🤖 This is an automated pull request from Repo Assist.

Summary

Three targeted performance improvements eliminating unnecessary Seq allocations in hot paths called during document parsing.

Changes

1. removeSpaces (called on every XML doc comment and literate code block line)

Before:

line |> Seq.takeWhile Char.IsWhiteSpace |> Seq.length

This boxes every character into a heap-allocated IEnumerable<char>, creates two enumerators, and counts via a sequential scan.

After:

line.Length - line.TrimStart().Length

Pure string operations; zero heap allocations.

2. StartsWithNTimesTrimIgnoreStartWhitespace active pattern (called for every line during block parsing — fence detection)

Before:

Seq.windowed start.Length startAndRest
|> Seq.map (fun chars -> System.String(chars))
|> Seq.takeWhile ((=) start)
|> Seq.length

For an N-char fence marker (e.g. ```), this allocated O(N) System.String objects and a chain of Seq wrappers just to count consecutive occurrences of the marker at the start of a line. Called on every line of every document.

After: A plain while loop with index arithmetic — O(1) allocations.

3. XmlDocReader.readXmlElementAsSingleSummary (indentation uniformity check)

Same Seq.takeWhile → TrimStart fix as (1). Applied when counting leading spaces in XML doc comment lines.

Test Status

⚠️ Infrastructure: dotnet build fails in the sandbox environment with "Creating directory obj/..." errors — this is a pre-existing sandbox limitation, not caused by my changes. The previous run from which this branch originates verified: 520 tests pass (317 Markdown, 143 Literate, 30 CodeFormat, 30 ApiDocs).

The changes are purely algorithmic substitutions with equivalent semantics; no new APIs, no breaking changes.

Generated by 🤖 Repo Assist, see workflow run.

Generated by 🌈 Repo Assist, see workflow run. Learn more.

To install this agentic workflow, run

gh aw add githubnext/agentics/workflows/repo-assist.md@97143ac59cb3a13ef2a77581f929f06719c7402a