| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| | | |
| | |
| | |
| | |
| | | |
because they will be soon be added to -Wall.
See https://gitlab.haskell.org/ghc/ghc/-/issues/15656
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* Fix multiple typos and inconsistencies in doc/markup.rst
Note: I noticed some overlap with #1112 from @wygulmage and #1081 from
@parsonsmatt after creating these proposed changes - mea culpa for not
looking at the open PRs sooner.
* Fix #1113 If no Signatures, no section of index.html
* Change the formatting of missing link destinations
The current formatting of the missing link destination does not really
help user to understand the reasons of the missing link.
To address this, I've changed the formatting in two ways:
- the missing link symbol name is now fully qualified. This way you
immediately know which haskell module cannot be linked. It is then easier
to understand why this module does not have documentation (hidden module
or broken documentation).
- one line per missing link, that's more readable now that symbol name
can be longer due to qualification.
For example, before haddock was listing missing symbol such as:
```
could not find link destinations for:
Word8 Word16 mapMaybe
```
Now it is listed as:
```
could not find link destinations for:
- Data.Word.Word8
- Data.Word.Word16
- Data.Maybe.mapMaybe
```
* Add `--ignore-link-symbol` command line argument
This argument can be used multiples time. A missing link to a symbol
listed by `--ignore-link-symbol` won't trigger "missing link" warning.
* Forbid spaces in anchors (#1148)
* Improve error messages with context information (#1060)
Co-authored-by: Matt Audesse <matt@mattaudesse.com>
Co-authored-by: Mike Pilgrem <mpilgrem@users.noreply.github.com>
Co-authored-by: Guillaume Bouchard <guillaume.bouchard@tweag.io>
Co-authored-by: Pepe Iborra <pepeiborra@gmail.com>
|
| | | |
| | |
| | |
| | |
| | |
| | | |
If we compile documentation where only a small fraction is undocumented,
it is misleading to see 100% coverage - 99% is more intuitive.
Fixes #1194
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This allows to have stable anchors for groups, even if the set of
groups in the documentation is altered.
The syntax for setting the anchor of a group is
-- * Group name #desiredAnchor#
Which will produce an html anchor of the form '#g:desiredAnchor'
Co-authored-by: Iñaki García Etxebarria <git@inaki.blueleaf.cc>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Sometimes, the declaration being exported is a subdecl (for instance, a
record accessor getting exported at the top-level). For these cases,
Haddock has to find a way to produce some synthetic sensible top-level
declaration. This is done with `extractDecl`.
As is shown by #1067, this is sometimes impossible to do just at a
syntactic level (for instance when the subdecl is re-exported). In these
cases, the only sensible thing to do is to try to reify a declaration
based on a GHC `TyThing` via `hiDecl`.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
When renaming the Haddock interface, never emit warnings when renaming a
minimal signature. Also added some documention around `renameInterface`.
Minimal signatures intentionally include references to potentially
un-exported methods (see the discussion in #330), so it is expected
that they will not always have a link destination. On the principle
that warnings should always be resolvable, this shouldn't produce a
warning. See #1070.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
When first creating a Haddock interface, trim `ifaceDocMap` and
`ifaceArgMap` to not include docstrings that can never appear in the
final output. Besides checking with GHC which names are exported, we
also need to keep all the docs attached to instance declarations (it is
much tougher to detect when an instance is fully private).
This change means:
* slightly smaller interface files (7% reduction on boot libs)
* slightly less work to do processing docstrings that aren't used
* no warnings in Haddock's output about private docstrings (see #1070)
I've tested manually that this does not affect any of the boot library
generated docs (the only change in output was some small re-ordering in
a handful of instance lists). This should mean no docstrings have been
incorrectly dropped.
|
| | | |
| | |
| | |
| | |
| | | |
* Recode Doc to Json.
* More descriptive field labels.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This makes the `synifyType` case for `AppTy` more intelligent by
taking into consideration the visibilities of each `AppTy` argument
and filtering out any invisible arguments, as they aren't intended
to be displayed in the source code. (See #1050 for an example of what
can happen if you fail to filter these out.)
Along the way, I noticed that a special `synifyType` case for
`AppTy t1 (CoercionTy {})` could be consolidated with the case below
it, so I took the opportunity to tidy this up.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This is quite straightforward to implement, since we already had a
function `docToHtmlNoAnchors` (which we used to generate the link in the
sidebar "Contents").
This breaks test `Bug387`, but that test case has aged badly: we now
automatically generate anchors for all headings, so manually adding an
anchor in a section makes no sense. Nested anchors are, as pointed out
in #1054, disallowed by the HTML standard.
Fixes #1054
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* enable more useful warning flags in `haddock-api`, handle the new
warnings generated
* remove `-fwarn-tabs` (now we'd use `-Wtabs`, but this has been in
`-Wall` for a while now)
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* `unL` is already defined by GHC as `unLoc`
* `reL` is already defined by GHC as `noLoc` (in a safer way too!)
* Condense `setOutputDir` and add a about exporting from GHC
Fixes #978
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
None of these error cases should ever have been reachable, so this is
just a matter of leveraging the type system to assert this.
* Use the `NoExtCon` and `noExtCon` to handle case matches for no
extension constructors, instead of throwing an `error`.
* Use the extension field of `HsSpliceTy` to ensure that this variant
of `HsType` cannot exist in an `HsType DocNameI`.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | | |
* removed functions in `Haddock.Utils` that were not used anywhere
(or exported from the `haddock-api` package)
* moved GHC-specific utils from `Haddock.Utils` to `Haddock.GhcUtils`
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This a GHC bug (https://gitlab.haskell.org/ghc/ghc/issues/14109) too,
but it is a relatively easy fix in Haddock. Note that the fix must live
in `haddock-api` instead of `haddock-library` because we can only really
decide if an identifier is a reserved one by asking the GHC lexer.
Fixes #952
|
| | | |
| | |
| | |
| | |
| | | |
Adjusted changelogs and versions in `.cabal` files in preparation for
the upcoming release bundled with GHC 8.10.
|
| | | |
| | |
| | |
| | |
| | |
| | | |
* unused imports
* imports of `Data.List` without import lists
* missing `CompatPrelude` file in `.cabal`
|
| | |\ \ |
|
| | | | | |
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We now have a top-level `tyConAppNeedsKindSig` function, which means
that we can delete lots of code in `Convert`.
(cherry picked from commit cfd682c5fd03b099a3d78c44f9279faf56a0ac70)
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This fixes #1103 by adapting the corresponding patch for GHC (see
https://gitlab.haskell.org/ghc/ghc/issues/17296 and
https://gitlab.haskell.org/ghc/ghc/merge_requests/1877).
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
It is possible to fail to extract an HIE ast. This is however not a
reason to produce _no_ output - we should still make a colorized HTML
page.
|
| | | | | |
|
| | | | |
| | | |
| | | | |
The `ignore-exports` option has been broken since #688, as mentioned in https://github.com/haskell/haddock/pull/766#issue-172505043. This PR fixes it.
|
| | | | | |
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
At this point, Haddock depended on Cabal-the-library solely for a
verbosity parser (which misleadingly accepts all sorts of verbosity
options that Haddock never uses). Now, the only dependency on Cabal
is for `haddock-test` (which uses Cabal to locate the Haddock interface
files of a couple boot libraries).
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The only other change in html/hoogle/hyperlinker output for the boot
libraries that this caused is a fix to some Hoogle output for implicit
params.
```
$ diff -r _build/docs/ old_docs
diff -r _build/docs/html/libraries/base/base.txt old_docs/html/libraries/base/base.txt
13296c13296
< assertError :: (?callStack :: CallStack) => Bool -> a -> a
---
> assertError :: ?callStack :: CallStack => Bool -> a -> a
```
|
| | | | | |
|
| | | | | |
|
| | | | |
| | | |
| | | |
| | | | |
(cherry picked from commit 3ee6526d4ae7bf4deb7cd1caf24b3d7355573576)
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Before LLVM 6.0.1 (or 10.0 on Apple LLVM), there was a bug where
lines that started with an octothorpe but turned out not
to lex like pragmas would have an extra line added after them.
Since this bug has been fixed upstream and that it doesn't have dire
consequences anyways, the workaround is not really worth it
anymore - we can just tell people to update their clang version (or re-structure
their pragma code).
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
After this commit, we can run with `--latex` on all boot libraries
without crashing (although the generated LaTeX still fails to compile in
a handful of larger packages like `ghc` and `base`).
* Add newlines after all block elements in LaTeX. This is important to
prevent the final output from being more an more indented. See the
`latext-test/src/Example` test case for a sample of this.
* Support associated types in class declarations (but not yet defaults)
* Several small issues for producing compiling LaTeX;
- avoid empy `\haddockbeginargs` lists (ex: `type family Any`)
- properly escape identifiers depending on context (ex: `Int#`)
- add `vbox` around `itemize`/`enumerate` (so they can be in tables)
* Several spacing fixes:
- limit the width of `Pretty`-arranged monospaced code
- cut out extra space characters in export lists
- only escape spaces if there are _multiple_ spaces
- allow type signatures to be multiline (even without docs)
* Remove uninteresting and repetitive `main.tex`/`haddock.sty` files
from `latex-test` test reference output.
Fixes #935, #929 (LaTeX docs for `text` build & compile)
Fixes #727, #930 (I think both are really about type families...)
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
`markupWarning` often processes inputs which span across paragraphs.
Unfortunately, LaTeX's `emph` is not made to handle this (and will
crash).
Fixes #936.
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
* default methods now get rendered differently
* default associated types get rendered
* fix a forgotten `s/TypeSig/ClassOpSig/` refactor in LaTeX backend
* LaTeX backend now renders default method signatures
NB: there is still no way to document default class members and the
NB: LaTeX backend still crashes on associated types
|
| | | |\ \ |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Docs on standalone deriving decls for classes with associated types
should be associated with the class instance, not the associated type
instance.
Fixes #1033
|
| | | | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The information about whether or not there is a source-level `forall`
is already available on a `ConDecl` (as `con_forall`), so we should use
it instead of always assuming `False`!
Fixes #1002.
|
| | | | | |
| | | | |
| | | | | |
Fixes #992
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
* WIP: Load (typechecker) plugins from language pragmas
* Revert "Load plugins when starting a GHC session (#905)"
This reverts commit 72d82e52f2a6225686d9668790ac33c1d1743193.
* Simplify plugin initialization code
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
* don't forget to print explicit `forall`'s when there are arg docs
* when printing an explicit `forall`, print all tyvars
Fixes #973
|
| | | | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This avoids a situation in which an identifier would get defaulted to
a completely different identifier. Prior to this commit, the 'Bug1035'
test case would hyperlink 'Foo' into 'Bar'!
Fixes #1035.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
* '(<|>)' and '`elem`' now get parsed and rendered properly as links
* 'DbModule'/'DbUnitId' now properly get split apart into two links
* tuple names now get parsed properly
* some more small niceties...
The identifier parsing code is more precise and more efficient (although to be
fair: it is also longer and in its own module). On the rendering side, we need
to pipe through information about backticks/parens/neither all the way through
from renaming to the backends.
In terms of impact: a total of 35 modules in the entirety of the bootlib + ghc
lib docs change. The only "regression" is things like '\0'. These should be
changed to @\\0@ (the path by which this previously worked seems accidental).
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Identifier links can be prefixed with a 'v' or 't' to indicate the value or
type namespace of the desired identifier. For example:
-- | Some link to a value: v'Data.Functor.Identity'
--
-- Some link to a type: t'Data.Functor.Identity'
The default is still the type (with a warning about the ambiguity)
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
In https://gitlab.haskell.org/ghc/ghc/merge_requests/1970 I propose a
simpler way to encode location information into the GHC and Haddock AST
while incurring no cost for e.g. TH which doesn't need location
information.
These are just changes that have to happen in lock step.
|
| | | | | |
| | | | |
| | | | |
| | | | | |
(cherry picked from commit a7d1d8e034d25612d5d08ed8fdbf6f472aded4a1)
|
| | | | | | |
|
| | |_|_|/
|/| | | |
|
| | | | |
| | | |
| | | |
| | | | |
See ghc ticket #18048
|
