Update mcp/nlp.py
Browse files- mcp/nlp.py +14 -7
mcp/nlp.py
CHANGED
|
@@ -1,14 +1,21 @@
|
|
| 1 |
# mcp/nlp.py
|
| 2 |
|
| 3 |
import spacy
|
|
|
|
| 4 |
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
def extract_keywords(text: str):
|
| 11 |
-
"""Extract biomedical
|
| 12 |
doc = nlp(text)
|
| 13 |
-
|
| 14 |
-
return keywords
|
|
|
|
| 1 |
# mcp/nlp.py
|
| 2 |
|
| 3 |
import spacy
|
| 4 |
+
import subprocess
|
| 5 |
|
| 6 |
+
def load_model():
|
| 7 |
+
try:
|
| 8 |
+
return spacy.load("en_core_sci_sm")
|
| 9 |
+
except OSError:
|
| 10 |
+
try:
|
| 11 |
+
subprocess.run(["python", "-m", "spacy", "download", "en_core_web_sm"], check=True)
|
| 12 |
+
return spacy.load("en_core_web_sm")
|
| 13 |
+
except Exception as e:
|
| 14 |
+
raise RuntimeError("Failed to download spaCy model. Details: " + str(e))
|
| 15 |
+
|
| 16 |
+
nlp = load_model()
|
| 17 |
|
| 18 |
def extract_keywords(text: str):
|
| 19 |
+
"""Extract biomedical or general entities from text."""
|
| 20 |
doc = nlp(text)
|
| 21 |
+
return list(set(ent.text for ent in doc.ents if len(ent.text.strip()) > 2))
|
|
|