fix(wheels): drop unused python deps, GraphML \r, walkdir+wait-timeout

Author: nikolay-e

diffctx/Cargo.lock

@@ -117,12 +117,13 @@ rustc-hash = "2"
 similar = "2"
 glob = "0.3"
 smallvec = "1"
+walkdir = "2"
+wait-timeout = "0.2"
 pyo3 = { version = "0.24", features = ["extension-module", "abi3-py310"], optional = true }
 
 [dev-dependencies]
 tempfile = "3"
 serde_yaml = "0.9"
-walkdir = "2"
 libtest-mimic = "0.7"
 
 [[test]]

diffctx/Cargo.toml

@@ -2,6 +2,7 @@ use std::path::{Path, PathBuf};
 
 use rayon::prelude::*;
 use rustc_hash::FxHashSet;
+use walkdir::WalkDir;
 
 use crate::config::graph_filtering::GRAPH_FILTERING;
 use crate::config::limits::LIMITS;
@@ -45,45 +46,34 @@ pub fn collect_candidate_files(root_dir: &Path, included_set: &FxHashSet<PathBuf
     }
 
     let mut fallback: Vec<PathBuf> = Vec::new();
-    if let Ok(entries) = walkdir(root_dir) {
-        for f in entries {
-            if fallback.len() >= GRAPH_FILTERING.fallback_max_files {
-                break;
+    for entry in WalkDir::new(root_dir)
+        .sort_by_file_name()
+        .into_iter()
+        .filter_entry(|e| {
+            if e.depth() == 0 || !e.file_type().is_dir() {
+                return true;
             }
-            if is_candidate_file(&f, root_dir, included_set) {
-                fallback.push(f);
+            match e.file_name().to_str() {
+                Some(name) => {
+                    !name.starts_with('.') && name != "node_modules" && name != "__pycache__"
+                }
+                None => true,
             }
+        })
+        .filter_map(|e| e.ok())
+    {
+        if !entry.file_type().is_file() {
+            continue;
         }
-    }
-    fallback
-}
-
-fn walkdir(root: &Path) -> std::io::Result<Vec<PathBuf>> {
-    let mut result = Vec::new();
-    walk_recursive(root, &mut result)?;
-    Ok(result)
-}
-
-fn walk_recursive(dir: &Path, result: &mut Vec<PathBuf>) -> std::io::Result<()> {
-    let mut entries: Vec<PathBuf> = std::fs::read_dir(dir)?
-        .filter_map(|e| e.ok().map(|e| e.path()))
-        .collect();
-    entries.sort();
-    for path in entries {
-        if path.is_dir() {
-            let name = path
-                .file_name()
-                .map(|n| n.to_string_lossy().to_string())
-                .unwrap_or_default();
-            if name.starts_with('.') || name == "node_modules" || name == "__pycache__" {
-                continue;
-            }
-            walk_recursive(&path, result)?;
-        } else {
-            result.push(path);
+        if fallback.len() >= GRAPH_FILTERING.fallback_max_files {
+            break;
+        }
+        let path = entry.into_path();
+        if is_candidate_file(&path, root_dir, included_set) {
+            fallback.push(path);
         }
     }
-    Ok(())
+    fallback
 }
 
 pub fn normalize_path(path: &Path, root_dir: &Path) -> PathBuf {

diffctx/src/candidate_files.rs

@@ -8,6 +8,7 @@ use std::time::Duration;
 use once_cell::sync::Lazy;
 use regex::Regex;
 use rustc_hash::FxHashSet;
+use wait_timeout::ChildExt;
 
 use crate::config::git::{self, GIT};
 use crate::types::DiffHunk;
@@ -105,19 +106,12 @@ fn wait_with_timeout(
         })
     });
 
-    let start = std::time::Instant::now();
-    let status = loop {
-        match child.try_wait() {
-            Ok(Some(status)) => break status,
-            Ok(None) => {
-                if start.elapsed() >= timeout {
-                    let _ = child.kill();
-                    let _ = child.wait();
-                    return Err(GitError::Timeout(timeout.as_secs()));
-                }
-                std::thread::sleep(Duration::from_millis(GIT.poll_interval_ms));
-            }
-            Err(e) => return Err(GitError::Io(e)),
+    let status = match child.wait_timeout(timeout)? {
+        Some(status) => status,
+        None => {
+            let _ = child.kill();
+            let _ = child.wait();
+            return Err(GitError::Timeout(timeout.as_secs()));
         }
     };
 
@@ -576,8 +570,8 @@ impl CatFileBatch {
         if let Some(mut child) = self.child.take() {
             drop(child.stdin.take());
             match child.wait_timeout(Duration::from_secs(GIT.catfile_termination_timeout_seconds)) {
-                Ok(_) => {}
-                Err(_) => {
+                Ok(Some(_)) => {}
+                _ => {
                     let _ = child.kill();
                     let _ = child.wait();
                 }
@@ -591,25 +585,3 @@ impl Drop for CatFileBatch {
         self.close();
     }
 }
-
-trait WaitTimeout {
-    fn wait_timeout(&mut self, dur: Duration) -> std::result::Result<std::process::ExitStatus, ()>;
-}
-
-impl WaitTimeout for Child {
-    fn wait_timeout(&mut self, dur: Duration) -> std::result::Result<std::process::ExitStatus, ()> {
-        let start = std::time::Instant::now();
-        loop {
-            match self.try_wait() {
-                Ok(Some(status)) => return Ok(status),
-                Ok(None) => {
-                    if start.elapsed() >= dur {
-                        return Err(());
-                    }
-                    std::thread::sleep(Duration::from_millis(GIT.poll_interval_ms));
-                }
-                Err(_) => return Err(()),
-            }
-        }
-    }
-}

diffctx/src/git.rs

@@ -234,6 +234,7 @@ fn escape_graphml(text: &str) -> String {
             '>' => out.push_str(">"),
             '"' => out.push_str("""),
             '\'' => out.push_str("'"),
+            '\r' => out.push_str("
"),
             other => out.push(other),
         }
     }

diffctx/src/graph_export.rs

@@ -54,10 +54,8 @@ classifiers = [
 ]
 dynamic = [ "version" ]  # Version is still managed in version.py
 dependencies = [
-  "networkx>=3.0,<4.0",
   "numpy>=1.24,<3.0",
   "pathspec>=0.11,<2.0",
-  "tiktoken==0.12.0",
 ]
 optional-dependencies.dev = [
   "black>=23.0.0,<27.0",
@@ -84,9 +82,9 @@ optional-dependencies.dev = [
   "ruamel-yaml>=0.18,<1.0",
   # Quality checks
   "ruff>=0.4,<1.0",
+  "tiktoken==0.12.0",
   "treemapper[tree-sitter]",
   # Type stubs
-  "types-networkx>=3.0,<4.0",
   "types-pyyaml>=6.0,<7.0",
 ]
 optional-dependencies.embeddings = [

pyproject.toml

@@ -373,24 +373,15 @@ def _write_to_file_path(output_file: Path, writer: Callable[[TextIO], None]) ->
             os.fsync(f.fileno())
         os.replace(tmp_path, output_file)
     except PermissionError:
-        try:
-            os.unlink(tmp_path)
-        except OSError:
-            pass
+        Path(tmp_path).unlink(missing_ok=True)
         logger.error("Unable to write to file '%s': permission denied", output_file)
         raise
     except OSError as e:
-        try:
-            os.unlink(tmp_path)
-        except OSError:
-            pass
+        Path(tmp_path).unlink(missing_ok=True)
         logger.error("Unable to write to file '%s': %s", output_file, e)
         raise
     except BaseException:
-        try:
-            os.unlink(tmp_path)
-        except OSError:
-            pass
+        Path(tmp_path).unlink(missing_ok=True)
         raise