When looking for related posts, Jekyll was indexing `Jekyll::Post`
objects, but finding related posts based on `Jekyll::Post#content`. This
caused two problems:
1. Ruby 2.2 will warn on == if <=> throws an exception (and future Ruby
versions will surface that exception). Because `String`s can't be
compared with `Jekyll::Post`s, this warning was appearing all the time
while searching for related posts.
2. LSI won't return a post itself when searching for related posts. But
LSI could never tell that we were searching on a post, since Jekyll
passed post content, not a post object. With this fix, we can remove the
`- [post]` from `Jekyll::RelatedPosts#find_related`.
This is a more accurate fix for #3484.
Clean up the destination modified check in `source_modified_or_dest_missing?` to be easier to read. Note that it can now return `nil` instead of `false` for an unmodified `source_path` and a `nil` `dest_path`, but in a discussion on 706007ead9 we decided that was okay.
When adding a dependency, also add the dependency to the metadata hash.
Addresses part 1 of #3591. Prior to this fix, the regnerator only paid attention the mtime of the first dependency it checked, so for posts/pages with N multiple dependencies (i.e., every layout file used to render them), it continues to regenerate the post/page approximately N times, at which point it's seen all of the dependencies.
- Replaced occurrences of #array += with concat
operations.(performance)
- Corrected alignment.
- Removed rebase artifact.
Signed-off-by: Martin Rogalla <martin@martinrogalla.com>
To address part of #3591, clear the regenerator's cache every time the
site is processed. This ensures that the regenerator doesn't incorrectly
believe a file hasn't changed based on stale information.
Organized the draft, post and layout reader into the *readers* classes.
Fixed all references and ran tests.
Signed-off-by: Martin Jorn Rogalla <martin@martinrogalla.com>
After carefully looking at these two methods, as of right now they do not
belong in the reader, as they should also be used by the writer. Thus the
decision was made to move them back into the class containing the source
and dest fields, site.rb.
Signed-off-by: Martin Jorn Rogalla <martin@martinrogalla.com>
Extracted `in_source_dir` from site.rb into reader.rb.
Updated all the references and tests.
Signed-off-by: Martin Jorn Rogalla <martin@martinrogalla.com>
This updates the default permalink style for pages and collections to
match the site-wide 'permalink' setting. If the permalink setting
contains a trailing slash, either explicitly or by being set to
':pretty', then pages and collections permalinks will contain trailing
slashes by default as well. Similarly, if the permalink setting
contains a trailing ':output_ext', so will pages and collections. If
the permalink setting contains neither a trailing slash or extension,
neither will pages or collections.
This impacts only the default permalink structure for pages and
collections. Permalinks set in the frontmatter of an individual page
take precedence, as does the permalink setting for a specific
collection.
Fixes#2691
- Single Quotes
- Fixed Typo's in variable names.
- Removed Redundant Escaping in Regular Expressions.
Signed-off-by: Martin Jorn Rogalla <martin@martinrogalla.com>
This ensures that destination files for HTML posts, pages and
collections always include the proper file extension (as defined by
output_ext) regardless of permalink structure. This allows for URLs
that contain no extension or trailing slash to still result in proper
destination files with .html extensions.
Because this change relies so heavily on output_ext accurately
identifying the extension of the destination file, this change also
removes the feature test that tested support for permalinks with a .htm
extension. In order to support alternate file extensions, a future
patch or plugin will need to modify the output_ext value, at which point
everything else should work as expected.
- Added a test to check if the sort filter will raise the correct
exception on given nil input.
- Improved error message and used "nil" consistently.
Signed-off-by: Martin Jorn Rogalla <martin@martinrogalla.com>
Sort will now throw an error when a nil object array is given as input.
See issue #3491 for more information.
Signed-off-by: Martin Jorn Rogalla <martin@martinrogalla.com>
For a full rebuild, we certainly don't want to *read* from
.jeykll-metadata, but we should still write it. Otherwise, a subsequent
incremental build would have to do a full rebuild again since there is
no metadata file to start from.
Found by @EdMcBane in https://github.com/jekyll/jekyll/pull/3435
The strange regexp we were doing to replace the <pre><code></pre></code>
bits in the Pygments output were wreaking havoc on Rouge output
because Rouge uses <pre>'s to wrap line numbers.
To be consistent, the output from render_* should *not* include
the wrapping <div> and <pre> tags. It should just be what was
inside. We can then wrap it in our own custom tags without using
any regular expressions, as God intended. Death to regular
expressions and HTML manipulation!
This shows people how to kill Jekyll without knowing the PID using `pkill` (you could also do `kill -9 $(pgrep -f jekyll)` but that is just the long way around doing `pkill -f` so it shouldn't be shown in the Jekyll logger but can be documented here for people.
Justin Weiss
Jordon Bedwell
Nicholas Burlett
Martijn den Hoedt
Parker Moore
Martin Jorn Rogalla
Matt Rogers
ChaYoung You
Will Norris
Philipp Rudloff
Joop Aué
Tony Eichelberger
Ryan Tomayko
Alfred Xing
Mark Tareshawty
Tanguy Krotoff
eksperimental
Seth Warburton