| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 37 hours | Update GeneratePMCFG.hsoptimize | adelon | |
| 46 hours | Better names, less code | adelon | |
| 46 hours | Use non-strict functions (makes little difference) | adelon | |
| 47 hours | Roll back spacing changes for minimal diff | adelon | |
| 2 days | Remove cache (no clear benefit) | adelon | |
| 2 days | Comments, `count` -> `areaSum` | adelon | |
| 3 days | Slightly more explicit naming | adelon | |
| 3 days | Tests | adelon | |
| 3 days | Benchmarks, initial sketches | adelon | |
| 2025-08-13 | add import Control.Monad when generating .hs with GADTs | Inari Listenmaa | |
| 2025-08-02 | define return in terms of pure, >> as *>, mappend as <> | Inari Listenmaa | |
| In preparation for deprecation, see https://gitlab.haskell.org/ghc/ghc/-/wikis/proposal/semigroup-monoid and https://gitlab.haskell.org/ghc/ghc/-/wikis/proposal/monad-of-no-return | |||
| 2025-08-02 | add whitespace on list comprehensions, applications etc. | Inari Listenmaa | |
| text editor interprets these things as errors (e.g. unterminated qq for list comprehension) and underlines red, even though there is no real error. | |||
| 2023-09-25 | Indent each line of an error message | Andreas Källberg | |
| By indenting each line instead of just the first, we simplify the work of the gf-lsp parser, so we can see which errors are the same | |||
| 2021-07-12 | Clean up redundant case expressions | Andreas Källberg | |
| 2021-07-12 | Remove last traces of the Either in value2term | Andreas Källberg | |
| 2021-07-12 | Remove the `Either Int` from value2term | Andreas Källberg | |
| This prevents HUGE space leak and makes compiling a PGF a LOT faster For example, an application grammar moved from taking over 50GB of ram and taking 5 minutes (most of which is spent on garbage colelction) to taking 1.2 seconds and using 42mb of memory The price we pay is that the "variable #n is out of scope" error is now lazy and will happen when we try to evaluate the term instead of happening when the function returns and allowing the caller to chose how to handle the error. I don't think this should matter in practice, since it's very rare; at least Inari has never encountered it. | |||
| 2021-07-08 | Fix bug introduced in cdbe73eb475cf44e1a45b2abacb12756e394016a | John J. Camilleri | |
| Apparently I don't understand how pattern-matching works in Haskell | |||
| 2021-07-08 | Remove two missing-methods warnings | John J. Camilleri | |
| 2021-07-07 | Replace tabs for whitespace in source code | John J. Camilleri | |
| 2021-07-06 | Merge branch 'master' into concrete-new | Inari Listenmaa | |
| 2021-07-06 | Merge pull request #118 from GrammaticalFramework/canonical | Inari Listenmaa | |
| Fixes to canonical compilation | |||
| 2021-07-02 | Merge pull request #57 from inariksit/cc-bugfix-rgl-only | Inari Listenmaa | |
| Hotfix for #56 (cc doesn't work for many RGL languages) | |||
| 2021-07-01 | Add --haskell=pgf2 flag | John J. Camilleri | |
| 2021-07-01 | Make imports in CheckGrammar a little more explicit | John J. Camilleri | |
| 2021-07-01 | Rename GF.Compile.TypeCheck.RConcrete to GF.Compile.TypeCheck.Concrete | John J. Camilleri | |
| 2021-07-01 | Rename GF.Compile.Compute.ConcreteNew to GF.Compile.Compute.Concrete | John J. Camilleri | |
| 2021-07-01 | Make cleanupRecordFields also recurse into variants | John J. Camilleri | |
| It's possible that more constructors need to be handled | |||
| 2021-07-01 | Remove record fields not in lincat | John J. Camilleri | |
| Fixes #100, #101 | |||
| 2021-07-01 | Add top-level signatures and general code cleanup | John J. Camilleri | |
| 2021-06-30 | Sort record fields in lin definitions | John J. Camilleri | |
| Fixes #102 | |||
| 2021-06-30 | Make GF.Grammar.Canonical.Id a type synonym for GF.Infra.Ident.RawIdent | John J. Camilleri | |
| This avoids a lot of conversion back and forth between Strings and ByteStrings This commit was cherry-picked from d0c27cdaae78c670b098740bfb49b428d900e640 (lpgf branch) | |||
| 2021-06-29 | Merge pull request #84 from ffrixslee/issue-46 | Inari Listenmaa | |
| Issue 46 (various deprecations during compilation of GF) | |||
| 2021-01-20 | Don't print out the error msg for pattern matching unnecessarily | Inari Listenmaa | |
| 2020-11-10 | Removed fromValue for boolV | Liyana | |
| 2020-11-10 | Deleted redundant pattern match | Liyana | |
| 2020-11-10 | Added explicit implementation for 'fromValue' in instance declaration for ↵ | Liyana | |
| 'Predef Bool' | |||
| 2020-09-05 | Merge remote-tracking branch 'origin/master' into fix-newer-cabal | Andreas Källberg | |
| 2020-09-05 | MonadFail: Make backwards-compatible | Andreas Källberg | |
| 2020-08-05 | Fix most build errors | Andreas Källberg | |
| 2020-08-05 | fix newer ghc: Don't try to be backwards compatible | Andreas Källberg | |
| 2020-08-05 | First attempt at fixing incompabilities with newer cabal | Andreas Källberg | |
| 2020-08-03 | Replace deprecated pragma with up-to-date one. (#17) | Inari Listenmaa | |
| 2020-07-31 | Remove accidentally added space character in deriving clause. | Inari Listenmaa | |
| 2020-07-31 | Remove characters that aren't allowed in Haskell data types. | Inari Listenmaa | |
| GF allows more characters in its types, as long as they are inside single quotes. E.g. 'VP/Object' is a valid name for a GF category, but not for a Haskell data type. | |||
| 2020-07-31 | Add option "data" to Haskell options. | Inari Listenmaa | |
| Imports Data.Data, all GF types derive Data, and uses DeriveDataTypeable. | |||
| 2020-07-06 | an attempt to solve record extension overloading bug, commented out for the ↵ | aarneranta | |
| moment | |||
| 2020-07-06 | fixed issue #67 on order of record fields in overloading | aarneranta | |
| 2020-06-06 | Hotfix for https://github.com/GrammaticalFramework/gf-core/issues/56 | Inari Listenmaa | |
| 2020-06-05 | Condense the unsupported token gluing as per John's suggestion | Inari Listenmaa | |
| 2020-06-04 | Don't output "\n **" if helpfulMsg is empty. | Inari Listenmaa | |
