feature/acp-cancel-request
claude/github-issue-40-0au5ua
claude/linked-issue-acp-tui-u4vvpa
claude/minor-release-cut-vkugow
claude/sigit-acp-local-chat-cx6380
claude/sigit-cloud-agent-expansion-reox0a
claude/tool-permission-system
claude/zen-feynman-0u78dk
development
feature/acp-cancel-request
feature/acp-session-list-cursor
feature/acp-v1-conformance
feature/agent-tools-multiedit-glob-todos-remember
feature/cargo-workspace
feature/hooks
feature/load-local-model-explicitly
feature/sigit-agent-profile
feature/sigit-code-cloud
feature/xcode-mcp-server
fix/acp-replay-loaded-session-history
fix/acp-replay-reasoning-and-attachments
fix/dsml-doubled-bar-markers
main
release/v1.5.10
release/v1.5.6
release/v1.6.1
| 1 | //! Shared parsing for Markdown files with a leading YAML-ish frontmatter |
| 2 | //! block, e.g.: |
| 3 | //! |
| 4 | //! ```markdown |
| 5 | //! --- |
| 6 | //! name: foo |
| 7 | //! description: bar |
| 8 | //! --- |
| 9 | //! |
| 10 | //! Body text. |
| 11 | //! ``` |
| 12 | //! |
| 13 | //! Used by [`crate::skills`] (`SKILL.md`) and [`crate::commands`] |
| 14 | //! (`.sigit/commands/*.md`) — both need the same "top-level `key: value` |
| 15 | //! scalars, then a Markdown body" shape, so the parsing lives here once |
| 16 | //! rather than being copied per format. |
| 17 | |
| 18 | /// Extract the YAML frontmatter block from `contents`: the text between a |
| 19 | /// leading `---` line and the next `---` line. Returns `None` if absent. |
| 20 | pub fn extract_frontmatter(contents: &str) -> Option<&str> { |
| 21 | // Strip an optional UTF-8 BOM and leading blank lines before the opener. |
| 22 | let trimmed = contents.trim_start_matches('\u{feff}'); |
| 23 | let mut rest = trimmed; |
| 24 | loop { |
| 25 | let line_end = rest.find('\n').map(|i| i + 1).unwrap_or(rest.len()); |
| 26 | let (line, after) = rest.split_at(line_end); |
| 27 | if line.trim().is_empty() { |
| 28 | rest = after; |
| 29 | continue; |
| 30 | } |
| 31 | if line.trim() != "---" { |
| 32 | return None; |
| 33 | } |
| 34 | // `after` now begins just past the opening `---` line. |
| 35 | let body = after; |
| 36 | let mut search = body; |
| 37 | let mut offset = 0; |
| 38 | loop { |
| 39 | let end = search.find('\n').map(|i| i + 1).unwrap_or(search.len()); |
| 40 | let (l, a) = search.split_at(end); |
| 41 | if l.trim() == "---" { |
| 42 | return Some(&body[..offset]); |
| 43 | } |
| 44 | if a.is_empty() { |
| 45 | return None; |
| 46 | } |
| 47 | offset += end; |
| 48 | search = a; |
| 49 | } |
| 50 | } |
| 51 | } |
| 52 | |
| 53 | /// Parse top-level `key: value` scalar pairs from frontmatter, skipping |
| 54 | /// nested mappings (indented lines) and comments. Quoted values are |
| 55 | /// unquoted. |
| 56 | pub fn parse_frontmatter_fields(frontmatter: &str) -> Vec<(String, String)> { |
| 57 | let mut fields = Vec::new(); |
| 58 | for line in frontmatter.lines() { |
| 59 | // Indented lines belong to a nested mapping/sequence — skip them. |
| 60 | if line.starts_with(char::is_whitespace) { |
| 61 | continue; |
| 62 | } |
| 63 | let line = line.trim_end(); |
| 64 | if line.is_empty() || line.starts_with('#') { |
| 65 | continue; |
| 66 | } |
| 67 | let Some((key, value)) = line.split_once(':') else { |
| 68 | continue; |
| 69 | }; |
| 70 | let key = key.trim(); |
| 71 | if key.is_empty() { |
| 72 | continue; |
| 73 | } |
| 74 | let value = unquote(value.trim()); |
| 75 | fields.push((key.to_string(), value)); |
| 76 | } |
| 77 | fields |
| 78 | } |
| 79 | |
| 80 | /// Strip a single layer of matching single or double quotes; otherwise |
| 81 | /// return the value unchanged. |
| 82 | pub fn unquote(value: &str) -> String { |
| 83 | if value.len() >= 2 { |
| 84 | let bytes = value.as_bytes(); |
| 85 | let first = bytes[0]; |
| 86 | let last = bytes[value.len() - 1]; |
| 87 | if (first == b'"' && last == b'"') || (first == b'\'' && last == b'\'') { |
| 88 | return value[1..value.len() - 1].to_string(); |
| 89 | } |
| 90 | } |
| 91 | value.to_string() |
| 92 | } |
| 93 | |
| 94 | /// Return the content after the frontmatter block, or the whole input if |
| 95 | /// there is no frontmatter. |
| 96 | pub fn strip_frontmatter(contents: &str) -> &str { |
| 97 | // Mirror `extract_frontmatter`'s leniency exactly: tolerate a UTF-8 BOM, |
| 98 | // leading blank lines before the opener, and CRLF line endings (compare |
| 99 | // trimmed lines). If the two disagree, a file whose frontmatter parses |
| 100 | // still leaks that raw block into the body — see the round-trip test. |
| 101 | let trimmed = contents.trim_start_matches('\u{feff}'); |
| 102 | let mut rest = trimmed; |
| 103 | loop { |
| 104 | let line_end = rest.find('\n').map(|i| i + 1).unwrap_or(rest.len()); |
| 105 | let (line, after) = rest.split_at(line_end); |
| 106 | if line.trim().is_empty() { |
| 107 | rest = after; |
| 108 | continue; |
| 109 | } |
| 110 | if line.trim() != "---" { |
| 111 | return contents; |
| 112 | } |
| 113 | // `after` begins just past the opening `---` line; find the closing one. |
| 114 | let mut search = after; |
| 115 | loop { |
| 116 | let end = search.find('\n').map(|i| i + 1).unwrap_or(search.len()); |
| 117 | let (l, a) = search.split_at(end); |
| 118 | if l.trim() == "---" { |
| 119 | return a; |
| 120 | } |
| 121 | if a.is_empty() { |
| 122 | return contents; |
| 123 | } |
| 124 | search = a; |
| 125 | } |
| 126 | } |
| 127 | } |
| 128 | |
| 129 | #[cfg(test)] |
| 130 | mod tests { |
| 131 | use super::*; |
| 132 | |
| 133 | #[test] |
| 134 | fn extract_frontmatter_reads_block() { |
| 135 | let md = "---\nname: foo\ndescription: bar\n---\n\nBody here.\n"; |
| 136 | let fm = extract_frontmatter(md).expect("frontmatter"); |
| 137 | assert!(fm.contains("name: foo")); |
| 138 | assert!(fm.contains("description: bar")); |
| 139 | assert!(!fm.contains("Body here")); |
| 140 | } |
| 141 | |
| 142 | #[test] |
| 143 | fn extract_frontmatter_none_without_delimiter() { |
| 144 | assert!(extract_frontmatter("no frontmatter here").is_none()); |
| 145 | assert!(extract_frontmatter("---\nname: foo\n").is_none()); |
| 146 | } |
| 147 | |
| 148 | #[test] |
| 149 | fn parse_fields_handles_quotes_and_nesting() { |
| 150 | let fm = "name: pdf-processing\ndescription: \"Extract PDF text\"\nmetadata:\n author: me\n version: \"1.0\"\nlicense: Apache-2.0\n"; |
| 151 | let fields = parse_frontmatter_fields(fm); |
| 152 | let get = |k: &str| { |
| 153 | fields |
| 154 | .iter() |
| 155 | .find(|(key, _)| key == k) |
| 156 | .map(|(_, v)| v.clone()) |
| 157 | }; |
| 158 | assert_eq!(get("name").as_deref(), Some("pdf-processing")); |
| 159 | assert_eq!(get("description").as_deref(), Some("Extract PDF text")); |
| 160 | assert_eq!(get("license").as_deref(), Some("Apache-2.0")); |
| 161 | // Nested keys under `metadata:` are skipped. |
| 162 | assert!(get("author").is_none()); |
| 163 | assert!(get("version").is_none()); |
| 164 | } |
| 165 | |
| 166 | #[test] |
| 167 | fn strip_frontmatter_returns_body() { |
| 168 | let md = "---\nname: foo\ndescription: bar\n---\n\n# Heading\n\nText.\n"; |
| 169 | assert_eq!(strip_frontmatter(md).trim(), "# Heading\n\nText."); |
| 170 | } |
| 171 | |
| 172 | #[test] |
| 173 | fn strip_frontmatter_passes_through_without_block() { |
| 174 | assert_eq!(strip_frontmatter("just body"), "just body"); |
| 175 | } |
| 176 | |
| 177 | #[test] |
| 178 | fn extract_and_strip_agree_on_crlf_and_leading_blanks() { |
| 179 | // Whenever extract_frontmatter finds a block, strip_frontmatter must |
| 180 | // remove exactly that block and never leak it into the body. These |
| 181 | // inputs (CRLF endings, a leading blank line, a BOM) parse fine in |
| 182 | // extract but used to defeat strip's byte-exact opener check. |
| 183 | for md in [ |
| 184 | "---\r\nname: foo\r\ndescription: bar\r\n---\r\n\r\nBody.\r\n", |
| 185 | "\n---\nname: foo\ndescription: bar\n---\n\nBody.\n", |
| 186 | "\u{feff}---\nname: foo\ndescription: bar\n---\n\nBody.\n", |
| 187 | ] { |
| 188 | assert!( |
| 189 | extract_frontmatter(md).is_some(), |
| 190 | "extract should find frontmatter in {md:?}" |
| 191 | ); |
| 192 | let body = strip_frontmatter(md); |
| 193 | assert!( |
| 194 | !body.contains("name: foo"), |
| 195 | "frontmatter leaked into body for {md:?}: {body:?}" |
| 196 | ); |
| 197 | assert_eq!(body.trim(), "Body."); |
| 198 | } |
| 199 | } |
| 200 | } |