about summary refs log tree commit diff stats
path: root/src/loader
Commit message (Collapse)AuthorAgeFilesLines
...
* Remove CLONE BufferSource; cache document sources in tmpdirbptato2024-02-126-89/+197
| | | | | | | | | | | | | | | | | | | | | | | At last all BufferSources are unified. To achieve the same effect as the previous CLONE source type, we now use the "fromcache" flag in Request. This *forces* the document to be streamed from the disk; if the file no longer exists for some reason, an error is returned (i.e. the document is not re-downloaded). For a document to be cached, it has to be the main document of the buffer (i.e. no additional resources requested with fetch()), and also not an x-htmloutput HTML file (for those, the original source is saved). The result is that toggleSource now always returns the actual source for e.g. markdown files, not the HTML-transformed version. Also, it is now possible to view the source of a document that is still being downloaded. buffer.sstream has almost been eliminated; it still exists, but only as a pseudo-buffer to interface with EncoderStream and DecoderStream. It no longer holds the entire source of a buffer at any point, and is cleared as soon as the buffer is completely loaded.
* simplify newLoaderBufferbptato2024-02-111-4/+3
|
* loader: significantly more efficient loadingbptato2024-02-112-98/+137
| | | | | | | | | | The previous version was running the processor on 100% because select would immediately return for writes even when no buffers to send were available. (This has been the case since I added asynchronous sending, but the previous commit put the console buffer's fd in loader too and that made the problem quite obvious.)
* Get rid of LOAD_PIPE BufferSourcebptato2024-02-112-40/+78
| | | | | Instead, use a stream: scheme and associate hostnames with file descriptors directly from the pager.
* loader: add rejectHandlebptato2024-02-101-8/+8
|
* loader: fix teebptato2024-02-103-122/+225
| | | | | | | | | | | | My eyes are bleeding, but at least there is a chance that this does what I wanted. The previous tee implementation mixed buffer and loader fds, so it was fundamentally broken. Also, it used MultiStream which makes asynchronous streaming impossible. This time we use a flat array of output handles and link to them any buffers not written to the target yet.
* loader: fixes & cleanupbptato2024-02-102-70/+43
| | | | | | | | | | | * LoaderHandle.fd is no more, we now check ostream's fd * setBlocking converted to a PosixStream method * SocketStream now sets fd variable * handle sostream/fd redirection properly * fix suspend/resume This fixes non-HTML resource loading, mostly. However, tee is still broken :/
* loader: use recvData instead of readDatabptato2024-02-091-16/+15
| | | | | | | | recvData is a new method for PosixStream that does less weird magic than readData. Also, allow duplicates in unregWrite/unregRead; it's simpler to live with them than to prevent them.
* cgi: do not register closed istreamsbptato2024-02-091-1/+2
|
* loader: fix (another) double-close bugbptato2024-02-081-4/+8
|
* loader: fix memory leakbptato2024-02-081-1/+3
|
* loader: fix double-close bugsbptato2024-02-081-3/+3
|
* loader: fix eagain in fetch, only add URL to handle in debugbptato2024-02-082-7/+12
| | | | | | * eagain was causing fetch to add unnecessary null bytes to input streams * URL is now only added to handles in debug mode
* Incremental renderingbptato2024-02-072-51/+152
| | | | | | | | | | | | Yay! Admittedly, it is not very useful in its current form, except maybe on very slow networks. The problem is that renderDocument is *slow*, so we only run it when onload fails to consume all bytes from the network in a single pass. Even then, we are guaranteed to get a FOUC, since CSS is only downloaded in finishLoad(). Well, I think it's cool, anyway.
* loader: clean up error handlingbptato2024-01-263-74/+54
| | | | | | | | | * remove pointless exception -> bool conversions; usually they were ignored anyway + exceptions are more convenient here * add EPIPE handler to raisePosixIOError * fix socketstream to use raisePosixIOError * fix socketstream sendFileHandle error handling * cgi: immediately return on file not found error
* cgi: fix broken libexec path env variablebptato2024-01-262-7/+7
|
* js: allow specifying static function name, small refactoringbptato2024-01-241-4/+1
| | | | | | | | * static function names can now be defined using the syntax `Class:functionName' (or just use `Class' to take the default name * fix URL.canParse with 1 argument only * do not store JSFuncGenerator for constructors; just put the function node in BoundFunctions
* js: merge some type modules into jstypesbptato2024-01-111-1/+1
| | | | They only had type definitions, no need to put them in separate modules.
* Add urlenc, urldec; fix a URL encoding bug; improve trans.cgibptato2024-01-081-0/+2
| | | | | | | | | | * Fix incorrect internal definition of the fragment percent-encode set * urlenc, urldec: these are simple utility programs mainly for use with shell local CGI scripts. (Sadly the printf + xargs solution is not portable.) * Pass libexec directory as an env var to local CGI scripts * Update trans.cgi to use urldec and add an example for combining it with selections
* Use std/* imports everywherebptato2024-01-075-14/+14
|
* request: don't deny BodyInit that is not an objectbptato2024-01-061-3/+0
| | | | This breaks string conversions.
* loader: send error message str for disallowed URLbptato2023-12-291-3/+3
| | | | buffer was crashing with an EOFError otherwise
* Compile with styleCheck:usagesbptato2023-12-283-17/+17
| | | | much better
* dom: use JS_EvalFunction; add module fetching stubsbptato2023-12-251-1/+3
| | | | (still no module support in buffer...)
* cgi: return ConnectionError when script is not executablebptato2023-12-212-1/+3
|
* loader: refactor onRead/onErrorbptato2023-12-211-19/+22
| | | | | | It was originally written this way to accomodate for the broken std file API. We no longer use that in buffer, so we can use a more correct version now.
* Implement local CGI error message handlingbptato2023-12-154-3/+31
| | | | | | This was documented, but not implemented until now. Also, improve the loader module's protocol documentation.
* loaderhandle: fix ConversionDefect in getFdbptato2023-12-131-2/+7
| | | | | We must save fd in the constructor, because the stream type may be changed while loading.
* Move http out of main binarybptato2023-12-138-344/+98
| | | | | | | | | | | | Now it is (technically) no longer mandatory to link to libcurl. Also, Chawan is at last completely protocol and network backend agnostic :) * Implement multipart requests in local CGI * Implement simultaneous download of CGI data * Add REQUEST_HEADERS env var with all headers * cssparser: add a missing check in consumeEscape
* Move gopher to adapter/bptato2023-12-122-154/+0
| | | | Also, move default urimethodmap config to res.
* Move out ftp: protocol; fix some local CGI bugsbptato2023-12-123-217/+15
|
* Move out file: protocol handling to adapter/bptato2023-12-123-129/+2
|
* local CGI improvements, move data: to cgi-binbptato2023-12-124-58/+87
| | | | error codes are WIP, not final yet...
* local CGI: add mapped URI env vars; move about: to adaptersbptato2023-12-123-41/+21
| | | | | | | | | | | | | | | * Add MAPPED_URI_* as environment variables when a request is coming from urimethodmap It costs us compatibility with w3m, but it seems to be a massive improvement over smuggling in the URL as a query string and then writing an ad-hoc parser for every single urimethodmap script. The variables are set for every urimethodmap request, to avoid accidental leaking of global environment variables. * Move about: to adapters (an obvious improvement over the previous solution)
* Separate gopher conversion from main binarybptato2023-12-101-115/+10
| | | | | | | | | | | | | | | | | | | | | Now we use a (much simplified) gopher2html binary in libexec, instead of converting gopher directories to HTML in loader/gopher. This has two advantages: * Less ugly conversion logic in the loader module; we can just convert the file line by line. (The previous converter also had some correctness issues, that is fixed now as well.) * If the user desires, they can replace the gopher converter with another binary using the mailcap mechanism. The disadvantages are: * For now, source display is broken. This is a problem with all mailcap filters in general, and should be fixed in the future. (That said, the previous version also only displayed the converted HTML source, which was not really useful anyway.) * The proper directory structure is required for this to work; OTOH plenty of work has been done so that this is as frictionless as possible, so it should not really be a problem.
* http: use Accept-Encodingbptato2023-11-171-0/+3
| | | | just ask libcurl to decode
* reduce new() usagebptato2023-10-251-10/+10
|
* Remove trailing spacesbptato2023-10-231-1/+1
|
* loader: remove stray eprintbptato2023-10-211-1/+0
|
* loader/http: do not choke on 103 early hints, use copyMembptato2023-10-201-11/+39
|
* dirlist: quote path names in hrefbptato2023-10-181-1/+1
| | | | | | | so that it does not choke on files with an apos in them. (We could also htmlEscape it, but this should be enough since we percent-encode the paths already.)
* XHR progressbptato2023-10-141-1/+39
| | | | still non-functional
* Add w3m-cgi-compat optionbptato2023-10-013-7/+29
|
* Add urimethodmap supportbptato2023-09-302-39/+72
| | | | yay
* loader: add local-cgibptato2023-09-304-16/+198
| | | | | | | | | | | Add w3m-style local CGI support. It is not quite as powerful as w3m's local CGI, because it lacks an equivalent to W3m-control. Not sure if it's worth adding; we certainly shouldn't allow passing JS in headers, but a custom language for headers does not sound like a great idea either... eh, idk. also, TODO add multipart
* Response.text: assume utf-8bptato2023-09-271-5/+5
|
* Response.text: encode/decode properlybptato2023-09-271-2/+16
| | | | also, use blob() for images
* buffer: simplify contentType handlingbptato2023-09-272-5/+5
| | | | | | * remove contentType member of Buffer object * add ishtml to reduce string comparisons * consistent spelling: contenttype -> contentType
* Add about:licensebptato2023-09-231-0/+6
|
* buffer: make clone fork()bptato2023-09-232-8/+125
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Makes e.g. on-page anchor navigation near-instantaneous. Well, as instantaneous as a fork can be. In any case, it's a lot faster than loading the entire page anew. This involves duplicating open resources (file descriptors, etc.), which is not exactly trivial. For now we have a huge clone() procedure that does an ok-ish job at it, but there remains a lot of room for improvement. e.g. cloning is still broken in some cases: * As noted in the comments, TeeStream'ing the input stream for any buffer is a horrible idea, as readout in the cloned buffer now depends on the original buffer also reading from the stream. (So e.g. if you clone, then kill the old buffer without waiting for the new one to load, the new buffer gets stuck.) * Timeouts/intervals are broken in cloned buffers. The timeout module probably needs a redesign to fix this. * If you clone before connect2, the cloned buffer gets stuck. The previous solution was even worse (i.e. broken in more cases), so this is still an improvement. For example, this fixes some issues with mailcap handling (removes the "set the Content-Type of htmloutput buffers to text/html" hack), does not reload all resources, does not completely break if the buffer is cloned during loading, etc.
generated by cgit-pink 1.4.1-2-gfad0 (git 2.36.2.497.gbbea4dcf42) at 2025-01-10 14:49:45 +0000