def lookup_wiktionary(word):
logger = logging.getLogger(__name__)
try:
wikipedia = MediaWiki()
wikipedia.set_api_url('https://en.wiktionary.org/w/api.php')
matches = {}
search_results = wikipedia.opensearch(word)
if len(search_results) > 0:
page_title = search_results[0][0]
page = wikipedia.page(page_title)
parts = page.content.split("\n")
i = 0
while i < len(parts):
definition = ""
part = parts[i].strip()
if part.startswith("=== Verb ===") or part.startswith(
"=== Noun ===") or part.startswith(
"=== Adjective ==="):
#print(part)
# try to skip the first two lines after the marker
if (i + 1) < len(parts):
definition = parts[i + 1]
if (i + 2) < len(parts) and len(parts[i + 2].strip()) > 0:
definition = parts[i + 2]
if (i + 3) < len(parts) and len(parts[i + 3].strip()) > 0:
definition = parts[i + 3]
if part.startswith(
"=== Adjective ===") and not 'adjective' in matches:
matches['adjective'] = definition
if part.startswith("=== Noun ===") and not 'noun' in matches:
matches['noun'] = definition
if part.startswith("=== Verb ===") and not 'verb' in matches:
matches['verb'] = definition
i = i + 1
final = ""
# prefer verb, noun then adjective
if matches.get('adjective', False):
final = matches.get('adjective')
if matches.get('noun', False):
final = matches.get('noun')
if matches.get('verb', False):
final = matches.get('verb')
# strip leading bracket comment
if final[0] == '(':
close = final.index(")") + 1
final = final[close:]
matches['definition'] = final
return matches
except:
e = sys.exc_info()
logger.debug(e)
async def send_to_wikipedia(word, site):
logger = logging.getLogger(__name__)
try:
# lookup in wiktionary and send display message
wikipedia = MediaWiki()
wikipedia.set_api_url('https://en.wiktionary.org/w/api.php')
matches = {}
search_results = wikipedia.opensearch(word)
# logger.debug(search_results)
if len(search_results) > 0:
page_title = search_results[0][0]
page_link = search_results[0][2]
# page = wikipedia.page(page_title)
# parts = page.content.split("\n")
# logger.debug([page_title,page_link])
await publish('hermod/' + site + '/display/show',
{'frame': page_link})
except:
e = sys.exc_info()
logger.debug(e)
"""Named-entity recognition with SpaCy"""
from collections import defaultdict
from sys import argv
import spacy
from mediawiki import MediaWiki
wikipedia = MediaWiki()
wikipedia.set_api_url('https://fr.wikipedia.org/w/api.php')
nlp = spacy.load('fr_core_news_sm')
def mediawiki(ask):
search = wikipedia.search(ask)
if len(search) >= 1:
try:
print(f'[ {search[0]} ]')
p = wikipedia.page(str(search[0]))
except:
print(f'[ {search[1]} ]')
p = wikipedia.page(str(search[1]))
print(p.summary)
if len(search) > 1:
print(f'{ask} may refer to : {search}')
else:
print(f'There were no results matching with {ask}.')
def ren(input):
text = open(f"{dataP}{input}").read()[:1000000]
doc = nlp(text)
people = defaultdict(int)