Initial commit: document_analyzer with CI/CD pipeline

- 4 skill pipeline (doc_parser, conflict_detection, ir_generation, resolution_application) - CI workflow on push/PR (.gitea/workflows/ci.yml) - Auto-issue on CI failure (.gitea/workflows/auto-issue.yml) - Pytest smoke tests (tests/test_sample.py) Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-05-29 20:00:26 +08:00
commit 40567a4fb6
22 changed files with 2898 additions and 0 deletions
@@ -0,0 +1,49 @@
+import argparse
+import sys
+import yaml
+
+
+def print_ir(yaml_path: str) -> None:
+    with open(yaml_path, "r", encoding="utf-8") as f:
+        data = yaml.safe_load(f)
+
+    if not isinstance(data, list):
+        print(f"Expected a YAML list, got {type(data).__name__}")
+        sys.exit(1)
+
+    for i, item in enumerate(data):
+        yc = item.get("converted")
+        if yc is None:
+            continue
+
+        # yaml_content is a raw string that happens to look like YAML —
+        # parse it then pretty-print it back
+        if isinstance(yc, str):
+            stripped = yc.strip()
+            # Strip markdown code fences: ```yaml / ``` at start, ``` at end
+            if stripped.startswith("```"):
+                first_newline = stripped.find("\n")
+                stripped = stripped[first_newline + 1:] if first_newline != -1 else stripped[3:]
+            if stripped.endswith("```"):
+                stripped = stripped[:-3]
+            try:
+                parsed = yaml.safe_load(stripped)
+            except yaml.YAMLError:
+                print(yc)
+                raise
+        else:
+            parsed = yc
+
+        if isinstance(parsed, (dict, list)):
+            print(yaml.dump(parsed, allow_unicode=True, default_flow_style=False, sort_keys=False).rstrip())
+        else:
+            print(str(parsed))
+
+        print("=" * 60)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Pretty-print yaml_content fields from an analysis YAML file.")
+    parser.add_argument("input", metavar="file.yaml", help="Path to a _文字.yaml or _图片.yaml file")
+    args = parser.parse_args()
+    print_ir(args.input)