Update utils/metadata.py

utils/metadata.py

@@ -28,20 +28,35 @@ def extract_metadata(text):
     words = text.split()
     chunks = [" ".join(words[i:i + max_chunk_length]) for i in range(0, len(words), max_chunk_length)]
 
-    metadata = {
-        "DATE": [],
-        "PERSON": [],
-        "ORGANIZATION": [],
-        "LOCATION": []
-    }
-
-    for chunk in chunks:
-        ner_results = ner_pipeline(chunk)
-        for ent in ner_results:
-            label = ent["entity_group"]
-            word = ent["word"]
-            if label in metadata and word not in metadata[label]:
-                metadata[label].append(word)
-
-    return metadata
+  metadata = {
+    "EFFECTIVE_DATE": [],
+    "PARTIES": [],
+    "GOVERNING_LAW": [],  # derived from LOCATION
+}
+
+# Use label mappings
+label_mapping = {
+    "DATE": "EFFECTIVE_DATE",
+    "PERSON": "PARTIES",
+    "ORGANIZATION": "PARTIES",
+    "LOCATION": "GOVERNING_LAW"
+}
+
+for ent in ner_results:
+    label = ent["entity_group"]
+    word = ent["word"]
+    custom_label = label_mapping.get(label)
+
+    if custom_label and word not in metadata[custom_label]:
+        metadata[custom_label].append(word)
+
+import re
+
+def extract_governing_law(text):
+    match = re.search(r"(?i)governed by the laws of ([\w\s,]+)", text)
+    return match.group(1).strip() if match else None
+
+def extract_venue(text):
+    match = re.search(r"(?i)submitted to.*?in ([\w\s,]+)", text)
+    return match.group(1).strip() if match else None