dirsql 0.4.87__tar.gz → 0.4.89__tar.gz

{dirsql-0.4.87 → dirsql-0.4.89}/Cargo.lock

@@ -552,7 +552,7 @@ checksum = "6184e33543162437515c2e2b48714794e37845ec9851711914eec9d308f6ebe8"
 
 [[package]]
 name = "dirsql"
-version = "0.4.87"
+version = "0.4.89"
 dependencies = [
  "assert_cmd",
  "axum",
@@ -603,7 +603,7 @@ dependencies = [
 
 [[package]]
 name = "dirsql-py-ext"
-version = "0.4.87"
+version = "0.4.89"
 dependencies = [
  "dirsql",
  "pyo3",

{dirsql-0.4.87 → dirsql-0.4.89}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dirsql
-Version: 0.4.87
+Version: 0.4.89
 Requires-Dist: bin-shim>=0.1
 Summary: Ephemeral SQL index over a local directory
 Keywords: sql,filesystem,directory,sqlite,index

{dirsql-0.4.87 → dirsql-0.4.89}/packages/python/Cargo.toml

@@ -4,7 +4,7 @@ name = "dirsql-py-ext"
 # pypi/maturin handler can rewrite it via `write-version` before
 # `maturin build`. `pyproject.toml` declares `dynamic = ["version"]`
 # and maturin reads this field. Mirrors `packages/rust/Cargo.toml`.
-version = "0.4.87"
+version = "0.4.89"
 # abi3 note: the `abi3-py310` pyo3 feature (below) builds ONE stable-ABI
 # `cp310-abi3` wheel per platform that loads on every CPython >= 3.10
 # (matching `pyproject.toml`'s `requires-python = ">=3.10"`), instead of

dirsql-0.4.89/packages/python/changelog.d/2026-10-06-config-directory.md

@@ -0,0 +1,3 @@
+**Changed** A declared `[[table]] glob` that names a directory, `glob = "docs"`
+or `glob = "docs/"`, lists the files directly inside it, like `FROM './docs'`.
+It used to match nothing.

dirsql-0.4.89/packages/python/changelog.d/2026-10-06-config-gitignore.md

@@ -0,0 +1,3 @@
+**Changed** Declared `[[table]]` tables respect `.gitignore` by default, as
+path-tables do: inside a git repo, a file a `.gitignore` ignores is no longer a
+row, at startup or while watching. `--no-ignore` turns this off for them too.

dirsql-0.4.89/packages/python/e2e-attestations/claude-1258-config-directory.json

@@ -0,0 +1,7 @@
+{
+  "command": "uv run python -m pytest tests/e2e/glob_speed_one_level_test.py tests/e2e/glob_speed_matchers_test.py tests/e2e/glob_speed_recursive_test.py tests/e2e/query_contract_test.py -x -q",
+  "ran_at": 1791331926,
+  "exit_code": 0,
+  "commit": "04430dd3eae48a6936b5ff45c9c51a38900fbacc",
+  "branch": "claude/1258-config-directory"
+}

dirsql-0.4.89/packages/python/e2e-attestations/claude-1258-config-gitignore.json

@@ -0,0 +1,7 @@
+{
+  "command": "uv run python -m pytest tests/e2e/glob_speed_gitignore_test.py tests/e2e/glob_speed_dotfiles_test.py tests/e2e/glob_speed_recursive_test.py tests/e2e/glob_speed_matchers_test.py tests/e2e/query_contract_test.py -x -q",
+  "ran_at": 1791329772,
+  "exit_code": 0,
+  "commit": "f828a8181c981ab9c5c775117cb60edf3ccd9dcc",
+  "branch": "claude/1258-config-gitignore"
+}

dirsql-0.4.89/packages/python/migrations.d/2026-10-06-config-directory.md

@@ -0,0 +1,39 @@
+### A config glob naming a directory lists its files
+
+#### Summary
+
+`glob = "docs"` used to match nothing, because no file is named `docs`. It now
+lists the files directly inside `docs`, as `FROM './docs'` does. A config that
+relied on the empty result gains rows.
+
+#### Required changes
+
+| Before | After |
+| --- | --- |
+| `glob = "docs"` (no rows) | `glob = "docs"` lists `docs/*`; write `docs/**` for every depth |
+
+#### Deprecations removed
+
+_None._
+
+#### Behavior changes without code changes
+
+- A glob with no wildcard that names a directory expands to `<dir>/*`.
+- A glob with no wildcard that names a file still names only that file.
+
+#### Verification
+
+```bash
+cd "$(mktemp -d)"
+mkdir docs
+echo '[{"p":"a"}]' > docs/a.json
+cat > .dirsql.toml <<'TOML'
+[[table]]
+name = "t"
+glob = "docs"
+ddl = "CREATE TABLE t (p TEXT)"
+on-file = "jq -sc add"
+TOML
+dirsql query "SELECT p FROM t" -c .dirsql.toml
+# expected: [{"p":"a"}]
+```

dirsql-0.4.89/packages/python/migrations.d/2026-10-06-config-gitignore.md

@@ -0,0 +1,45 @@
+### Config tables respect .gitignore
+
+#### Summary
+
+A declared `[[table]]` used to read every file its glob matched, whatever
+`.gitignore` said. It now skips gitignored files, as a path-table does. This
+breaks a config that relied on indexing an ignored file; pass `--no-ignore`
+(or `no_ignore` in the SDKs) to restore the old behavior.
+
+#### Required changes
+
+| Before | After |
+| --- | --- |
+| a table indexed gitignored files | pass `--no-ignore`, or stop ignoring the file |
+
+#### Deprecations removed
+
+_None._
+
+#### Behavior changes without code changes
+
+- Inside a git repo (a directory holding `.git`), files a `.gitignore` ignores
+  are not rows of a `[[table]]`, at startup or while watching.
+- Outside a repo nothing changes.
+
+#### Verification
+
+```bash
+cd "$(mktemp -d)"
+git init -q
+printf 'skip.json\n' > .gitignore
+echo '[{"p":"keep"}]' > keep.json
+echo '[{"p":"skip"}]' > skip.json
+cat > .dirsql.toml <<'TOML'
+[[table]]
+name = "t"
+glob = "*.json"
+ddl = "CREATE TABLE t (p TEXT)"
+on-file = "jq -sc add"
+TOML
+dirsql query "SELECT p FROM t" -c .dirsql.toml
+# expected: [{"p":"keep"}]
+dirsql query "SELECT p FROM t ORDER BY p" -c .dirsql.toml --no-ignore
+# expected: [{"p":"keep"},{"p":"skip"}]
+```

{dirsql-0.4.87 → dirsql-0.4.89}/packages/rust/Cargo.toml

@@ -6,7 +6,7 @@ description = "Ephemeral SQL index over a local directory"
 # version stays for the binding crates (publish = false). The literal
 # intentionally lags the published version: putitoutthere rewrites it to
 # the planned release version at build time and never commits it back.
-version = "0.4.87"
+version = "0.4.89"
 edition.workspace = true
 # Literal `license` (not `license.workspace = true`) because
 # putitoutthere's preflight check reads `[package].license` directly

{dirsql-0.4.87 → dirsql-0.4.89}/packages/rust/src/lib.rs

@@ -732,6 +732,14 @@ impl DirSQL {
             if group.matcher.is_ignored(&rel_path_buf) {
                 continue;
             }
+            if group.matcher.respects_gitignore()
+                && scanner::is_gitignored_path(
+                    &abs_path,
+                    self.inner.fs.is_dir(&abs_path).unwrap_or(false),
+                )
+            {
+                continue;
+            }
 
             // A directory that appears whole (`mkdir`, or a populated directory
             // renamed into the tree) arrives as one event naming the directory and
@@ -1110,6 +1118,13 @@ impl DirSQL {
         } = resolved;
 
         let (groups, anchors, table_names) = compile_groups(&tables, &ignore, &root)?;
+        let groups = groups
+            .into_iter()
+            .map(|group| AnchorGroup {
+                matcher: group.matcher.with_gitignore(!no_ignore),
+                ..group
+            })
+            .collect::<Vec<_>>();
 
         // Resolve the persistent context before scanning, so the scan can
         // consult the cached file index.
@@ -1996,7 +2011,10 @@ fn compile_groups(
             }
             None => root.to_path_buf(),
         };
-        let mapping = (table.glob.clone(), table_name.clone());
+        let mapping = (
+            path_table::directory_as_glob(&table.glob, &anchor, &|p| p.is_dir()),
+            table_name.clone(),
+        );
         match grouped
             .iter_mut()
             .find(|(a, i, _)| *a == anchor && *i == table.ignore)
@@ -2935,6 +2953,23 @@ mod internal_tests {
         assert_eq!(found, vec!["a.txt"]);
     }
 
+    /// The startup walk honors `.gitignore` unless the build asked not to.
+    #[test]
+    fn prepare_hands_the_walk_a_matcher_that_honors_gitignore_unless_no_ignore() {
+        let dir = TempDir::new().unwrap();
+        for no_ignore in [false, true] {
+            let honored = Cell::new(None);
+            let mut resolved = resolved_over(dir.path(), vec![txt_table()], false);
+            resolved.no_ignore = no_ignore;
+            DirSQL::prepare_resolved_with(resolved, &|_, matcher, _| {
+                honored.set(Some(matcher.respects_gitignore()));
+                Vec::new()
+            })
+            .unwrap();
+            assert_eq!(honored.get(), Some(!no_ignore));
+        }
+    }
+
     /// Skipping the walk skips only the walk: with persist on, the cache is
     /// still opened and reconciled against the (empty) scan.
     #[test]

{dirsql-0.4.87 → dirsql-0.4.89}/packages/rust/src/matcher.rs

@@ -97,6 +97,7 @@ pub struct TableMatcher {
     entries: Vec<PatternEntry>,
     ignore_set: Vec<Pattern>,
     ignore_dir_set: Vec<Pattern>,
+    gitignore: bool,
 }
 
 /// Byte spans of the `{name}` placeholders in `pattern`, in order: `(start,
@@ -182,9 +183,20 @@ impl TableMatcher {
             entries,
             ignore_set,
             ignore_dir_set,
+            gitignore: false,
         })
     }
 
+    /// Whether a scan under this matcher honors `.gitignore` files.
+    pub fn with_gitignore(mut self, gitignore: bool) -> Self {
+        self.gitignore = gitignore;
+        self
+    }
+
+    pub(crate) fn respects_gitignore(&self) -> bool {
+        self.gitignore
+    }
+
     /// Returns one [`MatchResult`] per matching pattern, in declaration order.
     /// A file matching N patterns yields N results (fan-out). Empty when
     /// nothing matches.
@@ -230,6 +242,13 @@ mod tests {
         assert_eq!(names(&matcher, "report.csv"), vec!["data"]);
     }
 
+    #[test]
+    fn a_matcher_honors_gitignore_only_once_asked_to() {
+        let matcher = TableMatcher::new(&[], &[]).unwrap();
+        assert!(!matcher.respects_gitignore());
+        assert!(matcher.with_gitignore(true).respects_gitignore());
+    }
+
     #[test]
     fn match_all_returns_empty_for_no_match() {
         let matcher = TableMatcher::new(&[("*.csv", "data")], &[]).unwrap();

{dirsql-0.4.87 → dirsql-0.4.89}/packages/rust/src/path_table.rs

@@ -149,6 +149,17 @@ fn resolve_as<S: Syntax>(
     }
 }
 
+/// `glob` as a path-table reads the same string after `FROM './`: a trailing
+/// separator or a wholly literal name that is a directory under `index_root`
+/// lists the files directly inside it.
+pub fn directory_as_glob(glob: &str, index_root: &Path, is_dir: &dyn Fn(&Path) -> bool) -> String {
+    let target = trailing_separator_as_star::<Utf8UnixEncoding>(glob);
+    if !target.is_empty() && !has_glob_metacharacter(&target) && is_dir(&index_root.join(&target)) {
+        return format!("{target}/{DIRECTORY_GLOB}");
+    }
+    target
+}
+
 /// Where a config `[[table]] glob` anchors, and the glob to match beneath it.
 ///
 /// A glob with a path-table prefix (`/`, `../`, `~/`, a drive or UNC root)
@@ -360,6 +371,45 @@ mod tests {
         }
     }
 
+    fn directory_glob(glob: &str, is_dir: &dyn Fn(&Path) -> bool) -> String {
+        directory_as_glob(glob, Path::new(ROOT), is_dir)
+    }
+
+    #[test]
+    fn a_literal_directory_name_lists_one_level() {
+        assert_eq!(directory_glob("docs", &everything_is_a_dir), "docs/*");
+    }
+
+    #[test]
+    fn a_literal_directory_is_looked_up_under_the_index_root() {
+        let only_docs = |p: &Path| p == Path::new("/index/docs");
+        assert_eq!(directory_glob("docs", &only_docs), "docs/*");
+        assert_eq!(directory_glob("other", &only_docs), "other");
+    }
+
+    #[test]
+    fn a_literal_file_name_is_left_alone() {
+        assert_eq!(directory_glob("top.md", &nothing_is_a_dir), "top.md");
+    }
+
+    #[test]
+    fn a_glob_is_left_alone_even_when_a_directory_has_its_name() {
+        assert_eq!(
+            directory_glob("docs/*.md", &everything_is_a_dir),
+            "docs/*.md"
+        );
+    }
+
+    #[test]
+    fn a_trailing_separator_lists_one_level() {
+        assert_eq!(directory_glob("docs/", &nothing_is_a_dir), "docs/*");
+    }
+
+    #[test]
+    fn an_empty_glob_is_left_alone() {
+        assert_eq!(directory_glob("", &everything_is_a_dir), "");
+    }
+
     #[test]
     fn default_ignores_cover_vcs_and_dependency_directories() {
         assert_eq!(DEFAULT_IGNORES, ["**/node_modules/**", "**/.git/**"]);

{dirsql-0.4.87 → dirsql-0.4.89}/packages/rust/src/scanner.rs

@@ -93,15 +93,23 @@ fn scan_below(
 ) -> Vec<(PathBuf, String)> {
     let mut results = Vec::new();
     let mut seen: u64 = 0;
+    let gitignore = matcher.respects_gitignore();
+    let repo = if gitignore {
+        enclosing_repo(start, &holds_git)
+    } else {
+        None
+    };
 
     // Match against relative path so globs like "comments/**/*.jsonl" work
     // regardless of the absolute root directory.
     let walker = Walk {
         ignore: matcher,
         glob: None,
-        gitignore: false,
-        in_repo: false,
-        frames: Vec::new(),
+        gitignore,
+        in_repo: repo.is_some(),
+        frames: repo.map_or_else(Vec::new, |top| {
+            gitignores_above(start, top, &load_gitignore)
+        }),
         dirs: false,
     };
     walk(
@@ -741,6 +749,26 @@ fn load_gitignore(dir: &Path) -> Option<Gitignore> {
     builder.build().ok()
 }
 
+/// Whether the `.gitignore` files of the repo enclosing `path` ignore it, or
+/// a directory above it, as the walk would have: for a path a watcher reports
+/// without having walked to it.
+pub(crate) fn is_gitignored_path(path: &Path, is_dir: bool) -> bool {
+    let Some(top) = path
+        .parent()
+        .and_then(|dir| enclosing_repo(dir, &holds_git))
+    else {
+        return false;
+    };
+    for frame in gitignores_above(path, top, &load_gitignore).iter().rev() {
+        match frame.matched_path_or_any_parents(path, is_dir) {
+            Match::Ignore(_) => return true,
+            Match::Whitelist(_) => return false,
+            Match::None => {}
+        }
+    }
+    false
+}
+
 /// Whether the `.gitignore` files in force mark `path` ignored. Deeper files
 /// take precedence (git's rule), and a whitelisting `!pattern` un-ignores.
 fn is_gitignored(frames: &[Arc<Gitignore>], path: &Path, is_dir: bool) -> bool {
@@ -1664,4 +1692,14 @@ mod tests {
         let frames = [frame("", &["!keep.log"]), frame("sub", &["*.log"])];
         assert!(is_gitignored(&frames, Path::new("sub/keep.log"), false));
     }
+
+    #[test]
+    fn is_gitignored_path_ignores_a_path_the_repo_gitignore_matches() {
+        let repo = tempfile::tempdir().unwrap();
+        fs::create_dir(repo.path().join(".git")).unwrap();
+        fs::write(repo.path().join(".gitignore"), "*.log\n").unwrap();
+
+        assert!(is_gitignored_path(&repo.path().join("debug.log"), false));
+        assert!(!is_gitignored_path(&repo.path().join("app.js"), false));
+    }
 }