Commit Graph

159 Commits (71efa549540ee085b7abe49a73b5dd95b97b3a4e)

Author SHA1 Message Date
Lennart Spitzner 71efa54954 Move tests, minor Refactoring, Add comments 2018-09-23 23:32:05 +02:00
Sergey Vinokurov 9755db1d05
Unify "(#", "#)" under single name, docParenHashL and docParenHashR 2018-09-18 09:14:20 +01:00
Sergey Vinokurov 6898d3ef44
Lay out unboxed tuples in types 2018-09-18 09:07:15 +01:00
Sergey Vinokurov 281d7a2f81
Lay out unboxed tuples with spaces
This avoids clashes with names like foo#
2018-09-18 09:07:13 +01:00
Evan Rutledge Borden 92a1d89983 Consolidate record expression layouter
Both record construction and record layouting have very similar
constructions. These each had their own layouter with slightly different
variations. Variations here lead to subtley different bugs in layout for
nearly identicle syntactic forms.

The record update logic is more advanced and respects `IndentPolicyLeft`.
Instead of keeping these layouters distinct we can consolidate
construction logic into the update logic. This results in a smaller
volume of code and more uniform layouting of syntax for these simlilar
forms.

Record constructors with fields and wildcards are not included in this
consolidation. A TODO has been left to handle this consolidation later.
2018-09-16 15:47:04 -04:00
Lennart Spitzner 8c5cce5070 Prevent package environment file GHC API stupidities
This fixes the code-path for the commandline interface
(unfortunately not covered by our testsuite).
The other code-path should be covered by the changes in
https://github.com/alanz/ghc-exactprint/pull/68.
2018-08-20 21:19:10 +02:00
Phil Hazelden a5a24b4220 Fix imports of type operators.
Previously, we could only import a type operator with no subsequent
list, i.e.

    import Foo ( (:.) )

was fine, but

    import Foo ( (:.)(..) )
    import Foo ( (:.)((:.) )
    import Foo ( (:.)(A, b) )

would all break. Brittany would attempt to output them as

    import Foo ( :.(..) )
    import Foo ( :.((:.) )
    import Foo ( :.(A, b) )

I believe the problem was that although `ieName <$> lie` was returning
an `IEWrappedName` with the same contents as used in `layoutWrapped`,
it had different location annotations; and the parentheses are
apparently saved in the location annotations.
2018-08-10 10:01:11 +01:00
Lennart Spitzner 4497fa927f Add a comment about one superfluous docSetParSpacing 2018-07-06 22:57:33 +02:00
Lennart Spitzner ab27825b7b Core layouting algo improvement (non-bottom more effective)
During alt-transform, when gather spacings, previously
we tracked different non-bottom spacings separately even
though they would be treated in the same way during any
future transformations (apart from certain exceptions that
don't practically give better results). Instead we now
merge such spacings into one, giving more space for other
spacings when pruning to the spacings limit.
2018-07-06 22:41:31 +02:00
Lennart Spitzner 3c5670d5cd Fix layouting for OpApps with comments (fixes 159) 2018-07-05 21:31:28 +02:00
Lennart Spitzner 95f42061d2 Fixup for the fix of ExplicitNames/PatternSynonyms for ghc-8.0 2018-07-04 22:42:20 +02:00
Lennart Spitzner ab389fe66f Support for -XExplicitNamespaces and -XPatternSynonyms
Properly round-trip export items of the forms "type OPERATOR"
or "pattern SYNONYM"

fixes #158
2018-07-04 21:42:46 +02:00
Lennart Spitzner 0b40dd7c32 Fix deprecation warnings with latest versions of yaml 2018-07-04 21:14:06 +02:00
Lennart Spitzner cc0718e964 Add indentation to import-hiding-paragraph 2018-06-04 18:18:44 +02:00
Lennart Spitzner 5816704658 Fix import-hiding-paragraph with policy/=free (fixes #150) 2018-06-04 17:30:50 +02:00
Lennart Spitzner 57c48f64c1 Apply inline config to module header (fixes #151) 2018-06-04 16:57:07 +02:00
Lennart Spitzner bdee27cb59 Add spaces around record braces (single-line) (fixes #126) 2018-05-25 17:57:54 +02:00
Lennart Spitzner f75127b3a5 Make non-bottom-spacing affect sameline, too (fixes #144) 2018-05-16 21:19:36 +02:00
Lennart Spitzner 4973298f30 Support same-line let decl when indentAmount>=4 2018-05-01 23:21:31 +02:00
Lennart Spitzner dd53948a23 Fix IndentPolicyMultiple for indentAmount>4 2018-05-01 23:21:09 +02:00
Lennart Spitzner e91bb6aec9 Clean up IndentPolicyMultiple 2018-05-01 23:19:44 +02:00
Bryan Richter 9ab17cc899 Implement IndentPolicyMultiple 2018-05-01 22:07:05 +02:00
Lennart Spitzner 929e465fd4 Refactor a bit (Internal.ppDecl -> Decl.layoutDecl) 2018-04-25 20:17:36 +02:00
Lennart Spitzner a237e591b2 Implement `--obfuscate`
Support replacing all (non-keyword) identifiers with
randomly generated characters
2018-04-24 00:47:50 +02:00
Lennart Spitzner 696f72d336 Fixup last two commits (tests were effectively disabled
- re-enable tests
- un-break tests by fixing the alignment code behaviour
2018-04-24 00:38:40 +02:00
Lennart Spitzner 315a7e1ee1 Fix overflowing due to alignment; Add docs (alignment algorithm) 2018-04-23 19:33:04 +02:00
Lennart Spitzner 18704e403f Fix inline disabling of brittany 2018-04-19 17:09:10 +02:00
Lennart Spitzner 1fc007591c Merge branch 'master' into inlineconfig
# Conflicts:
#	src/Language/Haskell/Brittany/Internal.hs
2018-04-17 20:05:33 +02:00
Lennart Spitzner e559a2cbf7 Implement inline configuration
e.g. "-- brittany --indent=4"

respects the following comment forms as input:

  source comment                       affected target
  ======================================================
  "-- brittany CONFIG"                 whole module
  "-- brittany-next-binding CONFIG"    next binding
  "-- brittany-disable-next-binding"   next binding
  "-- brittany @ myExampleFunc CONFIG" `myExampleFunc`

multiline-comments are supported too, although
the specification must still be a single line. E.g.

  "{- brittany --columns 50 -}"

CONFIG is either:

  1) one or more flags in the form of what brittany accepts
     on the commandline, e.g. "-- columns 50", or
  2) one or more specifications in the form of what brittany
     accepts in its config files for the layouting config
     (a one-line yaml document), e.g. "{ lconfig_cols: 50 }"

see #30
2018-04-17 20:04:40 +02:00
Lennart Spitzner 10e1c19788 Add support for ghc-8.4 2018-04-17 17:50:41 +02:00
Lennart Spitzner 213e82b476 Fix if-then-else paragraph layouting 2018-04-17 06:19:13 +02:00
Lennart Spitzner ae162403ae Support multistate-0.9 (ghc-8.4 preparation) 2018-04-13 22:13:15 +02:00
Lennart Spitzner 8b67a028ea Do not put `where` on newline when no export list
(even when lconfig_allowSingleLineExportList False)
2018-04-09 00:09:07 +02:00
Lennart Spitzner 21ef8b296c Adapt for czipwith-1.0.1.0 2018-04-05 21:06:44 +02:00
Sergey Vinokurov 9bd3bfbe4c
Review suggestions 2018-04-03 22:49:06 +01:00
Sergey Vinokurov 7a60229673
Fix some hlint hints 2018-04-02 22:44:49 +01:00
Sergey Vinokurov 631d9e181d
Replace 'fmap f $' with 'f <$>' 2018-04-02 22:44:49 +01:00
Sergey Vinokurov 545eff9e4f
Remove redundant parens 2018-04-02 22:42:06 +01:00
Sergey Vinokurov 226da07815
Improve vertical alignment 2018-04-02 22:41:15 +01:00
Sergey Vinokurov 0dad5051df
Remove redundant '$'s 2018-04-02 22:41:14 +01:00
Sergey Vinokurov 2ed9a13fdb
Replace 'docAltFilter' with 'runFilteredAlternative' 2018-04-02 21:25:44 +01:00
Sergey Vinokurov 8410fbff8e
Trailing whitespace 2018-04-02 21:25:44 +01:00
Sergey Vinokurov e9689394b1
Add Semigroup instance for 'Max' monoid
Otherwise ghc 8.4 will complain since Semigroup became a superclass
of Monoid.
2018-04-02 21:25:43 +01:00
Lennart Spitzner b43ee43220 Fix/Implement empty type constraint handling (fixes #133) 2018-04-02 17:11:53 +02:00
Lennart Spitzner 0845142727 Fix let-in comment placement bug (fixes #110) 2018-03-23 17:34:36 +01:00
Lennart Spitzner bdd3b155f3 Fix HsPar comment placement bug (see #111) 2018-03-23 17:11:39 +01:00
Lennart Spitzner 3847325fd5 Omit single-line layout for OpApp with comments (fixes #111) 2018-03-23 17:02:58 +01:00
Lennart Spitzner a003b932a9 Fix comments in tuples being dropped (fixes #37) 2018-03-23 16:55:28 +01:00
Lennart Spitzner d634d34ff1 Fix module-import-hiding-items layout 2018-03-23 15:41:41 +01:00
Lennart Spitzner 90a2f65ba7 Align applications on for same function, plus minor fixup
Arguments of two function applications will only be aligned
if the same function is called in both cases.

The column transform was altered slightly to fix #65
properly as well.

fixes #65, #128
2018-03-22 01:19:56 +01:00