julienfroidefond/stripstream-librarian
1
0
Fork 0
Code Issues 4 Pull Requests Actions Packages Projects Releases Wiki Activity

perf(parsers): optimiser listing CBZ avec file_names(), ajouter magic bytes check RAR

Browse Source 
- Remplacer by_index() par file_names() pour lister les pages ZIP (zero I/O)
- Ajouter vérification magic bytes avant fallback RAR
- Ajouter tracing debug logs dans parsers
- Script docker-push avec version bump interactif

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Julien Froidefond
2026-03-15 13:01:04 +01:00
parent c7f3ad981d
commit 61bc307715
3 changed files with 86 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
crates/parsers/Cargo.toml
View File

@@ -14,3 +14,4 @@ regex = "1"
unrar.workspace = true
zip = { version = "8", default-features = false, features = ["deflate"] }
flate2 = "1"
tracing.workspace = true

35
crates/parsers/src/lib.rs
View File

@@ -166,13 +166,30 @@ fn analyze_cbz(path: &Path, allow_fallback: bool) -> Result<(i32, Vec<u8>)> {
Ok(a) => a,
Err(zip_err) => {
if allow_fallback {
// Try RAR fallback first (file might be a RAR with .cbz extension)
tracing::debug!(target: "extraction", "[EXTRACTION] ZipArchive::new failed for {}: {} — trying fallbacks", path.display(), zip_err);
// Check magic bytes to avoid expensive RAR probe on ZIP files
let is_zip_magic = std::fs::File::open(path)
.and_then(|mut f| {
let mut magic = [0u8; 4];
std::io::Read::read_exact(&mut f, &mut magic)?;
Ok(magic[0] == b'P' && magic[1] == b'K')
})
.unwrap_or(false);
if !is_zip_magic {
// Try RAR fallback (file might be a RAR with .cbz extension)
if let Ok(result) = analyze_cbr(path, false) {
tracing::debug!(target: "extraction", "[EXTRACTION] RAR fallback succeeded for {}", path.display());
return Ok(result);
}
}
// Try streaming fallback: read local file headers without central directory
// (handles ZIP files with NTFS extra fields that confuse the central dir parser)
let t0 = std::time::Instant::now();
if let Ok(result) = analyze_cbz_streaming(path) {
tracing::debug!(target: "extraction", "[EXTRACTION] Streaming fallback succeeded for {} — {} pages in {:.0}ms", path.display(), result.0, t0.elapsed().as_secs_f64() * 1000.0);
return Ok(result);
}
}
@@ -180,17 +197,11 @@ fn analyze_cbz(path: &Path, allow_fallback: bool) -> Result<(i32, Vec<u8>)> {
}
};
let mut image_names: Vec<String> = Vec::new();
for i in 0..archive.len() {
let entry = match archive.by_index(i) {
Ok(e) => e,
Err(_) => continue, // skip corrupted entries
};
let name = entry.name().to_ascii_lowercase();
if is_image_name(&name) {
image_names.push(entry.name().to_string());
}
}
let mut image_names: Vec<String> = archive
.file_names()
.filter(|name| is_image_name(&name.to_ascii_lowercase()))
.map(|name| name.to_string())
.collect::<Vec<_>>();
image_names.sort_by(|a, b| natord::compare(a, b));
if image_names.is_empty() {

67
scripts/docker-push.sh
View File

@@ -4,29 +4,75 @@ set -e
# Docker Hub
REGISTRY="docker.io"
OWNER="julienfroidefond32"
VERSION=$(grep '^version = ' Cargo.toml | head -1 | sed 's/version = "\(.*\)"/\1/')
SERVICES=("api" "indexer" "backoffice")
# ─── Version bump ───────────────────────────────────────────────────────────
CURRENT_VERSION=$(grep '^version = ' Cargo.toml | head -1 | sed 's/version = "\(.*\)"/\1/')
IFS='.' read -r MAJOR MINOR PATCH <<< "$CURRENT_VERSION"
echo "=== Stripstream Librarian Docker Push ==="
echo "Current version: $CURRENT_VERSION"
echo ""
echo "Bump version?"
echo " 1) patch → $MAJOR.$MINOR.$((PATCH + 1))"
echo " 2) minor → $MAJOR.$((MINOR + 1)).0"
echo " 3) major → $((MAJOR + 1)).0.0"
echo " 4) skip → keep $CURRENT_VERSION"
echo ""
read -rp "Choice [1-4]: " BUMP_CHOICE
case "$BUMP_CHOICE" in
1) NEW_VERSION="$MAJOR.$MINOR.$((PATCH + 1))" ;;
2) NEW_VERSION="$MAJOR.$((MINOR + 1)).0" ;;
3) NEW_VERSION="$((MAJOR + 1)).0.0" ;;
4) NEW_VERSION="$CURRENT_VERSION" ;;
*) echo "Invalid choice"; exit 1 ;;
esac
if [ "$NEW_VERSION" != "$CURRENT_VERSION" ]; then
echo "Bumping version: $CURRENT_VERSION → $NEW_VERSION"
# Update Cargo.toml (workspace version)
sed -i.bak "s/^version = \"$CURRENT_VERSION\"/version = \"$NEW_VERSION\"/" Cargo.toml
rm -f Cargo.toml.bak
# Update backoffice package.json
sed -i.bak "s/\"version\": \"$CURRENT_VERSION\"/\"version\": \"$NEW_VERSION\"/" apps/backoffice/package.json
rm -f apps/backoffice/package.json.bak
# Update Cargo.lock
cargo update --workspace 2>/dev/null || true
# Commit version bump
git add Cargo.toml Cargo.lock apps/backoffice/package.json
git commit -m "chore: bump version to $NEW_VERSION"
echo "✓ Version bumped and committed"
else
echo "Keeping version $CURRENT_VERSION"
fi
VERSION="$NEW_VERSION"
echo ""
echo "Registry: $REGISTRY"
echo "Version: $VERSION"
echo "Services: ${SERVICES[@]}"
echo "Services: ${SERVICES[*]}"
echo ""
for service in "${SERVICES[@]}"; do
echo "Building $service..."
docker build -f apps/$service/Dockerfile -t $service:latest .
docker build -f "apps/$service/Dockerfile" -t "$service:latest" .
echo "Tagging $service..."
docker tag $service:latest $REGISTRY/$OWNER/stripstream-$service:$VERSION
docker tag $service:latest $REGISTRY/$OWNER/stripstream-$service:latest
docker tag "$service:latest" "$REGISTRY/$OWNER/stripstream-$service:$VERSION"
docker tag "$service:latest" "$REGISTRY/$OWNER/stripstream-$service:latest"
echo "Pushing stripstream-$service:$VERSION..."
docker push $REGISTRY/$OWNER/stripstream-$service:$VERSION
docker push "$REGISTRY/$OWNER/stripstream-$service:$VERSION"
echo "Pushing stripstream-$service:latest..."
docker push $REGISTRY/$OWNER/stripstream-$service:latest
docker push "$REGISTRY/$OWNER/stripstream-$service:latest"
echo "✓ $service pushed successfully"
echo ""
@@ -38,3 +84,10 @@ for service in "${SERVICES[@]}"; do
echo " - $REGISTRY/$OWNER/stripstream-$service:$VERSION"
echo " - $REGISTRY/$OWNER/stripstream-$service:latest"
done
# Tag git with version
if ! git tag -l "v$VERSION" | grep -q "v$VERSION"; then
git tag "v$VERSION"
echo ""
echo "Git tag v$VERSION created. Push with: git push origin v$VERSION"
fi