Update dependency floki to v0.38.3

[renovate]

This PR contains the following updates:

Package	Type	Update	Change
floki (source)	dev	minor	0.34.2 → 0.38.3

---

Release Notes

philss/floki (floki)

v0.38.3

Compare Source

Fixed

• Fix a regression with Floki.text/2 when a document contains a "processing instruction" tag.
  This is the case for a XML tag.

v0.38.2

Compare Source

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.

v0.38.1

Compare Source

Performance

This version contains major performance improvements in the following functions:

• Floki.filter_out/2.
• Floki.find/2 - with some improvements to specific selectors, like classes
  and attribute selectors.
• Floki.text/2.

Those functions are not only faster, but are now using less memory. Please check
the PRs related to this release if you want to better understand the numbers.

• Speed up do_classes_matches? - #​649
• Make filter_out faster - #​650
• Speed up attribute lookup - #​651
• Enable tuple traversal optimization for multiple selectors - #​652
• Optimize HTMLTree.to_tuple conversion using Enum.reduce - #​657
• Optimize Finder.get_descendant_ids/2 memory usage and speed - #​660
• Optimize Finder.get_siblings/2 memory usage and speed - #​663
• Optimize FlatText.get/3 memory usage and speed - #​664
• Optimize class matching in Floki.Selector - #​665

All the improvements in this version were made by Barna Kovacs - @​preciz,
so shout out and thanks to him!

Fixed

Remove a warning about an unused require Logger that pops up when using Elixir v1.20.

v0.38.0

Compare Source

Added

• This version adds initial support for the :has pseudo-selector.
  It is a great addition that enables finding elements containing
  matching children.

  Examples for selectors:

  • "div:has(h1)"
  • "div:has(h1, p, span)"
  • "div:has(p.foo)"
  • "div:has(img[src='https://example.com'])"
  • "tr:has(*:fl-contains('TEST'))"

  Note that combinators like ">" are not allowed yet.

  Thank you @​bvobart for this feature!

Fixed

• Add :style option documentation to Floki.text/2.
  Thanks @​s3cur3 for the fix.

• Fix deprecation warnings for upcoming Elixir 1.19.

• Prevent from crashing when selector is an empty string.

Removed

• Remove support for Elixir 1.14 and OTP 23.

• Remove deprecated functions and function clauses
  that were accepting strings (binaries).

  Affected functions:

  • parse/1 - removed function
  • map/2 - removed function
  • attr/4 - removed clause
  • find/2 - removed clause
  • text/3 - removed clause
  • text/3 - removed clause
  • attribute/2 - removed clause
  • filter_out/2 - removed clause

• HTML must be parsed before searching. Functions like Floki.find/2,
  Floki.attribute/2, and other HTML manipulation functions no longer work
  directly with HTML strings. The HTML must be parsed first using
  Floki.parse_fragment/2 or Floki.parse_document/2.

  Before:

  html = "<div class='foobar'><p>Hello</p></div>"
  Floki.find(html, "p")
  Floki.attribute(html, "div", "class")

  After:

  html = "<div class='foobar'><p>Hello</p></div>"
  parsed_html = Floki.parse_fragment!(html)
  Floki.find(parsed_html, "p")
  Floki.attribute(parsed_html, "div", "class")

v0.37.1

Compare Source

Fixed

• Move regex declaration from module tag to inside function.
  This is a fix to be compatible with the upcoming OTP 28.

v0.37.0

Compare Source

Added

• Add Floki.css_escape/1 - thanks @​SteffenDE.

Fixed

• Fix bug propagating identity encoder in raw_html/2 - thanks @​andyleclair.

Removed

• Remove support for Elixir 1.13 and OTP 22.

v0.36.3

Compare Source

This release contains some performance improvements, thanks to @​ypconstante.

Fixed

• Stop Floki.get_by_id/2 traversal on first match. Thanks @​ypconstante.

• Remove extra whitespaces from nodes without attributes on Floki.raw_html/1.
  Thank you @​ypconstante.

• Fix Floki.raw_html/1 typespecs. Thanks @​davydog187.

v0.36.2

Compare Source

Added

• Implement the Inspect protocol for the Floki.HTMLTree struct.
  This struct is currently private. Thank you @​vittoriabitton.

Fixed

• Fix regression to respect config option :encode in Floki.raw_html/2.
  Thanks @​Sgoettschkes.

• Make the Floki.raw_html/2 treat the contents of the <title> tag as plain text.
  The idea is to align with parse_document/2.
  Thank you @​aymanosman.

v0.36.1

Compare Source

Fixed

• Fix typespec of get_by_id/2.

v0.36.0

Compare Source

Added

• Add Floki.get_by_id/1 that returns one element by ID or nil.
  Thanks @​SteffenDE.

Changed

• Improve options validation with Keyword.validate!/2.
  This is not a change in APIs, but the error messages and opts validation
  should be standardized now.
  Thanks @​vittoriabitton.

Removed

• Drop support for Elixir v1.12.

v0.35.4

Compare Source

Besides the fix described below, this release also contains more performance
improvements, thanks to @​ypconstante.

Fixed

• Fix order of results for Floki.find/2. This was a regression from the
  previous version - thanks @​ypconstante.

v0.35.3

Compare Source

This release has great performance improvements, thanks to the PRs
from @​ypconstante!

Most of the main functions, such as Floki.raw_html/2 and Floki.find/2 are
faster and are using less memory. It's something like twice as fast, and half
usage of memory for find/2, for example.

Fixed

• Add :leex to Mix compilers. Fixes the build when running with dev version of Elixir.
  Thanks @​wojtekmach.

• Fix Floki.raw_html/2 when a tree using attributes as maps is given.
  Thanks @​SupaMic.

• Add a guard to Floki.find/2 so people can have a better error message when an
  invalid input is given. Thanks @​Hajto.

• Fix parsers to consider IO data as inputs. This may change in the next version
  of Floki, as I plan to drop support for IO data.
  Thanks @​ypconstante.

Removed

• Remove outdated Gleam wrapper code. The external functions syntax in Gleam
  has changed. So now
  the wrapper is not needed anymore.
  Thanks @​michallepicki.

v0.35.2

Compare Source

Fixed

• Enable usage of IO data by removing a guard for binaries in the main
  parser module.

v0.35.1

Compare Source

Fixed

• Fix a small regression of the mochiweb parser that was breaking when a
  malformed HTML was used. For more details, see the original issue:
  #​492

v0.35.0

Compare Source

Added

• Add support for parsing attributes as maps.

  This makes parse_document/2 and parse_fragment/2 accept the option
  :attributes_as_maps to change the behaviour and return attributes as
  maps instead of lists of tuples.
  The only parser that does not support it yet is the fast_html.

Changed

• Drop support for Elixir v1.11.

• Change the log level of parsing logger calls from "info" to "debug".
  This will help to reduce the amount of noise in production apps.

v0.34.3

Compare Source

Added

• Add boolean option :include_inputs to Floki.text/2 that changes the result
  of this function to include the values of inputs. So if there is any input with
  a "value" attribute, we now include that value if this option is set to true.
  Thanks @​viniciusmuller.

Fixed

• Fix find of elements by classes that contain colons. This is useful for when
  people are trying to find elements that contain Tailwind classes.
  Thanks @​viniciusmuller.

• Fix some typespecs that were using types from private modules. This is a fix
  to the documentation.

---

Configuration

📅 Schedule: (UTC)

• Branch creation
  • At any time (no schedule defined)
• Automerge
  • At any time (no schedule defined)

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.