v0.38.2

Performance

This is another juicy patch version with performance improvements made by @preciz.

• Optimize multiple selectors deduping - #667
• Optimize DeepText by using plain recursion - #668
• Optimize get_descendant_ids by removing Enum.reverse calls - #669
• Optimize HTMLTree.to_tuple_list/1 using tail recursion - #670
• Optimize Floki.TextExtractor by removing Enum iterations - #675
• Simplify class attribute matching - #677
• Refactor attribute_values to use for comprehension with pattern matching - #679
• Optimize CSS class matching to avoid list subtraction - #680
• Optimize iodata by avoiding empty separators in DeepText and FlatText - #682
• Simplify HTMLTree.build/1 by avoiding redundant build_tree calls - #683
• Optimize Floki.text/2 by extracting text in a single pass - #684
• Optimize HTMLTree updates and deletions - #687
• Refactor attribute retrieval functions for a 1.2-1.5x speedup - #689
• Optimize case-insensitive attribute includes match - #690
• Optimize selector traversal - #611

Please check the pull requests to see the improvements.

Fixed

• Fix compiler warnings for the upcoming Elixir v1.20.
• Fix typespecs of Floki.attribute/3.
• Fix documentation for some functions.

All pull requests

• Require Elixir v1.16 by @philss in #666 (reverted for this release)
• Optimize multiple selectors deduping by @preciz in #667
• Optimize DeepText by using plain recursion by @preciz in #668
• Optimize get_descendant_ids by removing Enum.reverse calls by @preciz in #669
• Optimize HTMLTree.to_tuple_list/1 using tail recursion by @preciz in #670
• Fix docs inaccuracies by @preciz in #671
• docs: update Mix.Config to import Config in HTMLParser docs by @preciz in #673
• docs: remove :html_parser option from Floki.text/2 docs by @preciz in #674
• Fix typespecs: remove binary from Floki.attribute/3 typespec by @preciz in #672
• Optimize Floki.TextExtractor by removing Enum iterations by @preciz in #675
• Fix Elixir 1.20 compiler warnings for bitstring match pin operators by @preciz in #676
• Simplify class attribute matching by @preciz in #677
• Bump credo from 1.7.17 to 1.7.18 by @dependabot[bot] in #678
• Refactor attribute_values to use for comprehension with pattern matching by @preciz in #679
• Optimize CSS class matching to avoid list subtraction by @preciz in #680
• Remove unused charref_html_after_buffer/2 clause by @preciz in #681
• Optimize iodata by avoiding empty separators in DeepText and FlatText by @preciz in #682
• Simplify HTMLTree.build/1 by avoiding redundant build_tree calls by @preciz in #683
• Optimize Floki.text/2 by extracting text in a single pass by @preciz in #684
• Optimize HTMLTree updates and deletions by @preciz in #687
• Refactor attribute retrieval functions for a 1.2-1.5x speedup by @preciz in #689
• Optimize case-insensitive attribute includes match by @preciz in #690
• Bump html5ever from 0.17.0 to 0.18.0 by @dependabot[bot] in #686
• Optimize selector traversal by @preciz in #691
• Bump jason from 1.4.4 to 1.4.5 by @dependabot[bot] in #692
• Bump ex_doc from 0.40.1 to 0.40.2 by @dependabot[bot] in #693
• Release v0.38.2 by @philss in #694

Full Changelog: v0.38.1...v0.38.2