dirsql 0.4.42__tar.gz → 0.4.44__tar.gz

{dirsql-0.4.42 → dirsql-0.4.44}/Cargo.lock

@@ -350,6 +350,18 @@ version = "1.0.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "1d07550c9036bf2ae0c684c4297d503f838287c83c53686d05370d0e139ae570"
 
+[[package]]
+name = "console"
+version = "0.16.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e96a4956774c13c126a8b5af4daa79384f4d826534c95a02d76afb39e2ab64e3"
+dependencies = [
+ "encode_unicode",
+ "libc",
+ "unicode-width",
+ "windows-sys 0.61.2",
+]
+
 [[package]]
 name = "constant_time_eq"
 version = "0.4.2"
@@ -540,12 +552,13 @@ checksum = "6184e33543162437515c2e2b48714794e37845ec9851711914eec9d308f6ebe8"
 
 [[package]]
 name = "dirsql"
-version = "0.4.42"
+version = "0.4.44"
 dependencies = [
  "assert_cmd",
  "axum",
  "blake3",
  "clap",
+ "console",
  "criterion",
  "eventsource-client",
  "futures",
@@ -554,6 +567,7 @@ dependencies = [
  "futures-util",
  "globset",
  "ignore",
+ "indicatif",
  "libc",
  "notify",
  "predicates",
@@ -583,7 +597,7 @@ dependencies = [
 
 [[package]]
 name = "dirsql-py-ext"
-version = "0.4.42"
+version = "0.4.44"
 dependencies = [
  "dirsql",
  "pyo3",
@@ -630,6 +644,12 @@ version = "1.15.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "48c757948c5ede0e46177b7add2e67155f70e33c07fea8284df6576da70b3719"
 
+[[package]]
+name = "encode_unicode"
+version = "1.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "34aa73646ffb006b8f5147f3dc182bd4bcb190227ce861fc4a4844bf8e3cb2c0"
+
 [[package]]
 name = "equivalent"
 version = "1.0.2"
@@ -1294,6 +1314,19 @@ dependencies = [
  "serde_core",
 ]
 
+[[package]]
+name = "indicatif"
+version = "0.18.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9433806cd6b4ec1aba79c021c7e4c58fb4c3b9977c085062e611ac929998fb0c"
+dependencies = [
+ "console",
+ "portable-atomic",
+ "unicode-width",
+ "unit-prefix",
+ "web-time",
+]
+
 [[package]]
 name = "inotify"
 version = "0.10.2"
@@ -2902,6 +2935,12 @@ version = "0.2.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "ebc1c04c71510c7f702b52b7c350734c9ff1295c464a03335b00bb84fc54f853"
 
+[[package]]
+name = "unit-prefix"
+version = "0.5.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "81e544489bf3d8ef66c953931f56617f423cd4b5494be343d9b9d3dda037b9a3"
+
 [[package]]
 name = "untrusted"
 version = "0.9.0"

{dirsql-0.4.42 → dirsql-0.4.44}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dirsql
-Version: 0.4.42
+Version: 0.4.44
 Requires-Dist: tomli>=2 ; python_full_version < '3.11'
 Requires-Dist: bin-shim>=0.1
 Summary: Ephemeral SQL index over a local directory

{dirsql-0.4.42 → dirsql-0.4.44}/docs/reference/cli.md

@@ -536,6 +536,11 @@ Override with `DIRSQL_PROGRESS`:
 Values are case-insensitive and surrounding whitespace is ignored; anything
 unrecognized reads as `auto`, so a typo cannot stop a scan from running.
 
+Each redraw clears the line and pads the new one out to the full width (the
+terminal's, or 80 columns when there is none), so a capture of an `always` run
+carries trailing spaces. It carries no escape sequences, and the summary that
+survives the phase is a whole, newline-terminated line.
+
 The setting is read by the **core**, not the CLI, so it governs an index built
 from any SDK as well — a Python or TypeScript program that builds a `DirSQL`
 with a terminal attached gets the same two phases on stderr, and the same

{dirsql-0.4.42 → dirsql-0.4.44}/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 = "0.4.42"
+version = "0.4.44"
 # 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-0.4.42/packages/rust → dirsql-0.4.44/packages/python}/docs/reference/cli.md

@@ -536,6 +536,11 @@ Override with `DIRSQL_PROGRESS`:
 Values are case-insensitive and surrounding whitespace is ignored; anything
 unrecognized reads as `auto`, so a typo cannot stop a scan from running.
 
+Each redraw clears the line and pads the new one out to the full width (the
+terminal's, or 80 columns when there is none), so a capture of an `always` run
+carries trailing spaces. It carries no escape sequences, and the summary that
+survives the phase is a whole, newline-terminated line.
+
 The setting is read by the **core**, not the CLI, so it governs an index built
 from any SDK as well — a Python or TypeScript program that builds a `DirSQL`
 with a terminal attached gets the same two phases on stderr, and the same

dirsql-0.4.44/packages/python/e2e-attestations/claude-1081-indicatif-progress.json

@@ -0,0 +1,7 @@
+{
+  "command": "just test-e2e",
+  "ran_at": 1789666612,
+  "exit_code": 0,
+  "commit": "991bcb921ba010ac18968740a79fd1d6fce05e11",
+  "branch": "claude/1081-indicatif-progress"
+}

dirsql-0.4.44/packages/python/e2e-attestations/claude-1119-ext-plan.json

@@ -0,0 +1,7 @@
+{
+  "command": "just test-e2e",
+  "ran_at": 1789672718,
+  "exit_code": 0,
+  "commit": "9f049fb0a4012807ce2b31c220df75b0b33e88e1",
+  "branch": "claude/1119-ext-plan"
+}

{dirsql-0.4.42 → dirsql-0.4.44}/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 = "0.4.42"
+version = "0.4.44"
 edition.workspace = true
 # Literal `license` (not `license.workspace = true`) because
 # putitoutthere's preflight check reads `[package].license` directly
@@ -59,12 +59,14 @@ cli = [
 axum = { version = "0.8", optional = true }
 blake3 = "1"
 clap = { version = "4", features = ["derive"], optional = true }
+console = "0.16.6"
 futures = { version = "0.3", optional = true }
 futures-channel.workspace = true
 globset = "0.4"
 # Hierarchical `.gitignore` matching for path-table scans (the fd/ripgrep
 # semantics), without adopting its walker: the scan keeps `walkdir`.
 ignore = "0.4"
+indicatif = "0.18.6"
 notify = "7"
 # The REPL's interactive line editor (history, Ctrl+R, multi-line entry).
 # CLI-only: a library consumer never pulls crossterm.

{dirsql-0.4.42/packages/python → dirsql-0.4.44/packages/rust}/docs/reference/cli.md

@@ -536,6 +536,11 @@ Override with `DIRSQL_PROGRESS`:
 Values are case-insensitive and surrounding whitespace is ignored; anything
 unrecognized reads as `auto`, so a typo cannot stop a scan from running.
 
+Each redraw clears the line and pads the new one out to the full width (the
+terminal's, or 80 columns when there is none), so a capture of an `always` run
+carries trailing spaces. It carries no escape sequences, and the summary that
+survives the phase is a whole, newline-terminated line.
+
 The setting is read by the **core**, not the CLI, so it governs an index built
 from any SDK as well — a Python or TypeScript program that builds a `DirSQL`
 with a terminal attached gets the same two phases on stderr, and the same

dirsql-0.4.44/packages/rust/src/extension_resolution.rs

@@ -0,0 +1,421 @@
+//! Planning the resolution of `[[dirsql.extension]]` entries.
+//!
+//! Each launcher used to carry its own port of this: read the configs, decide
+//! whether any entry names a package rather than a file, resolve the rest
+//! against the config's directory. Only "where is package `X` installed on
+//! this host" is genuinely host-specific (`importlib.util.find_spec` /
+//! `require.resolve`), so that stays with the launcher and comes back here as
+//! a set of directories.
+//!
+//! The suffix list is a parameter because it differs per host: Python counts
+//! `.pyd` as a loadable, Node counts `.node`.
+
+use std::fmt;
+use std::path::{Path, PathBuf};
+
+use crate::config::{ExtensionSpec, load_config_str};
+
+/// One config file's contents, or `None` when it is missing or unreadable.
+///
+/// `path` must be absolute: entry paths and the bare-name shadow probe resolve
+/// against its parent directory.
+#[derive(Debug, Clone, Copy)]
+pub struct ConfigSource<'a> {
+    pub path: &'a Path,
+    pub contents: Option<&'a str>,
+}
+
+/// One planned extension entry.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub enum PlanEntry {
+    /// A concrete path, already absolute against its config's directory.
+    Literal {
+        path: PathBuf,
+        entrypoint: Option<String>,
+    },
+    /// A bare package name. Use `shadow` when it is a file on disk; otherwise
+    /// locate the package on the host and hand its directories, plus every
+    /// file found under them, to [`select_loadable`].
+    Package {
+        name: String,
+        shadow: PathBuf,
+        entrypoint: Option<String>,
+    },
+}
+
+/// Why a package's loadable file could not be picked.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub enum SelectError {
+    NotFound {
+        name: String,
+        suffixes: Vec<String>,
+        dirs: Vec<PathBuf>,
+    },
+    Ambiguous {
+        name: String,
+        matches: Vec<PathBuf>,
+    },
+}
+
+fn join_display(paths: &[PathBuf], sep: &str) -> String {
+    paths
+        .iter()
+        .map(|p| p.display().to_string())
+        .collect::<Vec<_>>()
+        .join(sep)
+}
+
+impl fmt::Display for SelectError {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            SelectError::NotFound {
+                name,
+                suffixes,
+                dirs,
+            } => {
+                let patterns = suffixes
+                    .iter()
+                    .map(|s| format!("*{s}"))
+                    .collect::<Vec<_>>()
+                    .join(" / ");
+                write!(
+                    f,
+                    "no loadable extension file ({patterns}) found in package '{name}' (searched {})",
+                    join_display(dirs, ", ")
+                )
+            }
+            SelectError::Ambiguous { name, matches } => write!(
+                f,
+                "multiple loadable extension files found in package '{name}': {}; disambiguate with a literal path",
+                join_display(matches, ", ")
+            ),
+        }
+    }
+}
+
+impl std::error::Error for SelectError {}
+
+/// True when `path` names a package rather than a file.
+///
+/// Both separators count regardless of host, so a Windows-style path is a path
+/// everywhere.
+pub fn is_bare_name(path: &str, suffixes: &[&str]) -> bool {
+    if path.contains('/') || path.contains('\\') {
+        return false;
+    }
+    !suffixes.iter().any(|s| path.ends_with(s))
+}
+
+/// Plan the extension entries of several configs, in order.
+///
+/// `None` means "do not intervene": no config names an extension by bare
+/// package name, so the core's own config loading (and its error reporting)
+/// handles every entry. Otherwise **every** config's entries are planned, each
+/// against its own directory, concatenated in `configs` order — once the
+/// caller suppresses the core's config-extension loading it must supply the
+/// literal entries too.
+///
+/// A config whose contents are absent or do not parse is skipped rather than
+/// reported, leaving the error to the core.
+pub fn plan_config_extensions(
+    configs: &[ConfigSource<'_>],
+    suffixes: &[&str],
+) -> Option<Vec<PlanEntry>> {
+    let loaded: Vec<(PathBuf, Vec<ExtensionSpec>)> = configs
+        .iter()
+        .filter_map(|source| {
+            let config = load_config_str(source.contents?).ok()?;
+            Some((source.path.parent()?.to_path_buf(), config.extensions))
+        })
+        .collect();
+
+    if !loaded
+        .iter()
+        .flat_map(|(_, extensions)| extensions)
+        .any(|extension| spec_is_bare_name(extension, suffixes))
+    {
+        return None;
+    }
+
+    Some(
+        loaded
+            .into_iter()
+            .flat_map(|(base, extensions)| {
+                extensions.into_iter().map(move |extension| {
+                    let name = extension.path.to_string_lossy().into_owned();
+                    if is_bare_name(&name, suffixes) {
+                        PlanEntry::Package {
+                            shadow: base.join(&name),
+                            name,
+                            entrypoint: extension.entrypoint,
+                        }
+                    } else {
+                        // `join` already yields an absolute path verbatim.
+                        PlanEntry::Literal {
+                            path: base.join(&extension.path),
+                            entrypoint: extension.entrypoint,
+                        }
+                    }
+                })
+            })
+            .collect(),
+    )
+}
+
+fn spec_is_bare_name(extension: &ExtensionSpec, suffixes: &[&str]) -> bool {
+    extension
+        .path
+        .to_str()
+        .is_some_and(|path| is_bare_name(path, suffixes))
+}
+
+/// Pick the single loadable file for package `name` out of a host's listing.
+///
+/// `dirs` are the package's directories (used only for the not-found message);
+/// `candidates` is every file found under them. Zero matches and more than one
+/// are both errors — the user must disambiguate with a literal path.
+pub fn select_loadable(
+    name: &str,
+    dirs: &[PathBuf],
+    candidates: &[PathBuf],
+    suffixes: &[&str],
+) -> Result<PathBuf, SelectError> {
+    let mut matches: Vec<PathBuf> = candidates
+        .iter()
+        .filter(|candidate| {
+            candidate
+                .to_str()
+                .is_some_and(|path| suffixes.iter().any(|suffix| path.ends_with(suffix)))
+        })
+        .cloned()
+        .collect();
+    matches.sort();
+    matches.dedup();
+
+    match matches.len() {
+        0 => Err(SelectError::NotFound {
+            name: name.to_owned(),
+            suffixes: suffixes.iter().map(|s| (*s).to_owned()).collect(),
+            dirs: dirs.to_vec(),
+        }),
+        1 => Ok(matches.remove(0)),
+        _ => Err(SelectError::Ambiguous {
+            name: name.to_owned(),
+            matches,
+        }),
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::{
+        ConfigSource, PlanEntry, SelectError, is_bare_name, plan_config_extensions, select_loadable,
+    };
+    use std::path::{Path, PathBuf};
+
+    const PY: &[&str] = &[".so", ".dylib", ".dll", ".pyd"];
+    const NODE: &[&str] = &[".so", ".dylib", ".dll", ".node"];
+
+    fn source<'a>(path: &'a Path, contents: &'a str) -> ConfigSource<'a> {
+        ConfigSource {
+            path,
+            contents: Some(contents),
+        }
+    }
+
+    #[test]
+    fn a_value_with_no_separator_and_no_loadable_suffix_is_a_bare_name() {
+        assert!(is_bare_name("sqlite_vec", PY));
+        assert!(is_bare_name("vec0", NODE));
+    }
+
+    #[test]
+    fn a_value_carrying_either_separator_is_a_path() {
+        assert!(!is_bare_name("ext/vec0", PY));
+        assert!(!is_bare_name("ext\\vec0", PY));
+        assert!(!is_bare_name("/abs/vec0", PY));
+    }
+
+    #[test]
+    fn a_value_ending_in_a_loadable_suffix_is_a_path() {
+        for suffix in [".so", ".dylib", ".dll"] {
+            assert!(!is_bare_name(&format!("vec0{suffix}"), PY));
+            assert!(!is_bare_name(&format!("vec0{suffix}"), NODE));
+        }
+    }
+
+    #[test]
+    fn the_suffix_list_is_the_hosts_own() {
+        assert!(!is_bare_name("vec0.pyd", PY));
+        assert!(is_bare_name("vec0.pyd", NODE));
+        assert!(!is_bare_name("vec0.node", NODE));
+        assert!(is_bare_name("vec0.node", PY));
+    }
+
+    #[test]
+    fn configs_using_only_literal_paths_are_left_to_the_core() {
+        let path = Path::new("/project/.dirsql.toml");
+        let config = "[[dirsql.extension]]\npath = \"ext/vec0.so\"\n";
+        assert_eq!(plan_config_extensions(&[source(path, config)], PY), None);
+    }
+
+    #[test]
+    fn a_config_with_no_extension_array_is_left_to_the_core() {
+        let path = Path::new("/project/.dirsql.toml");
+        assert_eq!(
+            plan_config_extensions(&[source(path, "[dirsql]\nignore = [\"*.tmp\"]\n")], PY),
+            None
+        );
+    }
+
+    #[test]
+    fn an_absent_or_unparseable_config_is_skipped_rather_than_reported() {
+        let path = Path::new("/project/.dirsql.toml");
+        let missing = ConfigSource {
+            path,
+            contents: None,
+        };
+        assert_eq!(plan_config_extensions(&[missing], PY), None);
+        assert_eq!(
+            plan_config_extensions(&[source(path, "not valid = [")], PY),
+            None
+        );
+    }
+
+    #[test]
+    fn a_bare_name_plans_a_package_lookup_shadowed_by_the_configs_own_directory() {
+        let path = Path::new("/project/conf/.dirsql.toml");
+        let plan = plan_config_extensions(
+            &[source(
+                path,
+                "[[dirsql.extension]]\npath = \"sqlite_vec\"\n",
+            )],
+            PY,
+        )
+        .expect("a bare name makes the SDK intervene");
+
+        assert_eq!(
+            plan,
+            vec![PlanEntry::Package {
+                name: "sqlite_vec".into(),
+                shadow: PathBuf::from("/project/conf/sqlite_vec"),
+                entrypoint: None,
+            }]
+        );
+    }
+
+    #[test]
+    fn intervening_plans_the_literal_entries_too_and_keeps_config_order() {
+        let first = Path::new("/project/one/.dirsql.toml");
+        let second = Path::new("/project/two/.dirsql.toml");
+        let plan = plan_config_extensions(
+            &[
+                source(
+                    first,
+                    "[[dirsql.extension]]\npath = \"ext/vec0.so\"\nentrypoint = \"sqlite3_vec_init\"\n",
+                ),
+                source(second, "[[dirsql.extension]]\npath = \"sqlite_vec\"\n"),
+            ],
+            PY,
+        )
+        .expect("a bare name anywhere makes the SDK intervene for the whole set");
+
+        assert_eq!(
+            plan,
+            vec![
+                PlanEntry::Literal {
+                    path: PathBuf::from("/project/one/ext/vec0.so"),
+                    entrypoint: Some("sqlite3_vec_init".into()),
+                },
+                PlanEntry::Package {
+                    name: "sqlite_vec".into(),
+                    shadow: PathBuf::from("/project/two/sqlite_vec"),
+                    entrypoint: None,
+                },
+            ]
+        );
+    }
+
+    #[test]
+    fn an_absolute_entry_path_is_left_alone() {
+        let path = Path::new("/project/.dirsql.toml");
+        let plan = plan_config_extensions(
+            &[source(
+                path,
+                "[[dirsql.extension]]\npath = \"/opt/vec0.so\"\n\n[[dirsql.extension]]\npath = \"sqlite_vec\"\n",
+            )],
+            PY,
+        )
+        .expect("intervening");
+
+        assert_eq!(
+            plan[0],
+            PlanEntry::Literal {
+                path: PathBuf::from("/opt/vec0.so"),
+                entrypoint: None,
+            }
+        );
+    }
+
+    #[test]
+    fn select_loadable_returns_the_single_matching_file() {
+        let dir = PathBuf::from("/site-packages/sqlite_vec");
+        let found = select_loadable(
+            "sqlite_vec",
+            std::slice::from_ref(&dir),
+            &[
+                dir.join("__init__.py"),
+                dir.join("vec0.so"),
+                dir.join("README.md"),
+            ],
+            PY,
+        );
+        assert_eq!(found, Ok(dir.join("vec0.so")));
+    }
+
+    #[test]
+    fn select_loadable_dedupes_and_sorts_before_counting() {
+        let dir = PathBuf::from("/pkg");
+        let found = select_loadable(
+            "vec",
+            std::slice::from_ref(&dir),
+            &[dir.join("vec0.so"), dir.join("vec0.so")],
+            PY,
+        );
+        assert_eq!(found, Ok(dir.join("vec0.so")));
+    }
+
+    #[test]
+    fn select_loadable_reports_every_searched_directory_when_nothing_matches() {
+        let dirs = vec![PathBuf::from("/a"), PathBuf::from("/b")];
+        let err = select_loadable("vec", &dirs, &[PathBuf::from("/a/vec.py")], PY)
+            .expect_err("no loadable file");
+        assert_eq!(
+            err.to_string(),
+            "no loadable extension file (*.so / *.dylib / *.dll / *.pyd) found in package 'vec' (searched /a, /b)"
+        );
+    }
+
+    #[test]
+    fn select_loadable_rejects_an_ambiguous_package_in_sorted_order() {
+        let dir = PathBuf::from("/pkg");
+        let err = select_loadable(
+            "vec",
+            std::slice::from_ref(&dir),
+            &[dir.join("z.so"), dir.join("a.so")],
+            PY,
+        )
+        .expect_err("two loadable files");
+        assert_eq!(
+            err,
+            SelectError::Ambiguous {
+                name: "vec".into(),
+                matches: vec![dir.join("a.so"), dir.join("z.so")],
+            }
+        );
+        assert_eq!(
+            err.to_string(),
+            "multiple loadable extension files found in package 'vec': /pkg/a.so, /pkg/z.so; disambiguate with a literal path"
+        );
+    }
+}

{dirsql-0.4.42 → dirsql-0.4.44}/packages/rust/src/lib.rs

@@ -16,6 +16,8 @@ pub mod db;
 #[doc(hidden)]
 pub mod differ;
 #[doc(hidden)]
+pub mod extension_resolution;
+#[doc(hidden)]
 pub mod functions;
 #[doc(hidden)]
 pub mod infer;