| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
from ghc-9.2 on, the “default” langauge of GHC is expected to change
more wildly. To prepare for that (and unblock
https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4853), this sets the
language for all the test files to `Haskell2010`. This should insolate
this test suite against changes to the default.
Cherry-picked from https://github.com/haskell/haddock/pull/1341
|
|\ |
|
| | |
|
| | |
|
|/
|
|
|
|
|
|
|
|
|
| |
This means that `html-test`, `latex-test`, `hoogle-test`, and
`hypsrc-test` now only depend on GHC boot libs. So we should
now be able to build and run these as part of GHC's testsuite. \o/
The reference output has changed very slightly, in three ways:
* we don't convert quotes back into `"` as the `xml` lib did
* we don't add extra ` ` as the `xml` lib did
* we now remove the entire footer `div` (instead of just emptying it)
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
Correspondingly, we wrap all inline/diplay math in
<span class="mathjax"> ... the math .... </span>
This fixes #959.
|
| | |
|
| | |
|
| |
| |
| |
| |
| | |
This solves an issue reported about the content looking
incredibly small on mobile devices.
|
| |
| |
| |
| |
| | |
The package name element in the package-header is now a div instead of a
paragraph, and it is now above the menu ul.links instead of below.
|
|/ |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Add compile step that bundles and compresses JS files
Also, manage dependencies on third-party JS libraries using NPM.
* Compile JS from TypeScript
* Enable 'noImplicitAny' in TypeScript
* QuickJump: use JSX syntax
* Generate source maps from TypeScript for easier debugging
* TypeScript: more accurate type
* Separate quick jump css file from ocean theme
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
18ed871afb82560d5433b2f53e31b4db9353a74e switched to a new MathJax URL
but didn't update the tests.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
This was long overdue, now running ./accept.lhs on a clean test from
master will not generate a bunch of changes.
|
|
|
|
| |
These were accidentally left there when the tests were originally added
|
| |
|
|
|
|
|
|
|
| |
This test actually tests what #260 originally reported - I omitted the
RankNTypes scenario from the original fix because I realized it's not
relevant to the underlying issue and indeed, this renders as intended
now. Still good to have more tests.
|
|
This introduces a new precedence level for single contexts (because
implicit param contexts always need parens around them, but other types
of contexts don't necessarily, even when alone)
|