* add external_level and internal_level
* remove unnecessary debug derive on LinkDef
* clarify doc comment about link check levels
* simplify link checker logging
* add missing warn prefix
* simplify link level logging, remove "Level" from linklevel variants
* remove link level config from test site
* switch back to using bail! from get_link_domain
* move console's deps to libs
* remove unnecessary reference
* calling console::error/warn directly
* emit one error, or one warning, per link checker run
* various link checker level changes
* add docs about link checker levels
* remove accidentally committed test site
* remove completed TODO
* Add YAML to formats supported by load_data()
A fairly trivial addition; JSON and YAML are handled so similarly
that this was a matter of copying the JSON-relevant handlers and
editing the copies to handle YAML as well. The test file was
literally generated with 'json2yaml'.
The documentation has been updated to indicate that load_data() now
handles YAML code.
The CHANGELOG has been updated as well.
* After checking, I found that it's generally agreed the mime type is still application/x-yaml.
* Update comment, unify library importing.
I noticed one more place where the list of formats was supported,
and added YAML to that list.
I noticed that there's a singular place to load the `libs::` crate,
and unified by importing of serde_yaml in that place.
* Add `literal` as a new entry for `data source`, to be used by the `load_data` function
* Add tests to the module for plain text, json, xml, toml, and csv
* Update error messaging to include literal as a potential choice
* Update site documentation to include instructions for using `load_data` with a literal
* Make sure that zola serve/build can start from anywhere inside dir tree
* make clippy and rustfmt a bit happier
* replace unecessary if-else with unwrap_or and display which path could not get canonicalized if it fails at startup
* canonicalize config_path to avoid crash when config.toml changes
* Treat _index.LANG.md as section, not page (close#1694)
Display more advanced information in case of a future crash here
* Add link to translated local section in test_site_i18n to catch #1694
It appears linking to another translated page does not trigger the bug,
but linking to a section within the same (translated) page does.
* Allow (and test for) homepage internal links in broken link detection
Co-authored-by: southerntofu <southerntofu@thunix.net>
* Make ignored_content work with directories
Just had to remove strip_prefix(path) from file ignore code.
Added tests for subdirectory globbing.
* Add documentation
* add more tests
to confim that simple filename globs still match paths
without strip_prefix
Currently, init_extra_syntaxtes_and_highlight_themes will supress
loading anything if highlight_theme is "css". This makes sense for the
highlight themes, but will also skip loading any extra syntax
definitions too.
This change defers the theme = css check, so we still load syntaxes in
this case.
Fixes: #1723
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
* templates/load_data: add an optional parameter headers ...
... now `load_data` function supports setting extra headers
* docs/templates/overview: cover some edge-cases in the explanation
* templates/load_data: fix caching logic with headers
* docs/templates: change wording for load_data headers explanations
* Add heuristic checking for HTML anchors
Previously only anchors specified or generated in markdown could be
linked to, without complaint from the link checker. We now use a
simple heuristic check for `name` or `id` attributes.
Duplicate code has been refactored and all XML anchor checks updated
to use regex rather than substring match.
* Fix regexp and refactor
* Split checking for theme template off into separate function
Allows to check for fallbacks elsewhere in the code, without attempting to actually render the template.
* Add template fallback checking to taxonomy pages.
* Add template fallback checking to paginated taxonomies
Requires passing additional arguments to Paginator::from_taxonomy, which may not be desirable.
* Update documentation to reflect taxonomy template fallbacks.
* Update generic taxonomy template names.
* Make check_template_fallbacks() return &str.
* Add tests for check_template_fallbacks
* Next version
* Added tests for shortcode insertion
* Added TOC tests
* Added test for #1475 and #1355
* Basic internal / external links tests
* Added integration test
* Added pseudocode and started on logos
* Logos parsing for shortcodes
* Fixed string literal parsing
Moved string literal parsing to a new lexer in order to have greater
control of control characters which are parsed. This fixes the bug that
was present in the `string_from_quoted` function and also moves the
`QuoteType` to be in the `ArgValueToken`.
* Moved string literal logic to seperate module
* Added square bracket notation for variables
* Error handling rewritten
Remove the Result from the `fetch_shortcodes` function. Added proper
messages within the internal parsing.
* Reorganized and documented the shortcode submodule
* Added all logic for ShortcodeContext spans
* Added working insertion code for MD files
* Made functions generic over Markdown or HTML
* Add check for embedding bodies
* Structure of main function clear
* Added test for `new_with_transforms` function
* It runs!
* Added the code for handling p-ed html shortcodes
* Removed placeholders in markdown function
* Adjusted integration tests
* fetch_shortcodes now also returns a string
* Start of HTML insertion
* Kinda working everything
* Loading of shortcodes and builtins
* Fix tests
* Some missed fixes
* Tweaks + fmt
* Remove string literal handling
* Fix benches
* Grab shortcode def only once per site
* Fix benches
* Rewrite of parser
* Fix tests
* Add test for #1655
* Re-enable integration test
* Add test for #1601
* Add test for #1600
* Add test for #1500
* Add test for #1320
* Fix test on windows?
Co-authored-by: Gijs Burghoorn <g.burghoorn@gmail.com>
* Pass lang to shortcodes context
* Add tests for lang in shortcodes
* Lang is passed to anchor-link.html template
* Document passing lang to shortcodes/anchor-link.html
Add a test to make sure lang can be overriden by passing an explicit argument to the shortcode,
for usage in the markdown filter where the language context is not available.
* Update docs for more information on shortcodes+i18n+markdown() filter
Co-authored-by: southerntofu <southerntofu@thunix.net>
* Consider the site's output path in search_for_file
The search_for_file helper function now accepts an optional
output path. If passed, the file will also be searched there.
This is used in the get_url function to search in the
Site::output_path.
In practice, this means cachebust works for files in the
output path.
* Make output_dir required in search_for_file
* Update docs for file searching logic
* Add test for new file searching behavior
* cargo/manifests: allow user to use native-tls ...
... if `ring` can't be used on the user's platform (e.g. mips/ppc/riscv)
* CI: test for native tls build
* Add class based syntax higlighting + line numbers
* Use fork of syntect for now
* Fix tests
* Fix diff background on inline highlighter
Co-authored-by: evan-brass <evan-brass@protonmail.com>
* Add `num_format` filter for displaying formatted numbers
* Register the filter
* Update docs
* Make `locale` argument required
* Revert "Make `locale` argument required"
This reverts commit 9cdbf28591.
* Pull the default locale from the site config
* Add note about defaults to the docs
* Add missing borrow
* Move image loading code used by get_image_metadata from templates to imageproc
* resize_image: Compute image resize op beforehand and return metadata to caller
* get_image_metdata: Use a serializable response struct
* imageproc: Add integration tests
* imageproc: Assume webp is lossy
* get_image_metadata: Use webp crate for decoding WebP, image is still not very good at it
* imageproc: Test read_image_dimensions in integration tests
* imageproc: Rename read_image_dimensions() as read_image_metadata()
* imageproc: Fix a regression in hash computation
* imageproc: Don't hardcode hashes in integration tests, pattern match them instead
* imageproc: Style fixes
* imageproc: Fix integration tests on Windows (hopefully)
* Add new sorting: UpdateDate that sorts by `meta.updated`.
* Use `max(created, updated)` for sort-by-update-date
This prevents created but never updated articles from appearing at the end/not
appearing at all.
* Fixed failing tests on windows when user is not VssAdministrator.
* Fixed windows specific testcases related to \r
* Added the ability to perform POST requests to load_data
* make tests on windows deal with both \r being there on windows, and \r not being generated as on my personal windows system.
* undo earlier commit eaaa8c3ddd
because it fails on azure buildserver
* added new arguments to the hash for the cache function.
So caching now works as it should
* added new arguments to the hash for the cache function.
* improved documentation of load_data POST with better example.
* added basic derive traits
* changed load_data param contenttype to content_type
* fixed caching issues that went missing?
* format
* made code more idiomatic as suggested by keats
* link_checking: prevent rate-limiting
Fix for https://github.com/getzola/zola/issues/1056.
- assign all links for a domain to the same thread
- reduce number of threads from 32 to 8
- add sleep between HTTP calls
* Add get_link_domain(), use for loops
* Do not sleep after last link for domain
* Avoid quadratic complexity
* remove prints
This updates rendered markdown structures in order to keep track
of all internal links, not anymore limiting to only those targeting
an explicit anchor fragment.
The goal of this rework is to allow building other features, such
as backlinks, on top of the existing collection of internal links.
* fix can not build zola with search/indexing-zh feature
* fix can not build components/utils after enabled indexing-zh feature
error[E0252]: the name `Deserialize` is defined multiple times
--> components/utils/src/de.rs:2:5
|
1 | use serde::{Deserialize, Deserializer};
| ----------- previous import of the macro `Deserialize` here
2 | use serde_derive::Deserialize;
| ^^^^^^^^^^^^^^^^^^^^^^^^^ `Deserialize` reimported here
|
= note: `Deserialize` must be defined only once in the macro namespace of this module
* fix: webp quality level mismatch
* fix: hash collision using the same image ops
Using the same image operations, but for different formats, e.g. `.jpg` and
`.webp`, produced the same hash.
To differentiate between these, the image extension is added to the hash.
* Translations are also generated for the index page (fix#1332)
* More tests for translations
* Even better error message
* Update page count for test
* Patch to fix Windows tests
By @mtolk
Co-authored-by: southerntofu <southerntofu@thunix.net>
* Allow optional whitespace around dash/underscore in filename
Allow file names that are as follows:
2021-01-01 - test.md
To be parsed the same as if they were
2021-01-01-test.md
The slug for both will now just be "test" instead of previously the
first example would have become "2021-01-01-test".
* Add documentation for optional whitespace in filename
* Test that updated regex does not take space after dash
Without content it should be possible to create a file that contains no
newlines after the frontmatter at all, rather than forcing a newline
after the frontmatter.
* Add support for base64-encoded hash values
The global template function 'get_file_hash' can now return a
base64-encoded hash value when its 'base64' parameter is set to true.
See discussion in #519.
* Fix integrity attribute's value in test site
SRI hash values must be base64-encoded.
* Update documentation about 'get_file_hash'
* Fix 'can_get_hash_for_static_files' unit test
* Move `load_tera` to `templates`
I don't know if this is a good place for it, conceptually. I'm moving it
there because I need to use it from `templates`, and `templates` can't
depend on `site`, because there's already a dependency in the opposite
direction.
* Load templates in `markdown` filter
This enables the `markdown` filter to handle shortcodes, as long as
those shortcodes don't access any context variables.
Addresses #1350
* Update documentation of `markdown` filter
* Only load templates for `markdown` filter once
* Clarify `markdown` filter documentation
This is a lightly edited version of what @southerntofu suggested.
* load_data() template function takes a `required` boolean flag
* Update tests for load_data()
* Add test to make sure invalid data always fails in load_data
* Better documentation, fixing a few typos
Co-authored-by: southerntofu <southerntofu@thunix.net>
* Add sort_by=title
* Remove old comment.
* Remove println! debugging
* Minor: text spacing
* Use lexical_sort crate for sort_by title
Co-authored-by: David James <davidcjames@gmail.com>
* Internal links are resolved in tera markdown filter (close#1296#1316)
* Add a test for internal links in markdown filter
Co-authored-by: southerntofu <southerntofu@thunix.net>
* Highlight fallback from extra syntaxes to the theme (close#1309)
* Warning when codeblock language is unknown and cannot be highlighted
* page/section path in codeblock language missing warning
Co-authored-by: southerntofu <southerntofu@thunix.net>
* add shortcode 'invocation' variable to allow a shortcode to track how many times it has been invoked in a given Markdown file
* use closure (implicit struct) instead of explicit struct for invocation tracking
* update variable name to "nth"
* Accept dates presented as strings
Still, if we find a string we involve the TOML parser to make sure the
content of said string would be a correct TOML date. In a pure TOML
world this isn't exactly relevant, but this will come in handy when
using a YAML deserializer.
Signed-off-by: Kevin Ottens <ervin@ipsquad.net>
* Add serde_yaml and test-case dependencies
This will be necessary for the YAML frontmatter handling and
corresponding unit tests.
Signed-off-by: Kevin Ottens <ervin@ipsquad.net>
* Add YAML front matter handling
Signed-off-by: Kevin Ottens <ervin@ipsquad.net>
* Switch RawFrontMatter enum to wrap &str instead of String
Signed-off-by: Kevin Ottens <ervin@ipsquad.net>
* Update the documentation to mention YAML frontmatter
This is just a light update on purpose. There would be no point in
pushing YAML too much, this is mainly here to help people with a backlog
of posts to transition.
Signed-off-by: Kevin Ottens <ervin@ipsquad.net>
* Add support for loading Bibtex data.
* Add load_data() documentation for the bibtex format
* Force bibtex tags to be lower case.
Bibtex tags are case-insensitive, and this works around tera's case-sensitiveness.
* Improve the load_data() documentation for the bibtex format
* Next version
* Remove lines forcing minify_html to false
* Update documentation about minify_html
* Update changelog
Co-authored-by: Vincent Prouillet <balthek@gmail.com>
This introduces `relative-path`, a crate I've written for the specific
purpose of providing platform-neutral operations over paths the same way
they are used in URLs.
This means that `///hello///` == `/hello`, which should do the same as
the existing stripping minus the platform-specific path separators
causing the [bug being referenced](#1169).
* Allow site path to contain underscores
Fixes site.css is not being generated if any part of the path contains
underscores
* Add tests for path with underscores
* Add line highlighting to code blocks
* Fix highlighting of lines
Apparently every line to be highlighted is provided in one chunk.
* Add more documentation to codeblock.rs
* Turn FenceIter into an Iterator
* Move Range to fence.rs
* Add tests
