intarweb, branch 3.0.1Advanced HTTP library to make building HTTP clients and servers convenient.
https://code.more-magic.net/intarweb/atom?h=3.0.12026-08-24T07:36:34ZHack around to make read-bytevector work on custom (chunked) input port2026-08-24T07:36:34ZPeter Bexpeter@more-magic.net2026-08-24T07:36:34Zurn:sha1:7c9edbf53d5c31ffcb3a6981848d24c09c064eef
In CHICKEN 6.0.0, the signature was documented to be different, but
we were using the actual signature. For later CHICKEN versions, the
signature was updated to match the documentation so now we're kind of
stuck having to support both signatures.
Since we cannot cond-expand our way out of this because there are no
features that differ between 6.0.0 and 6.0.1, we'll have to use
case-lambda and convert the old signature to the new one. Oh well...
Add scheme.base to test imports for completeness2026-05-08T11:00:37ZPeter Bexpeter@more-magic.net2026-05-08T11:00:37Zurn:sha1:940c356b2eb41279b1618d1c3bb89533883d5baa
Apparently, "raise" is unavailable in the tests and it gets called in
some cases on Windows.
Remove 2.x releases from release-info file2024-09-12T10:16:40ZPeter Bexpeter@more-magic.net2024-09-12T10:16:40Zurn:sha1:6d3722f533f22c19736d1e62fd506e0d2e511b3e
These don't work with C6
Update intarweb for CHICKEN 62024-09-12T10:15:05ZPeter Bexpeter@more-magic.net2024-09-12T10:15:05Zurn:sha1:192458eb3c4e1d1a31f9b84ac4b2cca43fdc3e1c
This is now using read-bytevector instead of read-string, and the
custom port constructor uses keyword arguments now.
Don't bother to use cond-expand to make it compatible with C5.
Instead, we can cut new C5 releases from the intarweb-2.x branch if
necessary.
Update copyright year2024-01-11T13:04:50ZPeter Bexpeter@more-magic.net2024-01-11T13:04:50Zurn:sha1:e3940c300b993bd1256bdff3758d9c421d664ac7Add 2.1.0 release2024-01-11T13:04:05ZPeter Bexpeter@more-magic.net2024-01-11T13:04:05Zurn:sha1:e91a91b900481405f83880c3af663ce286e51e7fCorrected Range and Content-Range header handling2024-01-11T12:59:39ZBilly Browndruidofluhn@gmail.com2023-12-21T17:07:00Zurn:sha1:b9683af2da882d542208a15d67d8040bab8fa5c2
Corrected the handling of the Range request header and the Content-
Range response header. They now have separate parsers, to match the
standard, and the content-range unparser is now correctly named.
The Range header can be in one of three forms:
Range: bytes=<start>-<end>
Range: bytes=<start>-
Range: bytes=-<end>
The Content-Range header can be in one of three forms:
Content-Range: bytes <start>-<end>/<size>
Content-Range: bytes <start>-<end>/*
Content-Range: bytes */<size>
The unit tests were updated to reflect the changes.
Signed-off-by: Peter Bex <peter@more-magic.net>
Add 2.0.3 release2023-09-07T11:04:21ZPeter Bexpeter@more-magic.net2023-09-07T11:04:21Zurn:sha1:b2632661807a99fc7c3e610e8bd1b04c68b869c5Use string-downcase instead of string-downcase! in http-name->symbol2023-09-07T11:01:42ZPeter Bexpeter@more-magic.net2023-09-07T11:01:42Zurn:sha1:28478e4336e2f1acb990bad48bf2790d04aeee2f
Unlike the srfi-1 linear updating procedures with exclamation marks, in
srfi-13, the side-effecting operations do *not* necessarily return
the string. They just mutate it in-place.
This fixes #1826, so that it will work with utf CHICKEN (upcoming C6).
Publish 2.0.22021-09-13T14:02:11ZPeter Bexpeter@more-magic.net2021-09-13T14:02:11Zurn:sha1:ba82b081176e2b651f255996e5b4846d2b32e87e