Spaces:
Running
Running
Update smart_fallback.py
Browse files- smart_fallback.py +9 -1
smart_fallback.py
CHANGED
|
@@ -3,6 +3,7 @@ from Bio import Entrez, Medline
|
|
| 3 |
import mtdna_classifier
|
| 4 |
from NER.html import extractHTML
|
| 5 |
import data_preprocess
|
|
|
|
| 6 |
# Setup
|
| 7 |
def fetch_ncbi(accession_number):
|
| 8 |
try:
|
|
@@ -168,7 +169,14 @@ def filter_links_by_metadata(search_results, saveLinkFolder, accession=None):
|
|
| 168 |
keywords = [accession] + keywords
|
| 169 |
title_snippet = link.lower()
|
| 170 |
print("save link folder inside this filter function: ", saveLinkFolder)
|
| 171 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 172 |
print("article text")
|
| 173 |
#print(article_text)
|
| 174 |
try:
|
|
|
|
| 3 |
import mtdna_classifier
|
| 4 |
from NER.html import extractHTML
|
| 5 |
import data_preprocess
|
| 6 |
+
import pipeline
|
| 7 |
# Setup
|
| 8 |
def fetch_ncbi(accession_number):
|
| 9 |
try:
|
|
|
|
| 169 |
keywords = [accession] + keywords
|
| 170 |
title_snippet = link.lower()
|
| 171 |
print("save link folder inside this filter function: ", saveLinkFolder)
|
| 172 |
+
success_process, output_process = pipeline.run_with_timeout(data_preprocess.extract_text,args=(link,saveLinkFolder),timeout=60)
|
| 173 |
+
if success_process:
|
| 174 |
+
article_text = output_process
|
| 175 |
+
print("yes succeed for getting article text")
|
| 176 |
+
else:
|
| 177 |
+
print("no suceed, fallback to no link")
|
| 178 |
+
article_text = ""
|
| 179 |
+
#article_text = data_preprocess.extract_text(link,saveLinkFolder)
|
| 180 |
print("article text")
|
| 181 |
#print(article_text)
|
| 182 |
try:
|