GitLab

- Fixes #4476
- Fixes #5283

i.e. strict pair of PackageName and LibraryName
the legacy conversion is done via Pretty/Parsec instances.

Change of `Maybe UnqualComponentName` to `LibraryName` caused
a cascade of other changes, but they all seem to be good changes.
In the sense, they made many comments not-so-necessary.

Add Distribution.Types.PackageName.Magic for special package names.

Updates in cabal-install are mostly trivial type error driven changes.
I removed few (deprecated) `Text` instances: `MungedPackageId`,
`MungedPackageName` and `LibraryName`. Turns out only a `Pretty`
part was used, so it was easy to update. Note: `LibraryName`
doesn't have `Pretty` / `Parsec` instances as it's either parsed/printed
as a `ComponentName` or `UnqualComponentName`, never stand alone.

Split fields related functionality out of `Distribution.Parsec`
namespace which is not left only to handle "scalar" values.
This separation highlights that we use `parsec` for *structure* (think
JSON), and the *values* of the fields (think e.g. dates in JSON).

- New top-level `Distribution.Fields` and `Distribution.Parsec` modules
  should include most things most people would need.
- Also `Distribution.Pretty.Field` is moved to `Distribution.Fields.Pretty`,
  as now it has proper place.
- The commit is moving things around and fixing compilation errors.
- I fixed `cabal check` parse error printing as I was nearby.

- Warn about missplaced import (on top level)
- Warn in trees

- for main library default is True
- for named libraries default is False
- remove Monoid Library instance; i'd rather remove emptyLibrary too
  but now these (two?) usages are explicit:
  - creating Library from BuildInfo when merging common stanzas
  - creation of dummy package in `cabal new-repl`
- Documentation is left for Francesco to write
- Field is named `visible` because we have `buildable` and `exposed`
  adjectives. `visibility: True` is unclear.

quasicomputational authored Jul 10, 2018 and Mikhail Glushenkov committed Jul 31, 2018

This code existed to support building Cabal with 7.4 and earlier, but
we have recently dropped support for 7.4.

Lexically, they're identical to 2.2, but the semantics of globs have
changed slightly.

This is not a proper fix, but we are more conservative than is strictly
required. I.e. each "branch" must have `type`.

Also as `type` is set, then `main-is` (or module) have to be specified,
because of `validate*` functions.

The `onAllBranches` is wrong. But we can fix (i.e. relax) it for Cabal-2.4

/Note:/
regressions/issue-5055.format: it's broken as there isn't `main-is`
`os(windows) conditional branch in the formatted output.

A little refactoring in handling UTF8, it's validated as a preprocess
step, so consequent steps can simply use lenient streams.

Distribution.PackageDescription.Parsec had a locally defined empty
GenericPackageDescription, which turned out to be useful for unit testing the
NFdata instance of that type, too, so it's now available properly from the
Distribution.Types.GenericPackageDescription module.

There are 3957 warned files atm in Hackage index.
All of them have tab "inside the field", thus old parser acceps them.

See e.g added th-lift-instances file. The warning comes from the description field.

Herbert Valerio Riedel authored Sep 20, 2017 and Oleg Grenrus committed Jan 16, 2018

This provides a provisional (i.e. hacky) retrofitted implementation of
the forward-compat scheme described in #4899 for the cabal-2.2 branch

This hack works by constructing a dummy package description in case
the package description fails to be parsed via the standard parser,
and we detect a new-style cabal-spec declaration.

Fixes #5022

Tag Backpack fields (mixins, signatures) to `availableSince [2,0]`.
This "fixes" https://github.com/haskell/cabal/issues/4448, as fields
are recognised, warned, but parsed as empty if cabal-version < 2.0
(actual cut-off is ! (>= 1.25). For example, a file with

    cabal-version: >=1.10
    library:
      mixins: foo-indef requires (Foo42 as FooImpl)

will be accepted, yet warned, and parsed `mixins` in `BuildInfo` will be
an empty list.

Also availableSince is removed from `build-tool-depends`,
as we **want** to parse (and not warn) it in old Cabal files.
It can be thought as added retrospectively to old specs, but old `Cabal`
s don't know how to use it.

CabalSpecVersion type-class will allow to gather per-spec conditionals.
Currently it's used for selecting parsers / grammatical structure.

Leading (or trailing commas) for CommaFSep/CommaVSep fields,
i.e. fields with mandatory comma are (atm):

- build-depends
- build-tool-depends
- build-tools
- mixins
- pkgconfig-depends
- reexported-modules
- setup-depends

- common stanzas can be include other common stanzas
- `import: name1, name2` to import multiple stanzas
- Parse common stanzas in the same pass with other sections.
- Common stanzas have to be defined before use.
- Also negative tests
- Terse documentation, let's improve it as questions are asked

- Edit gen-extra-source-files to include golden files
- Amend elif warning to mention cabal-version: 2.2
- In regression golden tests, include also warnings

Note: ATM the common stanzas are completely handled inside parser,
GenericPackageDescription doesn't know about them anymore.
That can be changed, but the the flattening of
GenericPackageDescription to PackageDescription may fail.
I don't want to do that refactor now.

This commit reworks how GenericPackageDescription is parsed from
`[Field Position]` and pretty-printed to `Doc`.

This also fixes few issues:
- Fix #4697: `cabal format` doesn't output custom-setup stanza (nor
  foreign-lib stanzas)
- Fix #4719: `parse . pretty . parse = parse`  for all Hackage cabal files.
    - `parser-hackage-tests roundtrip` is the test program.

The handling of `license-file` and `license-files` is changed.
Now they behave the same.

At the moment we have very lax "name" lexer, which is essentially
*everything else* than otherwise specified characters: parens, space,
symbols (but not hyphen -)

This removes `TokNum` and `SecArgNum`, both are merged into `TokSym` and
`SecArgSym` respectively.

- Manually generate Lexer.hs
- Temporarily disable parser-hackage-tests on appveyor (needs
  01-index.tar)
- Add root Makefile to handle commmon dev tasks (genering Lexer.hs)

Edward Z. Yang authored Feb 13, 2017 and Edward Z. Yang committed Feb 19, 2017

This flips error handling around, so that 'die' now can format
an error message with call stacks and markers before raising
it to the top handler.  The top handler detects "verbatim"
deaths and prints them without formatting.
Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>