Skip to content

Fix vortex-jni local path handling, centralise logic in vortex-file #8650

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
robert3005 wants to merge 1 commit into
base: develop
Choose a base branch
from
Open

Fix vortex-jni local path handling, centralise logic in vortex-file #8650

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions Cargo.lock
View file
Open in desktop

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

39 changes: 39 additions & 0 deletions java/vortex-spark/src/test/java/dev/vortex/spark/VortexDataSourceWriteTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -129,6 +129,45 @@ public void testWriteAndReadVortexFiles() throws IOException {
verifyDataContent(originalDf, readDf);
}

@Test
@DisplayName("Write and read Vortex files with a bare path (no URI scheme)")
public void testWriteAndReadWithBarePath() throws IOException {
int numRows = 50;
Dataset<Row> originalDf = createTestDataFrame(numRows);

// A bare filesystem path, without a file:// scheme.
String barePath = tempDir.resolve("bare_path_output").toString();
originalDf
.write()
.format("vortex")
.option("path", barePath)
.mode(SaveMode.Overwrite)
.save();

assertFalse(findVortexFiles(tempDir.resolve("bare_path_output")).isEmpty(), "Write should create files");

// Reading a bare directory path exercises schema inference via file listing.
Dataset<Row> readDf =
spark.read().format("vortex").option("path", barePath).load();

assertSchemaEquals(originalDf.schema(), readDf.schema());
assertEquals(numRows, readDf.count(), "Read DataFrame should have same number of rows as original");
verifyDataContent(originalDf, readDf);

// Overwriting a bare path exercises file listing and deletion of the existing files.
Dataset<Row> replacementDf = createTestDataFrame(25);
replacementDf
.write()
.format("vortex")
.option("path", barePath)
.mode(SaveMode.Overwrite)
.save();

Dataset<Row> reread =
spark.read().format("vortex").option("path", barePath).load();
assertEquals(25, reread.count(), "Should have data from second write after overwrite");
}

@Test
@DisplayName("Write empty DataFrame as Vortex")
public void testWriteEmptyDataFrame() throws IOException {
Expand Down
180 changes: 2 additions & 178 deletions vortex-duckdb/src/multi_file.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,8 +1,6 @@
// SPDX-License-Identifier: Apache-2.0
// SPDX-FileCopyrightText: Copyright the Vortex contributors

use std::path::Path;
use std::path::absolute;
use std::sync::Arc;

use itertools::Itertools;
Expand All @@ -14,6 +12,7 @@ use vortex::error::VortexResult;
use vortex::error::vortex_bail;
use vortex::error::vortex_err;
use vortex::file::multi::MultiFileDataSource;
use vortex::file::multi::parse_uri_or_path;
use vortex::io::compat::Compat;
use vortex::io::filesystem::FileSystemRef;
use vortex::io::object_store::ObjectStoreFileSystem;
Expand All @@ -26,38 +25,6 @@ use crate::SESSION;
use crate::duckdb::BindInputRef;
use crate::duckdb::ExtractedValue;

/// Parse a glob string into a [`Url`].
///
/// Accepts full URLs (e.g. `s3://bucket/prefix/*.vortex`, `file:///data/*.vortex`) as well as
/// bare file paths. For bare paths, the path is made absolute (without requiring it to exist)
/// so that relative paths such as `./data/*.vortex` or `../data/*.vortex` are resolved correctly.
fn parse_glob_url(glob_url_str: &str) -> VortexResult<Url> {
Url::parse(glob_url_str).or_else(|_| {
let path = absolute(Path::new(glob_url_str))
.map_err(|e| vortex_err!("Failed making {glob_url_str} absolute: {e}"))?;
// `absolute()` does not normalize `..` components, so `/a/b/../c` stays as-is.
// Normalizing manually avoids `..` being percent-encoded in the resulting URL.
let path = normalize_path(path);
Url::from_file_path(path).map_err(|_| vortex_err!("Neither URL nor path: {glob_url_str}"))
})
}

/// Normalize a path by resolving `.` and `..` components without accessing the filesystem.
fn normalize_path(path: std::path::PathBuf) -> std::path::PathBuf {
use std::path::Component;
let mut normalized = std::path::PathBuf::new();
for component in path.components() {
match component {
Component::CurDir => {}
Component::ParentDir => {
normalized.pop();
}
c => normalized.push(c),
}
}
normalized
}

fn resolve_filesystem(base_url: &Url) -> VortexResult<FileSystemRef> {
let object_store: Arc<dyn ObjectStore> = match base_url.scheme() {
"file" => Arc::new(LocalFileSystem::new()),
Expand Down Expand Up @@ -104,7 +71,7 @@ pub fn bind_multi_file_scan(input: &BindInputRef) -> VortexResult<MultiLayoutDat
// Parse each glob URL and resolve its filesystem.
let mut glob_urls: Vec<Url> = Vec::with_capacity(glob_strings.len());
for glob_str in &glob_strings {
glob_urls.push(parse_glob_url(glob_str)?);
glob_urls.push(parse_uri_or_path(glob_str)?);
}

// Cache filesystems by base URL to avoid resolving the same filesystem multiple times.
Expand Down Expand Up @@ -134,146 +101,3 @@ pub fn bind_multi_file_scan(input: &BindInputRef) -> VortexResult<MultiLayoutDat
builder.build().await
})
}

#[cfg(test)]
mod tests {
use rstest::rstest;

use super::*;

#[test]
fn test_parse_glob_url_s3() -> VortexResult<()> {
let url = parse_glob_url("s3://my-bucket/prefix/*.vortex")?;
assert_eq!(url.scheme(), "s3");
assert_eq!(url.host_str(), Some("my-bucket"));
assert_eq!(url.path(), "/prefix/*.vortex");
Ok(())
}

#[test]
fn test_parse_glob_url_file_scheme() -> VortexResult<()> {
let url = parse_glob_url("file:///absolute/path/data.vortex")?;
assert_eq!(url.scheme(), "file");
assert_eq!(url.path(), "/absolute/path/data.vortex");
Ok(())
}

#[test]
fn test_parse_glob_url_absolute_glob_path() -> VortexResult<()> {
let tmpdir = tempfile::tempdir()?;
let glob = format!("{}/*.vortex", tmpdir.path().display());
let url = parse_glob_url(&glob)?;
assert_eq!(url.scheme(), "file");
assert!(url.path().ends_with("/*.vortex"));
Ok(())
}

#[test]
fn test_parse_glob_url_absolute_existing_path() -> VortexResult<()> {
let tmpfile = tempfile::NamedTempFile::new()?;
let canonical = std::fs::canonicalize(tmpfile.path())?;
let path_str = canonical
.to_str()
.ok_or_else(|| vortex_err!("canonical path is not valid UTF-8"))?;
let url = parse_glob_url(path_str)?;
assert_eq!(url.scheme(), "file");
assert_eq!(url.path(), path_str);
Ok(())
}

#[test]
fn test_parse_glob_url_relative_path() -> VortexResult<()> {
// Create a tempfile in the current working directory so we can refer to it
// by a relative name (just the filename, without any directory component).
let tmpfile = tempfile::NamedTempFile::new_in(".")?;
let filename = tmpfile
.path()
.file_name()
.ok_or_else(|| vortex_err!("temp file missing file name"))?
.to_str()
.ok_or_else(|| vortex_err!("temp file name is not valid UTF-8"))?;

let url = parse_glob_url(filename)?;
assert_eq!(url.scheme(), "file");
// The relative name must have been resolved to an absolute path.
assert!(url.path().ends_with(filename));
assert!(url.path().starts_with('/'));
Ok(())
}

#[test]
fn test_parse_glob_url_relative_glob_path() -> VortexResult<()> {
// A relative path with a glob character (e.g. `./data/*.vortex`) must also resolve
// correctly.
let tmpdir = tempfile::tempdir_in(".")?;
let dir_name = tmpdir
.path()
.file_name()
.ok_or_else(|| vortex_err!("temp dir missing file name"))?
.to_str()
.ok_or_else(|| vortex_err!("temp dir name is not valid UTF-8"))?;
let glob = format!("./{dir_name}/*.vortex");
let url = parse_glob_url(&glob)?;
assert_eq!(url.scheme(), "file");
assert!(url.path().starts_with('/'));
assert!(url.path().ends_with("/*.vortex"));
Ok(())
}

#[test]
fn test_parse_glob_url_nonexistent_path() -> VortexResult<()> {
// absolute() does not require the path to exist, so a non-existent path succeeds.
let url = parse_glob_url("/nonexistent/path/file.vortex")?;
assert_eq!(url.scheme(), "file");
assert_eq!(url.path(), "/nonexistent/path/file.vortex");
Ok(())
}

#[test]
fn test_parse_glob_url_parent_relative_path() -> VortexResult<()> {
// A path starting with `..` must be resolved to an absolute path without
// percent-encoding the `..` component in the resulting URL.
let tmpfile = tempfile::NamedTempFile::new_in("..")?;
let filename = tmpfile
.path()
.file_name()
.ok_or_else(|| vortex_err!("temp file missing file name"))?
.to_str()
.ok_or_else(|| vortex_err!("temp file name is not valid UTF-8"))?;
let relative = format!("../{filename}");

let url = parse_glob_url(&relative)?;
assert_eq!(url.scheme(), "file");
// The resolved path must be absolute and must not contain encoded dots.
assert!(url.path().starts_with('/'));
assert!(
!url.path().contains("%2E"),
"path must not contain percent-encoded dots"
);
assert!(url.path().ends_with(filename));
Ok(())
}

// Use absolute paths so the expected result is cwd-independent.
#[rstest]
#[case("/a/./b", "/a/b")]
#[case("/a/b/./c", "/a/b/c")]
#[case("/a/../b", "/b")]
#[case("/a/b/../c", "/a/c")]
#[case("/a/b/../../c", "/c")]
#[case("/a/./b/.././c", "/a/c")]
#[case("/a/b/../..", "/")]
fn test_parse_glob_url_dot_normalization(
#[case] input: &str,
#[case] expected_path: &str,
) -> VortexResult<()> {
let url = parse_glob_url(input)?;
assert_eq!(url.scheme(), "file");
assert_eq!(
url.path(),
expected_path,
"input {input:?} should normalize to {expected_path:?}"
);
Ok(())
}
}
1 change: 1 addition & 0 deletions vortex-file/Cargo.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,7 @@ parking_lot = { workspace = true }
pin-project-lite = { workspace = true }
tokio = { workspace = true, features = ["rt"], optional = true }
tracing = { workspace = true }
url = { workspace = true }
vortex-alp = { workspace = true }
vortex-array = { workspace = true }
vortex-btrblocks = { workspace = true }
Expand Down
2 changes: 2 additions & 0 deletions vortex-file/src/multi/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@
//! Builder for constructing a [`MultiLayoutDataSource`] from multiple Vortex files.

mod session;
mod uri;

use std::sync::Arc;

Expand All @@ -14,6 +15,7 @@ use futures::stream;
pub use session::MultiFileSession;
use session::MultiFileSessionExt;
use tracing::debug;
pub use uri::parse_uri_or_path;
use vortex_error::VortexError;
use vortex_error::VortexResult;
use vortex_error::vortex_bail;
Expand Down
Loading
Loading