0.11.0

Finally, the so long awaited math mode is here! Enable the option to use it.

This release also includes other improvements and bugfixes, please see the changelog below for more details. Thanks to all contributors that has made possible this release!

Breaking changes

• Change write_to_html to allow fmt::Write by @stepantubanov in #870

New features

• Math mode

Bugfixes

• [0.11] Don't exit scan_attribute with the ix pointing at block quote by @notriddle in #873
• (Re)introduce simd feature to pulldown-cmark-escape by @ollpu in #880
• fix: remove unnecessary end_newline set by @tomcur in #885

New Contributors

• @duskmoon314 made their first contribution in #874
• @stepantubanov made their first contribution in #870
• @tomcur made their first contribution in #885

Full Changelog: v0.10.3...v0.11.0

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.