Create utils/metadata_utils.py
Browse files- utils/metadata_utils.py +35 -0
utils/metadata_utils.py
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# utils/metadata_utils.py
|
| 2 |
+
|
| 3 |
+
from datetime import datetime
|
| 4 |
+
from urllib.parse import urlparse
|
| 5 |
+
|
| 6 |
+
def enhance_metadata(metadata):
|
| 7 |
+
"""
|
| 8 |
+
Enhance metadata with inferred fields based on known patterns.
|
| 9 |
+
|
| 10 |
+
Args:
|
| 11 |
+
metadata (dict): Original metadata dictionary
|
| 12 |
+
|
| 13 |
+
Returns:
|
| 14 |
+
dict: Enhanced metadata with additional standardized fields
|
| 15 |
+
"""
|
| 16 |
+
enhanced = metadata.copy()
|
| 17 |
+
|
| 18 |
+
url = enhanced.get("url", "")
|
| 19 |
+
parsed_url = urlparse(url)
|
| 20 |
+
|
| 21 |
+
# Set institution and short name
|
| 22 |
+
if "american.edu" in parsed_url.netloc:
|
| 23 |
+
enhanced.setdefault("institution", "American University")
|
| 24 |
+
enhanced.setdefault("institution_short", "AU")
|
| 25 |
+
|
| 26 |
+
# Determine source system from subdomain or path
|
| 27 |
+
if "catalog" in parsed_url.netloc:
|
| 28 |
+
enhanced.setdefault("source_system", "Course Catalog")
|
| 29 |
+
else:
|
| 30 |
+
enhanced.setdefault("source_system", "University Website")
|
| 31 |
+
|
| 32 |
+
# Standardized timestamp (ISO 8601)
|
| 33 |
+
enhanced.setdefault("timestamp", datetime.now().isoformat())
|
| 34 |
+
|
| 35 |
+
return enhanced
|