refactor: restore prev behaviour by introducing write_summary flag

Author: 404Simon

vesskel/_batch.py

@@ -19,6 +19,6 @@ def process_one(
     image = load_image(in_path)
     result = analyze_binary_image(image=image, base_name=in_path.stem, config=config)
 
-    save_analysis_outputs(out_dir, safe_name, result, config.output)
+    save_analysis_outputs(out_dir, safe_name, result, config.output, write_summary=False)
 
     return {"image": in_path.name, **result.summary_features}

vesskel/_io.py

@@ -80,8 +80,26 @@ def save_analysis_outputs(
     base_name: str,
     result: AnalysisResult,
     config: OutputConfig,
+    *,
+    write_summary: bool = True,
 ) -> None:
-    """Write all analysis results to disk according to *config*."""
+    """Write all analysis results to disk according to *config*.
+
+    Parameters
+    ----------
+    output_dir
+        Parent directory under which a *base_name* subdirectory is created.
+    base_name
+        Used as subdirectory name and as prefix for all output files.
+    result
+        Analysis result to save.
+    config
+        Output configuration controlling which files are written.
+    write_summary
+        When True (default), also write a per-image summary CSV inside the
+        output subdirectory.  Set to False in batch mode where the caller
+        produces an aggregated summary.csv at the top level.
+    """
     out = output_dir / base_name
     out.mkdir(parents=True, exist_ok=True)
 
@@ -102,7 +120,7 @@ def save_analysis_outputs(
     if config.write_node_csv and result.node_records:
         write_csv(out / f"{base_name}_nodes.csv", result.node_records)
 
-    if config.write_summary_csv and result.summary_features:
+    if write_summary and config.write_summary_csv and result.summary_features:
         write_csv(
             out / f"{base_name}_summary.csv",
             [{"image": base_name, **result.summary_features}],