Optimize applies_to parsing

[reakaleek]

Changes

Extract applies_to yaml parsing into its own class and reuse it in all the places.

Also, add memoization so we don't parse the same input over and over again.

[Copilot]

Pull Request Overview

This PR optimizes the parsing of applies_to YAML definitions by introducing a centralized parser with memoization capabilities. The change eliminates code duplication and improves performance by caching parsed results.

Key Changes:

• Created a new ApplicableToParser class that provides centralized YAML parsing with memoization
• Replaced direct YamlSerialization.Deserialize calls across multiple files with the new parser
• Updated test expectations to reflect changes in HTML output formatting

Reviewed Changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
src/Elastic.Documentation/AppliesTo/ApplicableToParser.cs	New parser class with memoization using ConcurrentDictionary
src/Elastic.Markdown/Myst/Roles/AppliesTo/AppliesToRole.cs	Updated to use centralized parser instead of direct deserialization
src/Elastic.Markdown/Myst/Directives/DirectiveHtmlRenderer.cs	Replaced local parsing method with centralized parser
src/Elastic.Markdown/Myst/Directives/AppliesSwitch/AppliesSwitchBlock.cs	Simplified sync key generation using new parser
src/Elastic.Markdown/Myst/CodeBlocks/EnhancedCodeBlockParser.cs	Updated applies_to directive processing to use new parser
tests/Elastic.Markdown.Tests/Directives/ApplicabilitySwitchTests.cs	Added import for new AppliesTo namespace
tests/authoring/Inline/AppliesToRole.fs	Updated test expectations for HTML output format changes

[Mpdreamz]

Conflicted about memoization here.

Deserialization should be fast (the serialization routines get source generated at build time) and these objects are typically short lived so should be GC'd away quickly too.

This does keep fewer instances around but I wonder how many cache hits we actually get with folks writing the yaml in different orders.

Applicability.TryParse() might be worth memoizing since it has a much higher chance of hitting the cache. But even there there are so many version lifecycle permutations around memoizing might eat up more memory then its worth easing GC pressure.