v0.10.3

The main change of this release is the simd feature was not being used in the escape functions since the version 0.10 due to a mistake during the separation of the crate pulldown-cmark-escape.

The crate pulldown-cmark-escape has been updated to the version 0.10.1.

What's Changed

• fix: fix dead code warning from nightly compiler by @rhysd in #876
• Eat all spaces after line break in link title by @notriddle in #877
• Mark Rule as a block item by @notriddle in #879

Full Changelog: v0.10.2...v0.10.3

v0.10.2

New release with some fixes and improvements. Note the 0.10.1 is missing (yanked from crates.io) due to a conflict with the clap version and the Rust minimum version (1.74 now instead of 1.70).

Thanks to all people that contributed to this release!

What's Changed

• Limit link def expansion by @notriddle in #845
• Do not look for HTML tags that start with backslash by @notriddle in #849
• Count a blank line at end of indented code block towards list by @notriddle in #851
• Use same limit for refdef as inline links by @notriddle in #854
• Don't exit scan_attribute with the ix pointing at block quote by @notriddle in #858
• Check indentation on the closing fence relative to the line by @notriddle in #862
• Adjust strikethrough flanking rule to better fit Rustdoc Crater run by @notriddle in #864
• perf: cargo-wizard default recommendations for runtime perf by @Martin1887 in #868

New Contributors

• @ehuss made their first contribution in #848
• @jimblandy made their first contribution in #865
• @max-heller made their first contribution in #866
• @blinxen made their first contribution in #875

Full Changelog: v0.10.0...v0.10.2

v0.10.0

This is a huge release after a a long time without a major release, with many bugfixes and several new features. Thanks to all people involved!

Breaking Changes

• feat!: allow custom headings attributes with optional value. Fixes #634 by @Martin1887 in #638
• feat!: metadata blocks. Fixes #580 by @Martin1887 in #641
• Make html and escape modules optional by @jfrimmel in #519
• feat: separate escape.rs in another crate by @Martin1887 in #838
• Allow dynamically generated header ids and classes by @SergioBenitez in #614
• Add separate enum for Tag ends by @marcusklaas in #517
• add link identifiers into emitted tags by @Gozala in #436
• Add a separate mode to parse footnotes the same way GitHub does by @notriddle in #654
• feat!: separate inline HTML and HTML blocks by @Martin1887 in #692
• Use generics for broken link callback by @ollpu in #701

New Features

• Add task list with a list item's first paragraph by @marcusklaas in #558
• feat: TextMergeStream utility to merge consecutive text events by @Martin1887 in #686
• Reads markdown from file by @phoepsilonix in #726
• feat(utils): text events merger with offsets by @Martin1887 in #743
• Add reject-broken-links CLI option by @PoignardAzur in #807
• feat: convert Tag to_end() and From impl by @Martin1887 in #824
• feat(escape): separate attribute and text HTML by @notriddle in #830
• Update to CommonMark 0.31.2 by @notriddle in #834
• Backport footnote handling without needing blank lines between them for old footnote format by @GuillaumeGomez in #836
• Implement missing cell cutoff to stop quadratic results by @notriddle in #837

Bugfixes

• Fix parser bug where link label gets broken by ] in code span by @lukas-code in #643
• fix: strikethrough can be intra-word by @Martin1887 in #648
• fix: shift right overflow panic in debug mode in rare cases by @Martin1887 in #652
• fix: tables can interrupt paragraphs by @Martin1887 in #653
• fix(parse): rare error searching in the emphasis stack by @Martin1887 in #650
• Skip leading space on leading pipe in scan_table_head by @notriddle in #665
• Fix index out of bounds when parsing GFM footnote by @mgeisler in #667
• Allow lowercase letters in HTML declarations by @notriddle in #668
• Fix fuzzing empty list items by @mgeisler in #669
• Fix, and document, some table/paragraph interrupt corner cases by @notriddle in #672
• Include trailing backslash at all break points in paragraph by @notriddle in #676
• Block quotes interrupt paragraphs even with no space at start by @notriddle in #675
• Fix table/paragraph interruption when nested by @notriddle in #679
• Use correct both flag when pushing remaining delimiters to stack by @notriddle in #682
• Fix #684 by @Martin1887 in #685
• Fix spec violation on empty-list-item-cannot-interrupt-paragrph rule by @notriddle in #681
• fix: empty text events in ATX headings with trailing # by @Martin1887 in #683
• Fix incorrect scan of bad email address by @raphlinus in #688
• Fix infinite loop with metadata blocks with indented lines by @Martin1887 in #690
• Add special handling for backslashed table pipes by @notriddle in #699
• Fixes rustdoc::bare_urls lint by @GuillaumeGomez in #719
• Fix elided lifetimes and unused lifetimes errors by @GuillaumeGomez in #723
• Allow arbitrary blank lines within list by @notriddle in #728
• Single pipes alone on lines are not tables by @notriddle in #732
• Use original run lengths to judge multiples of three by @notriddle in #730
• Fix oversight in table escape parsing by @notriddle in #736
• Check that parens in link def are nested by @notriddle in #738
• Fix emphasis stack truncation bug by @notriddle in #740
• Fix emphasis run length delimiter lower limit bug by @notriddle in #742
• fix: compilation error in utils.rs after merge 701 by @Martin1887 in #746
• Track underscore lower bound by % 3 modulo by @notriddle in #748
• Parse [^] as a link by @notriddle in #750
• Do not subtract list item indent if it has more than one blank line by @notriddle in #752
• Do not parse hard line breaks when doing link label blocks by @notriddle in #754
• Ignore block synstax when determining emphasis can_open/close by @notriddle in #756
• Allow up to three spaces indentation before footnote definition by @notriddle in #759
• Do not accept footnote definitions with line breaks by @notriddle in #761
• Apply laziness rules to link definition by @notriddle in #765
• When popping a list, clear begin_list_item state by @notriddle in #763
• Parse footnote definitions below the first line of nested blocks by @notriddle in #767
• Parse settext headers with higher priority than linkdefs by @notriddle in #769
• Clean up some more gratuitous differences in footnote parsing by @notriddle in #773
• Parse inline nested footnote definitions by @notriddle in #771
• Do not treat backslash at end of block as hard break by @notriddle in #775
• Correct instructions for running specs by @ollpu in #777
• Read lazy paragraph continuations off of refdefs by @notriddle in #782
• Allow setext headers to interrupt link definitions by @notriddle in #780
• Fix incorrect space handling at paragraph/linkref boundary by @notriddle in #783
• Fix handling of newlines in nested inline link by @notriddle in #784
• Allow setext headers to interrupt link def insides by @notriddle in #790
• Avoid updating linkdef when title isn't followed by NL by @notriddle in #792
• Trim spaces off start and end of link reference by @notriddle in #794
• Don't eat blank line when parsing unclosed fenced code by @notriddle in #796
• Do not allow inlines to cross over block structures by @notriddle in #798
• Fix parsing of spaces in code by @notriddle in #786
• Include last of line of link title by @notriddle in #809
• Calculate correct list indent for ordered lists by @notriddle in #811
• Do not treat zero-cell rows as valid by @notriddle in #812
• Trim spaces off end of parsed lines by @notriddle in #785
• Fix up corner cases in table and code unescaping by @notriddle in #776
• Reject table delimiter row with unexpecte...

0.9.6

cargo build fails when --locked is passed for 0.9.4 and 0.9.5 #833

0.9.5

Set range end to Paragraph if returned early because of footnote definition #831

0.9.4

• Allow footnote definitions to not be separated by blank lines #825
• chore: Update bitflags from 1.3.x to 2.x #694

0.9.3

This patch prepares the release of the minor version 0.9.3. It's forked off 5c7881c and includes a cherry-pick of 48b5931.

Changelog

Parser

Even though this is a minor release, it changes changes how some inputs are parsed. These changes are made to comply with the CommonMark and GitHub Flavored Markdown specifications.

First, pulldown-cmark now supports strikethrough with a pair of single tildes in addition to double tildes, if the strikethrough option is enabled. For example, ~this~ is now equal to ~~this~~ and will be parsed like this.

The second change fixes a bug in the parser where a specific edge case with brackets and inline code ([`]`]) was incorrectly parsed as a link. This change should not affect most users of pulldown-cmark.

impl Debug for Everything

All public types in the pulldown-cmark library now implement the Debug trait.

PRs included in this release

• #604
• #611
• #566
• #621
• #643

New Contributors

• @djc made their first contribution in #604
• @gifnksm made their first contribution in #611
• @joelparkerhenderson made their first contribution in #566
• @rhysd made their first contribution in #621

Full Changelog: v0.9.2...v0.9.3

v0.9.2

This release includes fixes for a few panics and other minor bugs.

0.9.1

Fixes minor parsing bug in nested lists.

0.9.0

This release brings a number of changes.

New features

• Thanks to @lo48576, pulldown now optionally supports custom header ids and classes for headers. Set ENABLE_HEADING_ATTRIBUTES in the options to enable.
• Users can now access reference definitions, information that was previously only exposed internally.
• Pulldown is now CommonMark 0.30 compliant.

Changes

• The function signature for the broken link callback has changed slightly to allow for FnMut functions.

There have also been a number of (small) parsing bug fixes.