Reduce ParseWorkflow regression by optimizing frontmatter delimiter parsing (#31583)

* Initial plan

* Plan ParseWorkflow regression investigation

Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>

* Optimize frontmatter delimiter checks in ParseWorkflow path

Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>

* Tighten delimiter whitespace fast path in frontmatter parser

Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>

* chore: merge main and recompile workflows

Agent-Logs-Url: https://github.com/github/gh-aw/sessions/355f4283-dea2-411d-a18d-579513ada771

Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>

* chore: drop unrelated formatter-only file churn

Agent-Logs-Url: https://github.com/github/gh-aw/sessions/355f4283-dea2-411d-a18d-579513ada771

Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Co-authored-by: Peli de Halleux <pelikhan@users.noreply.github.com>

Author: Copilot

pkg/parser/frontmatter_content.go

@@ -32,7 +32,7 @@ func ExtractFrontmatterFromContent(content string) (*FrontmatterResult, error) {
 	}
 
 	// Check if file starts with frontmatter delimiter.
-	if strings.TrimSpace(firstLine) != "---" {
+	if !isFrontmatterDelimiterLine(firstLine) {
 		log.Print("No frontmatter delimiter found, returning content as markdown")
 		// No frontmatter, return entire content as markdown
 		return &FrontmatterResult{
@@ -62,7 +62,7 @@ func ExtractFrontmatterFromContent(content string) (*FrontmatterResult, error) {
 			}
 		}
 
-		if strings.TrimSpace(content[lineStart:lineEnd]) == "---" {
+		if isFrontmatterDelimiterLine(content[lineStart:lineEnd]) {
 			frontmatterEndStart = lineStart
 			markdownStart = nextCursor
 			break
@@ -122,6 +122,44 @@ func ExtractFrontmatterFromContent(content string) (*FrontmatterResult, error) {
 	}, nil
 }
 
+// isFrontmatterDelimiterLine returns true when a line consists of "---" with optional surrounding whitespace.
+func isFrontmatterDelimiterLine(line string) bool {
+	// Fast path for common delimiters.
+	if line == "---" || line == "---\r" {
+		return true
+	}
+
+	// Fast path for ASCII-trimmable whitespace.
+	start, end := 0, len(line)
+	for start < end {
+		switch line[start] {
+		case ' ', '\t', '\n', '\r', '\v', '\f':
+			start++
+		default:
+			goto leftTrimmed
+		}
+	}
+leftTrimmed:
+	if start >= end {
+		return false
+	}
+	for end > start {
+		switch line[end-1] {
+		case ' ', '\t', '\n', '\r', '\v', '\f':
+			end--
+		default:
+			goto rightTrimmed
+		}
+	}
+rightTrimmed:
+	if end-start == 3 && line[start] == '-' && line[start+1] == '-' && line[start+2] == '-' {
+		return true
+	}
+
+	// Fallback keeps previous behavior for uncommon Unicode whitespace.
+	return strings.TrimSpace(line) == "---"
+}
+
 // ExtractFrontmatterFromBuiltinFile is a caching wrapper around ExtractFrontmatterFromContent
 // for builtin virtual files. Because builtin files are registered once at startup and never
 // change, the parsed FrontmatterResult is identical across calls. This function caches the

pkg/parser/frontmatter_extraction_test.go

@@ -64,6 +64,14 @@ This is a test workflow with empty frontmatter.`,
 			},
 			wantMarkdown: "# Content",
 		},
+		{
+			name: "frontmatter delimiters with surrounding whitespace",
+			content: " \t--- \t\r\non: push\r\n \t--- \t\r\n\r\n# Test Workflow\r\n",
+			wantYAML: map[string]any{
+				"on": "push",
+			},
+			wantMarkdown: "# Test Workflow",
+		},
 	}
 
 	for _, tt := range tests {

pkg/workflow/compiler_orchestrator_frontmatter.go

@@ -40,20 +40,21 @@ func (c *Compiler) parseFrontmatterSection(markdownPath string) (*frontmatterPar
 		// Intentionally not wrapping to avoid exposing internal path details
 		return nil, fmt.Errorf("failed to read file: %v", err) //nolint:errorlint // intentionally not wrapping to avoid exposing os.PathError
 	}
+	contentString := string(content)
 
 	log.Printf("File size: %d bytes", len(content))
 
 	// Parse frontmatter and markdown
 	orchestratorFrontmatterLog.Printf("Parsing frontmatter from file: %s", cleanPath)
-	result, err := parser.ExtractFrontmatterFromContent(string(content))
+	result, err := parser.ExtractFrontmatterFromContent(contentString)
 	if err != nil {
 		orchestratorFrontmatterLog.Printf("Frontmatter extraction failed: %v", err)
 		// Use FrontmatterStart from result if available, otherwise default to line 2 (after opening ---)
 		frontmatterStart := 2
 		if result != nil && result.FrontmatterStart > 0 {
 			frontmatterStart = result.FrontmatterStart
 		}
-		return nil, c.createFrontmatterError(cleanPath, string(content), err, frontmatterStart)
+		return nil, c.createFrontmatterError(cleanPath, contentString, err, frontmatterStart)
 	}
 
 	if len(result.Frontmatter) == 0 {
@@ -62,7 +63,7 @@ func (c *Compiler) parseFrontmatterSection(markdownPath string) (*frontmatterPar
 	}
 
 	// Preprocess schedule fields to convert human-friendly format to cron expressions
-	if err := c.preprocessScheduleFields(result.Frontmatter, cleanPath, string(content)); err != nil {
+	if err := c.preprocessScheduleFields(result.Frontmatter, cleanPath, contentString); err != nil {
 		orchestratorFrontmatterLog.Printf("Schedule preprocessing failed: %v", err)
 		return nil, err
 	}