diff --git a/crates/huginn-daemon/src/sources/git.rs b/crates/huginn-daemon/src/sources/git.rs index af41e21..c471cf0 100644 --- a/crates/huginn-daemon/src/sources/git.rs +++ b/crates/huginn-daemon/src/sources/git.rs @@ -24,13 +24,27 @@ const SKIPPED_NAMES: &[&str] = &["package-lock.json", "pnpm-lock.yaml", "yarn.lo const LFS_POINTER: &[u8] = b"version https://git-lfs.github.com/spec/v1"; +/// Directories of editor settings and generated packets, not the repo's own +/// writing: a path with one of these as a component is not indexed. +const SKIPPED_DIRECTORIES: &[&str] = &[".obsidian", ".voidbot"]; + +/// Licence texts: a file named exactly one of these stems, bare or with a prose +/// extension. They crowd out a front page; `license.rs` and `Copying.cs` are code. +const SKIPPED_STEMS: &[&str] = &["license", "licence", "copying"]; +const LICENSE_EXTENSIONS: &[&str] = &["", "md", "txt", "rst"]; + /// Whether a path is not worth reading, by name alone. pub fn skipped_path(path: &str) -> bool { - let name = path.rsplit('/').next().unwrap_or(path).to_ascii_lowercase(); + let mut components = path.split('/'); + let name = components.next_back().unwrap_or(path).to_ascii_lowercase(); + if components.any(|directory| SKIPPED_DIRECTORIES.contains(&directory.to_ascii_lowercase().as_str())) { + return true; + } if SKIPPED_NAMES.contains(&name.as_str()) { return true; } - name.rsplit_once('.').is_some_and(|(_, extension)| SKIPPED_EXTENSIONS.contains(&extension)) + let (stem, extension) = name.rsplit_once('.').unwrap_or((&name, "")); + (SKIPPED_STEMS.contains(&stem) && LICENSE_EXTENSIONS.contains(&extension)) || extension == "provenance" || SKIPPED_EXTENSIONS.contains(&extension) } /// The text of a blob worth indexing, or `None`: a NUL byte, bytes that are @@ -186,10 +200,14 @@ mod tests { #[test] fn binary_unity_and_lock_paths_are_skipped_by_name_and_source_is_not() { - for path in ["a.png", "dir/B.PNG", "x.lock", "Assets/a.prefab", "Assets/a.unity", "a.cs.meta", "p/package-lock.json", "Cargo.lock"] { + for path in ["a.png", "dir/B.PNG", "x.lock", "Assets/a.prefab", "Assets/a.unity", "a.cs.meta", "p/package-lock.json", "Cargo.lock", + ".obsidian/app.json", "vault/.obsidian/appearance.json", ".voidbot/packets/a.json", "LICENSE", "LICENSE.md", "sub/licence.txt", "COPYING", "COPYING.rst", "x/.Obsidian/a.md", "x/.VoidBot/p.json", + "bundle.provenance", "x/bundle.provenance", ".provenance"] { assert!(skipped_path(path), "{path}"); } - for path in ["src/main.rs", "README.md", "docs/lock.md", "Assets/Scripts/Player.cs", "Makefile", ".gitignore", "locksmith"] { + for path in ["src/main.rs", "README.md", "docs/lock.md", "Assets/Scripts/Player.cs", "Makefile", ".gitignore", "locksmith", + "docs/x.md", "src/obsidian_notes.rs", "docs/licensing.md", "docs/obsidian/notes.md", "provenance.md", + "src/license.rs", "crates/x/src/license.rs", "Copying.cs", "LICENSE-MIT", "LICENSE.md.bak"] { assert!(!skipped_path(path), "{path}"); } }