager allocation of whole branches for leaves; now it's done only after the first conflict (or repeated color). This makes quantization ~2x faster for images without large runs of the same color and ~2x slower for images with a lot of those. No noticeable loss in quality. I also disabled refc for the octree and turned the nodes into tagged unions, which again resulted in a ~2x speedup.
* sixel: misc optimizationsbptato2024-09-091-42/+43
|
* sixel: small optimizationbptato2024-09-081-29/+27
| | | | | | For large palettes, looping over the entire chunk list is quite slow. Store the list of active chunks in a separate sequence so we can skip the unused ones.
* md2html: code, pre, inline fixesbptato2024-09-071-120/+193
|
* term: sixel sizing & output fixesbptato2024-09-051-7/+7
| | | | | | | | | | | | * round down to number divisible by 6 for height * make pager's dispw match term's dispw even after width clamping * make *BE procs actually emit/consume big-endian (lol) * fix borked sixel set raster attributes & control string I mixed up SRA with the device control string's parameters, so instead of toggling transparency in the DCS, I was setting the second SRA parameter to 0. Which, by the way, defines the aspect ratio's denominator, and has nothing to do with transparency. Whoops.
* sixel, stbi, sandbox: fix fstat sandbox violationbptato2024-09-042-50/+44
| | | | | | | Until recently, glibc used to implement it as fstatat. So don't trap for fstatat (and for consistency, fstat), but return EPERM. Just to be sure, rewrite sixel & stbi to never call fread.
* md2html: misc inline fixesbptato2024-09-011-114/+160
| | | | | * fix whitespace-surrounded asterisk turning into emphasis * slightly refactor parseInline (it's still quite ugly...)
* sixel: special case search for few color registersbptato2024-09-011-17/+58
| | | | | | See comment for details. (Also, set the palette size based on quality for toBlob.)
* canvas: make sure we don't link to QJSbptato2024-09-011-2/+1
|
* canvas: move to separate CGI scriptbptato2024-09-013-3/+148
| | | | | | | | | | * stream: and passFd is now client-based, and accessible for buffers * Bitmap's width & height is now int, not uint64 * no more non-network Bitmap special case in the pager for canvas I just shoehorned it into the static image model, so it still doesn't render changes after page load. But at least now it doesn't crash the browser.
* sixel: fix oobbptato2024-08-301-3/+2
| | | | whoops
* term, sixel: misc image fixes, sixel optimizationbptato2024-08-301-55/+97
| | | | | | | | * expand allowed color range somewhat * update maximum sixel size on window resize * fix kitty image cropping * use faster algorithm for sixel compression (also produces less wasteful output)
* md2html: support blockquotebptato2024-08-302-10/+28
| | | | + update todo, readme
* sixel: ditheringbptato2024-08-291-42/+73
| | | | | | | | | | | | | uses floyd-steinberg - I originally tried atkinson, but liked the results with fs a bit more. also, I got rid of the half hearted attempt to skip tree traversals for color lookups, as it performs horribly with dithering. it *did* work if I set Y as the key, but it still felt wrong - well, as it turns out, octree traversal is faster anyway, so just do that. it works out nicely because we can fill in holes (from dithering) with a linear search for the nearest match as we go.
* sixel: fix borked approximation schemebptato2024-08-291-41/+29
| | | | | | | | | | | ok, of course if you search for the nearest entry after setting blue as the lowest index bits the search will be biased towards blue... let's just skip the trouble of reconstructing closest color relations by storing hashes in the octree - this way, I can ensure that the final merged colors are present in the buckets that getColor then looks at (while we're at it, also fix the color run merging code in quantize.)
* sixel: minor optimizationsbptato2024-08-281-51/+86
| | | | | | * don't emit transparency request when we don't need it * add lookup table for missing hash entries * write(2)'ize
* sixel, term: fix off-by-1's in croppingbptato2024-08-281-2/+6
| | | | & clean up outputSixelImage in general
* sixel: proper color quantizationbptato2024-08-271-37/+215
| | | | | | | | just use an octree. works fine afaict, though obviously somewhat slower than the static method (encoding is 2-pass now) & still has banding issues with many colors (will need dithering) also, fixed a bug that caused initial masks of bands to get misplaced
* stbi, jebp: use read/write instead of fread/fwritebptato2024-08-263-61/+131
| | | | | | | | glibc likes to do weird things (such as calling stat) when you use fread(3) and friends, so try to use functions that are more likely to just do a single syscall. Also, copy over some more paranoid read/write procedures to http.
generated by cgit-pink 1.4.1-2-gfad0 (git 2.36.2.497.gbbea4dcf42) at 2025-08-16 03:40:16 +0000