This fixes an issue where global keypresses were being caught when they
shouldn't when there is a shadow DOM element with focus. This can happen
if the user has included their own extensions. Arguably the extension
should do a stopPropagation, but they don't always do that. I think this
is relatively safe way to approach this.
Fixes https://github.com/rust-lang/mdBook/issues/2507
This fixes an issue where pressing `?` inside the ACE editor was opening
the help popup. The solution here is to ensure that the global keypress
handler works the same as the one used in the search handler.
Fortunately other keypresses like left and right were working OK because
ACE was doing something like stopPropagation (I'm not sure exactly).
Fixes https://github.com/rust-lang/mdBook/issues/3064
These tests have been randomly failing on CI. I'm not entirely certain
why there is a race for these conditions. I'm not sure how
browser-ui-test waits when the document reloads on navigation.
There's also some javascript triggers that happen after the page loads,
and I don't remember if that runs before DOMContentLoaded.
This test seems to be a little flaky. This tries to make it more
reliable by using `wait-for-` commands instead of `assert-`. My guess is
that clicking on an element doesn't necessarily wait for all the
javascript code to finish and the DOM to be done updating.
Fixes CI failure in sidebar-scroll.goml where mdbook-sidebar-scrollbox custom element not found (timing issue in headless Linux Chrome). The class is always present in generated HTML (<mdbook-sidebar-scrollbox class="sidebar-scrollbox">).
This fixes an issue where the sidebar was scrolling incorrectly when
clicking on chapters when a heading nav is currently in view. The
problem was that it was storing the scrollTop of the sidebar, but it was
not considering that when navigating to another chapter that the heading
nav of the previous chapter would disappear.
The solution is to keep of an offset instead so that it can make sure
that the vertical position of the clicked chapter stays in the same
relative position when the new page loads.
Fixes https://github.com/rust-lang/mdBook/issues/2967
This fixes an issue where when a heading goes from a larger to a smaller
level, it was going back too far so that subsequent headings would be at
a lower level than they should have been.
Fixes https://github.com/rust-lang/mdBook/issues/2944
This checks the entire structure of the heading nav for this test. It
currently is not handling nesting properly.
Having a large inline string like this may not be the easiest to
maintain or be able to see the structure. I don't see a way to format it
in goml.
This makes sure that the sidebar headings don't have the `<mark>` tag.
When these are created, the Marker is unable to remove them from the
sidebar (and we don't want them there in the first place).
I suspect we'll want more filtering in the future, but I'm not sure
exactly what to filter. Alternatively, it could have an allow list of
tags, and filter all others out.
This updates the header navigation so that:
- Added a colored bar to break it apart from the chapter navigation.
- Removed the colored circle and just use link color to make it
look cleaner.
This fixes an issue when folding is enabled. The folding was not
properly hiding the sub-chapters because it was assuming it could hide
the next list element. However, the heading nav was the next list
element, so the remaining chapters remained visible.
The solution required some deeper changes to how the chapters were
organized in the sidebar. Instead of nested chapters being a list
element *sibling*, the nested chapter's `ol` is now a *child* of its
parent chapter. This makes it much easier to just hide everything
without regard of the exact sibling order.
This required wrapping the chapter title and the toggle chevron inside a
span so that the flex layout could be localized to just those elements,
and allow the following `ol` elements to lay out regularly.
Closes https://github.com/rust-lang/mdBook/issues/2880
This is no longer used, as individual books have been added to support
different GUI tests. If there is anything here that we later decide we
need to keep for whatever reason, the needed content can be brought back
as new GUI test books. Otherwise, the guide and other books should cover
most things here.
This adds the all-summary GUI test book which can be used for general
purpose tests that need a few pages to exercise all the different kinds
of items.
This adds the ability to use multiple books for the GUI tests. This is
helpful since some tests need special configuration, and sharing the
same book can make it difficult or impossible to test different
configurations. It also makes it difficult to make changes to the
test_book since it can affect other tests.
This works by placing the books in the tests/gui/books directory. The
test runner will automatically build all the books. The gui tests can
then just access the DOC_PATH with the name of the book.
Books are now saved in a temp directory to make it easier to use the
DOC_PATH variable, instead of being tests/gui/books/book_name/book which
is a little awkward.
Following commits will restructure the existing book. This is just a
mechanical move.
This test is failing on CI. I don't know why, as I cannot reproduce
locally. Perhaps it is due to the update to browser-ui-test? Or perhaps
some events from the new nav bar are delaying something?
This adds dynamic navigation of headers of the current page in the
sidebar. This is intended to help the user see what is on the current
page, and to be able to more easily navigate it. The "current" header is
tracked based on the scrolling behavior of the user, and is marked with
a small circle. This includes automatic folding to help keep it from
being too unwieldy on a page with a lot of nested headers.
This includes the `output.html.sidebar-header-nav` option to disable it.
I'm sure there are tweaks, fixes, and improvements that can be made. I'd
like to get this out now, and iterate on it over time to make
improvements.
This adds the ability to pass options to browser-ui-test, which can help
with debugging or doing things like snapshot work. It's maybe not the
cleanest since it doesn't support space-separated args, but should be
good enough.
This changes all HTML IDs so that they have the `mdbook-` prefix. This
should help avoid ID conflicts between internal IDs and IDs from user
content such as section headers.
This is a relatively disruptive change and has a high risk of breaking
something. However, I think I have covered everything, and if anything
is missed, hopefully it will get detected.
I did not change class names since the chance of a collision is much
smaller than with IDs. However, that is something that could be
considered in the future.
Closes https://github.com/rust-lang/mdBook/issues/880
These tests have been flaky, and in general it was probably unwise to
try to rely on an external site like this. I was unable to determine
exactly why the test is failing. The page loads, and then puppeteer
throws an error.
I don't know if it is really feasible to bring these back in some form.
It's probably more effort than it is worth.
Closes https://github.com/rust-lang/mdBook/issues/2765
This adds the ability to redirect URLs with `#` fragments. This is
useful when section headers get renamed or moved to other pages.
This works both for deleted pages and existing pages.
The implementation requires the use of JavaScript in order to manipulate
the location. (Ideally this would be handled on the server side.)
This also makes it so that deleted page redirects preserve the fragment
ID. Previously if you had a deleted page redirect, and the user went to
something like `page.html#foo`, it would redirect to `bar.html` without
the fragment. I think preserving the fragment is probably a better
behavior. If the new page doesn't have the fragment ID, then no harm is
really done. This is technically an open redirect, but I don't think
that there is too much danger with preserving a fragment ID?
The reason the ACE editor was failing to load the rust syntax
highlighting is because the syntax highlighting was being created
*after* the editor was created. If the editor is created first, then ACE
tries to load `ace/mode/rust`. Since it isn't already defined, it tried
to compute the URL and load it manually. However, since the URLs now
have a hash in it (via https://github.com/rust-lang/mdBook/pull/1368),
it was unable to load.
The solution here is to make sure `ace/mode/rust` is defined before
creating the editors. Then ACE knows that it can just load the module
directly instead of trying to fetch it from the server.
Fixes https://github.com/rust-lang/mdBook/issues/2700
