6bcced31a0
fix 32-bit compilation
2025-07-31 22:10:11 -07:00
42b10f08cc
std.compress.flate.Decompress: delete bad unit tests
...
if I remove the last input byte from "don't read past deflate stream's
end" (on master branch), the test fails with error.EndOfStream. what,
then, is it supposed to be testing?
2025-07-31 22:10:11 -07:00
5f790464b0
std.compress.flate.Decompress: hashing is out of scope
...
This API provides the data; applications can verify their own checksums.
2025-07-31 22:10:11 -07:00
f3a38e30fa
std.Io: delete SeekableStream
...
Alternative is to use File.Reader and File.Writer directly.
2025-07-31 22:10:11 -07:00
4741a16d9a
putting stuff back does not require mutation
2025-07-31 22:10:11 -07:00
05ce1f99a6
compiler: update to new flate API
2025-07-31 22:10:11 -07:00
2569f4ff85
simplify tossBitsEnding
2025-07-31 22:10:11 -07:00
5bc63794cc
fix takeBitsEnding
2025-07-31 22:10:11 -07:00
63f496c4f9
make takeBits deal with integers only
2025-07-31 22:10:11 -07:00
c00fb86db6
fix peekBitsEnding
2025-07-31 22:10:11 -07:00
8ab91a6fe9
error.EndOfStream disambiguation
2025-07-31 22:10:11 -07:00
f644f40702
implement tossBitsEnding
2025-07-31 22:10:11 -07:00
2d8d0dd9b0
std.compress.flate.Decompress: unfuck the test suite
2025-07-31 22:10:11 -07:00
c684b21b4f
simplify test cases
2025-07-31 22:10:11 -07:00
ac4fbb427b
std.compress.flate.Decompress: don't compute checksums
...
These have no business being in-bound; simply provide the expected
values to user code for maximum flexibility.
2025-07-31 22:10:11 -07:00
5f571f53d6
refactor gzip test cases
...
zig newbies love using for loops in unit tests
2025-07-31 22:10:11 -07:00
e73ca2444e
std.compress.flate.Decompress: implement peekBitsEnding and writeMatch
2025-07-31 22:10:11 -07:00
7bf91d705c
fix bit read not at eof
2025-07-31 22:10:11 -07:00
73e5594c78
std.compress.flate.Decompress: fix bit read at eof
2025-07-31 22:10:11 -07:00
9c8cb777d4
std.compress.flate.Decompress: implement more bit reading
2025-07-31 22:10:11 -07:00
6509fa1cf3
std.compress.flate.Decompress: passing basic test case
2025-07-31 22:10:11 -07:00
88ca750209
std.compress.flate.Decompress: add rebase impl
2025-07-31 22:10:11 -07:00
fa410cc234
std.Io: delete BitReader
2025-07-31 22:10:11 -07:00
1b43551190
std.Io: remove BitWriter
2025-07-31 22:10:11 -07:00
824c157e0c
std.compress.flate: finish reorganizing
2025-07-31 22:10:11 -07:00
73c98ca0e6
simplify std.hash.Adler32
2025-07-31 22:10:11 -07:00
a4f05a4588
delete flate implementation
2025-07-31 22:10:11 -07:00
83513ade35
std.compress: rework flate to new I/O API
2025-07-31 22:10:11 -07:00
a2d21d6327
enable pwd.h functions for other OSes
...
also add the layout of `struct passwd` for DragonflyBSD
and FreeBSD:
- https://github.com/DragonFlyBSD/DragonFlyBSD/blob/c267aac0072dae6cf4ae874605f3f0659a2fc820/include/pwd.h#L112
- https://cgit.freebsd.org/src/tree/include/pwd.h?id=d66f9c86fa3fd8d8f0a56ea96b03ca11f2fac1fb#n114
2025-08-01 04:34:45 +02:00
0294e91451
std.Io.Reader: fix readVec at end
2025-07-31 19:26:12 -07:00
ac1e73e249
std.enums: fix EnumIndexer branch quota
...
It's quite silly to have this override which nonetheless makes
assumptions about the input type. Encode the actual complexity of the
sort.
Also, simplify the sorting logic, and fix a bug (grab min and max
*after* the sort, not *before*!)
2025-07-31 22:10:22 +01:00
04d7b491b4
Merge pull request #24632 from mlugg/lossy-int-to-float-coercion
...
Sema: compile error on lossy int to float coercion
2025-07-31 21:49:37 +01:00
982c387753
Merge pull request #24633 from linusg/more-serenity-fixes
...
std: A few more fixes for serenity
2025-07-31 09:56:26 -07:00
03facba496
std.Target: require libc for Android API levels prior to 29
...
Emulated TLS depends on libc pthread functions.
Closes #24589 .
2025-07-31 17:08:25 +02:00
64bf8bb146
std: stop relying on precision-losing coercions
2025-07-31 10:57:04 +01:00
e941ce3e68
add grp.h functions to c.zig
2025-07-30 21:28:42 -07:00
ce776d3245
std: Add serenity to more OS checks
2025-07-30 23:28:58 +01:00
813a0f125e
std.posix: Default ACCMODE to NONE for serenity
...
Unlike all other platforms where RDONLY is 0 it does not work as a
default for the O flags on serenity - various syscalls other than
'open', e.g. 'pipe', return EINVAL if unexpected bits are set in the
flags.
2025-07-30 23:27:32 +01:00
f5e9384335
std.c: Fix MAP for serenity
...
I accidentally translated MAP_ constants representing the type as
individual fields. MAP_FILE is for compatibility only and not needed
here.
2025-07-30 23:22:06 +01:00
467a1f4a1c
std.c: Fix msghdr_const for serenity
2025-07-30 23:19:29 +01:00
de23ccfad1
build system: print captured stderr on Run step failure
...
when a Run step that captures stderr fails, no output from it is visible
by the user and, since the step failed, any downstream step that would
process the captured stream will not run, making it impossible for the
user to see the stderr output from the failed process invocation, which
makes for a frustrating puzzle when this happens in CI.
2025-07-30 22:40:36 +01:00
259b7c3f3f
std.Target: pull Os.requiresLibC() up to Target
2025-07-30 18:58:47 +02:00
cf7a28febb
std.Io.Reader: introduce readVec back into the VTable
...
simplifies and fixes things
addresses a subset of #24608
2025-07-30 09:26:34 -07:00
135a34c963
Update doc comment for ptr_type and ptr_type_bit_range to data of .extra_and_node
...
The other pointer types are `.opt_node_and_node` but `ptr_type` and `ptr_type_bit_range` contain `.extra_and_node` in their `data` field
2025-07-30 10:02:41 +01:00
1fcaf90dd3
std.Io.Reader: make fillUnbuffered respect prexisting buffer
...
addresses only one usage pattern in #24608
2025-07-29 23:10:50 -07:00
08f1d63be1
disable more failing tests
...
Wow, *lots* of backends were reliant on Sema doing the heavy lifting for
them. CBE, Wasm, and SPIR-V have all regressed in places now that they
actually need to, like, initialize unions and such.
2025-07-29 22:44:01 +01:00
ecd3ea9bd2
DeprecatedReader.Adapted: fix EndOfStream handling
2025-07-28 21:41:19 -07:00
c334956a54
aarch64: workaround some optional/union issues
2025-07-28 09:03:17 -07:00
bb29846732
std.compress.xz: eliminate dependency on std.Io.bitReader
2025-07-27 22:08:52 -07:00
e12dc4947c
std.zig: fmtId returns a FormatId
...
Changes fmtId to return the FormatId type directly, and renames the
FormatId.render function to FormatId.format, so it can be used in a
format expression directly.
Why? Since `render` is private, you can't create functions that wrap
`fmtId` or `fmtIdFlags`, since you can't name the return type of those
functions outside of std itself.
The current setup _might_ be intentional? In which case I can live with
it, but I figured I'd make a small contrib to upstream zig :)
2025-07-26 21:53:23 -07:00
Andrew Kelley