evaluation for rag systems

services/rag/langchain/cli.py

@@ -1,4 +1,6 @@
 import os
+import csv
+import json
 from pathlib import Path
 
 import click
@@ -126,5 +128,60 @@ def chat(collection_name, model):
         click.echo(f"Error: {str(e)}")
 
 
+@cli.command(
+    name="export-supported-paths",
+    help="Filter JSON paths by enrichment-supported extensions and export JSON/CSV",
+)
+@click.argument("input_json", type=click.Path(exists=True, dir_okay=False, path_type=Path))
+def export_supported_paths(input_json: Path):
+    """Export supported document paths into yadisk_imported_paths.json and yadisk_imported_paths.csv."""
+    logger.info(f"Filtering supported paths from input file: {input_json}")
+
+    try:
+        from enrichment import SUPPORTED_EXTENSIONS
+
+        with input_json.open("r", encoding="utf-8") as source_file:
+            raw_data = json.load(source_file)
+
+        if not isinstance(raw_data, list):
+            raise ValueError("Input JSON must contain an array of file paths")
+
+        filtered_paths = []
+        seen_paths = set()
+        for item in raw_data:
+            path_str = str(item).strip()
+            if not path_str:
+                continue
+            if path_str in seen_paths:
+                continue
+
+            extension = Path(path_str).suffix.lower()
+            if extension in SUPPORTED_EXTENSIONS:
+                filtered_paths.append(path_str)
+                seen_paths.add(path_str)
+
+        output_json = Path.cwd() / "yadisk_imported_paths.json"
+        output_csv = Path.cwd() / "yadisk_imported_paths.csv"
+
+        with output_json.open("w", encoding="utf-8") as output_json_file:
+            json.dump(filtered_paths, output_json_file, ensure_ascii=False, indent=2)
+
+        with output_csv.open("w", encoding="utf-8", newline="") as output_csv_file:
+            writer = csv.writer(output_csv_file)
+            writer.writerow(["path"])
+            for path_item in filtered_paths:
+                writer.writerow([path_item])
+
+        click.echo(
+            f"Export complete: {len(filtered_paths)} supported paths saved to {output_json.name} and {output_csv.name}"
+        )
+        logger.info(
+            f"Exported {len(filtered_paths)} supported paths to {output_json} and {output_csv}"
+        )
+    except Exception as error:
+        logger.error(f"Failed to export supported paths: {error}")
+        click.echo(f"Error: {error}")
+
+
 if __name__ == "__main__":
     cli()