Commit ca436213 authored by PkSM3's avatar PkSM3

[BUGFIX] pubdates format for istex

parent 15dbf721
......@@ -27,8 +27,11 @@ class ISTex(FileParser):
"authorsRAW" : 'author',
"keywords" : "keywords"
}
suma = 0
for json_doc in json_docs:
hyperdata = {}
for key, path in hyperdata_path.items():
try:
......@@ -83,33 +86,47 @@ class ISTex(FileParser):
else:
hyperdata["language_iso3"] = "eng"
RealDate = hyperdata["publication_date"]
if "publication_date" in hyperdata: hyperdata.pop("publication_date")
if "publication_date" in hyperdata:
RealDate = hyperdata["publication_date"]
if "publication_date" in hyperdata:
hyperdata.pop("publication_date")
Decision=""
if len(RealDate)>4:
if len(RealDate)>8:
try: Decision = datetime.strptime(RealDate, '%Y-%b-%d').date()
except:
try: Decision = datetime.strptime(RealDate, '%Y-%m-%d').date()
except: Decision=False
if isinstance(RealDate, list):
RealDate = RealDate[0]
# print( RealDate ," | length:",len(RealDate))
Decision=""
if len(RealDate)>4:
if len(RealDate)>8:
try: Decision = datetime.strptime(RealDate, '%Y-%b-%d').date()
except:
try: Decision = datetime.strptime(RealDate, '%Y-%m-%d').date()
except: Decision=False
else:
try: Decision = datetime.strptime(RealDate, '%Y-%b').date()
except:
try: Decision = datetime.strptime(RealDate, '%Y-%m').date()
except: Decision=False
else:
try: Decision = datetime.strptime(RealDate, '%Y-%b').date()
except:
try: Decision = datetime.strptime(RealDate, '%Y-%m-%d').date()
except: Decision=False
else:
try: Decision = datetime.strptime(RealDate, '%Y-%m-%d').date()
except: Decision=False
if Decision!=False:
hyperdata["publication_year"] = str(Decision.year)
hyperdata["publication_month"] = str(Decision.month)
hyperdata["publication_day"] = str(Decision.day)
hyperdata_list.append(hyperdata)
# print("\t||",hyperdata["title"])
# print("\t\t",Decision)
# print("=============================")
try: Decision = datetime.strptime(RealDate, '%Y').date()
except: Decision=False
if Decision!=False:
hyperdata["publication_year"] = str(Decision.year)
hyperdata["publication_month"] = str(Decision.month)
hyperdata["publication_day"] = str(Decision.day)
hyperdata_list.append(hyperdata)
# print("\t||",hyperdata["title"])
# print("\t\t",Decision)
# print("=============================")
# else:
# suma+=1
# if "pubdate" in json_doc:
# print ("\tfail pubdate:",json_doc["pubdate"])
# print ("nb_hits:",len(json_docs))
# print("\t - nb_fails:",suma)
# print(" -- - - - - - -- - -")
return hyperdata_list
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment