| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
| |
This reverts commit bddecda1a4c96da21e3f5211743ce5e4c78793a2.
This implements the first step in the plan formulated in #20025 to
improve the communication and migration strategy for the proposed
changes to Data.List.
Requires changing the haddock submodule to update the test output.
|
| |
|
|
|
|
|
| |
- Remove GHC.OldList
- Remove Data.OldList
- compat-unqualified-imports is no-op
- update haddock submodule
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The first change makes the array ones use the proper fixed-size types,
which also means that just like before, they can be used without
explicit conversions with the boxed sized types. (Before, it was Int# /
Word# on both sides, now it is fixed sized on both sides).
For the second change, don't use "extend" or "narrow" in some of the
user-facing primops names for conversions.
- Names like `narrowInt32#` are misleading when `Int` is 32-bits.
- Names like `extendInt64#` are flat-out wrong when `Int is
32-bits.
- `narrow{Int,Word}<N>#` however map a type to itself, and so don't
suffer from this problem. They are left as-is.
These changes are batched together because Alex happend to use the array
ops. We can only use released versions of Alex at this time, sadly, and
I don't want to have to have a release thatwon't work for the final GHC
9.2. So by combining these we get all the changes for Alex done at once.
Bump hackage state in a few places, and also make that workflow slightly
easier for the future.
Bump minimum Alex version
Bump Cabal, array, bytestring, containers, text, and binary submodules
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This replaces all Word<N> = W<N># Word# and Int<N> = I<N># Int# with
Word<N> = W<N># Word<N># and Int<N> = I<N># Int<N>#, thus providing us
with properly sized primitives in the codegenerator instead of pretending
they are all full machine words.
This came up when implementing darwinpcs for arm64. The darwinpcs reqires
us to pack function argugments in excess of registers on the stack. While
most procedure call standards (pcs) assume arguments are just passed in
8 byte slots; and thus the caller does not know the exact signature to make
the call, darwinpcs requires us to adhere to the prototype, and thus have
the correct sizes. If we specify CInt in the FFI call, it should correspond
to the C int, and not just be Word sized, when it's only half the size.
This does change the expected output of T16402 but the new result is no
less correct as it eliminates the narrowing (instead of the `and` as was
previously done).
Bumps the array, bytestring, text, and binary submodules.
Co-Authored-By: Ben Gamari <ben@well-typed.com>
Metric Increase:
T13701
T14697
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
Our new CPP linter enforces this.
|
| |
|
|
| |
Refs #9586
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This removes language pragmas from Haskell modules which are implicitly
active with `default-language: Haskell2010`. Specifically, the following
language extension pragmas are removed by this commit:
- PatternGuards
- ForeignFunctionInterface
- EmptyDataDecls
- NoBangPatterns
Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Because MultiByteToWideChar/WideCharToMultiByte have a rather unhelpful
interface, we have to use a lot of binary searching tricks to get them
to match the iconv-like interface that GHC requires.
Even though the resulting encodings are slow, it does at least mean that we
now support all of Window's code pages. What's more, since these codecs are
basically only used for console output there probably won't be a huge volume
of text to deal with in the common case, so speed is less of a worry.
Note that we will still use GHC's faster table-based custom codec for supported
SBCSs.
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
patch series fixes #5061, #1414, #3309, #3308, #3307, #4006 and #4855.
The major changes are:
1) Make Foreign.C.String.*CString use the locale encoding
This change follows the FFI specification in Haskell 98, which
has never actually been implemented before.
The functions exported from Foreign.C.String are partially-applied
versions of those from GHC.Foreign, which allows the user to supply
their own TextEncoding.
We also introduce foreignEncoding as the name of the text encoding
that follows the FFI appendix in that it transliterates encoding
errors.
2) I also changed the code so that mkTextEncoding always tries the
native-Haskell decoders in preference to those from iconv, even on
non-Windows. The motivation here is simply that it is better for
compatibility if we do this, and those are the ones you get for
the utf* and latin1* predefined TextEncodings anyway.
3) Implement surrogate-byte error handling mode for TextEncoding
This implements PEP383-like behaviour so that we are able to
roundtrip byte strings through Strings without loss of information.
The withFilePath function now uses this encoding to get to/from CStrings,
so any code that uses that will get the right PEP383 behaviour automatically.
4) Implement three other coding failure modes: ignore, throw error, transliterate
These mimic the behaviour of the GNU Iconv extensions.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
Add explicit {-# LANGUAGE xxx #-} pragmas to each module, that say
what extensions that module uses. This makes it clearer where
different extensions are used in the (large, variagated) base package.
Now base.cabal doesn't need any extensions field
Thanks to Bas van Dijk for doing all the work.
|
| | |
|
| | |
|
|
|
We keep all of the code page tables in the module
GHC.IO.Encoding.CodePage.Table. That file was generated automatically
by running codepages/MakeTable.hs; more details are in the comments at the
start of that script.
Storing the lookup tables adds about 40KB to each statically linked executable;
this only increases the size of a "hello world" program by about 7%.
Currently we do not support double-byte encodings (Chinese/Japanese/Korean), since
including those codepages would increase the table size to 400KB. It will be
straightforward to implement them once the work on library DLLs is finished.
|
