def check_player_against_rankings(player, rankings):
rankings = csv.DictReader(open(rankings))
matches = []
#print("checking for \"%s\" in rankings with initial search" % player)
for row in rankings:
split_name = player.lower().split(" ")
stored_name = row['playername']
# using this for the refinement search so we get better results
stored_name_lower = row['playername'].lower()
# remove any single character tuples as it will unnecessarily cause
# it to search any row with that character in the name
split_name = [el for el in split_name if len(el) > 1]
if any(s in stored_name_lower for s in split_name):
match_score = name_tools.match(player, stored_name)
# if we get a perfect match, automatically return that
if match_score == 1.0:
return (stored_name)
elif match_score > 0.61:
#print(match_score, stored_name)
matches.append((match_score, stored_name))
if matches == []:
#print("could not find a match for \"%s\"" % player)
return None
else:
sorted_matches = sorted(matches, key=lambda tup: tup[0], reverse=True)
best_match = sorted_matches[0]
#print("best match for %s is %s" % (player, best_match))
return best_match[1]
def get(self, name):
print(name)
t = 0.1
with open(config.database) as json_file:
data = json.load(json_file)
for p in data['nodes']:
dist = name_tools.match(p['id'], name)
if dist > t: t = dist
return t * 10
def check_player_by_nickname(player, nicknames):
#print("Seeing if \"%s\" is a known nickname..." % player)
nicknames = csv.DictReader(open(nicknames))
for row in nicknames:
stored_nickname = row['nickname']
match_score = name_tools.match(player, stored_nickname)
if match_score >= 0.95:
identified_player = row['playername']
#print("Found a match: " + player + "=" + stored_nickname,identified_player, match_score)
return identified_player
else:
#("Didn't find %s in nicknames list" % player)
return None
def first_pass(self):
# first check for an explicit same as
if self.value1 and self.value2:
self.match.append('match= ' + self.value1 + ' and ' + self.value2 + ' --')
if self.sameas(self.value1, self.value2):
self.match.append('match-sameas--confidence=' + str(
1.0) + '--There is an explicit sameas (or equivalent) relationship between ' + self.value1label + ' (' + self.value1 + ') and ' + self.value2label + ' (' + self.value1 + '). These are the same person.--')
self.overallconfidence = 1.0
return self.match
# if no explicit sameas is in place, check the names
else:
self.match.append('match-sameas--There is no explicit sameas relationship between ' + self.value1label + ' (' + self.value1 + ') and ' + self.value2label + ' (' + self.value2 + '). Invstigate further ... --')
names1 = self.checkname(self.value1, 1)
names2 = self.checkname(self.value2, 2)
scores = []
if names1 and names2:
for nm in names1:
for n in names2:
# get a confidence score using name_tools
scores.append(name_tools.match(name_tools.canonicalize(nm), name_tools.canonicalize(n)))
confidence = 0
for score in scores:
if score > confidence:
confidence = score
if confidence >= 0.75:
self.match.append('match-name--confidence=' + str(
confidence) + '--The names are alike. Checking for dates ... --')
elif confidence < 0.75 and confidence > 0.5:
self.match.append('match-name--confidence=' + str(
confidence) + '--The names have some similarities, they could be the same person. Checking dates to be sure ... --')
elif confidence < 0.5:
self.match.append('match-name--confidence=' + str(
confidence) + '--The names are different, indicating that these are not likely to be the same person. Checking dates to be sure ... --')
self.overallconfidence = score
# check the dates
self.dates1 = self.checkdates(self.value1)
self.dates2 = self.checkdates(self.value2)
# if there are no dates for either check for sameas links to other sources
if not self.dates1 and not self.dates2:
self.match.append('match-dates--Neither of the names have associated dates. Checking for sameas links to other sources ... )' + '--')
# run a second pass on value 1
dates = self.secondpass(self.value1)
if dates:
for d in dates:
self.dates1 = d
self.match.append(
'match-dates--These dates were found: ' + d + ' but there is nothing to compare them to--')
else:
self.match.append(
'match-dates--No dates were found for ' + self.value1label + ' (' + self.value1 + ')--')
# run a second pass on value 2
dates = self.secondpass(self.value2)
if dates:
for d in dates:
self.dates2 = d
if self.dates1 and d in self.dates1:
self.match.append(
'match-dates--confidence=1.0--There is an exact match with birth and death dates (' + str(d) + ' and ' + str(self.dates1) + '). These are very likely the same person.--')
elif self.dates1 and d not in self.dates1:
self.match.append(
'match-dates--confidence=0--The dates do not match (' + str(d) + ' and ' + str(self.dates1) +'). These are not likely to be the same person.--')
else:
self.match.append(
'match-dates--These dates were found: ' + d + ' but there is nothing to compare them to--')
else:
self.match.append(
'match-dates--No dates were found for ' + self.value2label + ' (' + self.value2 + ')--')
#return self.match
return self.match
# if there is one date, look for sameas links for the other
elif not self.dates1 and self.dates2:
self.match.append('match-dates--There are no birth or death dates for ' + self.value1label + ' (' + self.value1 + '). Checking for sameas links to other sources ... --')
# we could have more than one
dates = self.secondpass(self.value1)
if dates:
for d in dates:
if d in self.dates2:
self.match.append(
'match-dates--confidence=1.0--There is an exact match with birth and death dates (' + str(d) + ' and ' + str(self.dates2) + '). These are very likely the same person.--')
self.overallconfidence = 1.0
return self.match
else:
self.match.append(
'match-dates--confidence=0--The dates do not match (' + str(d) + ' and ' + str(self.dates2) +'). These are not likely to be the same person.--')
self.overallconfidence = 0
return self.match
else:
self.match.append(
'match-dates--confidence=0--We only have one date (' + str(self.dates2) + ') and nothing to compare it to.--')
self.overallconfidence = 0
return self.match
# if there is one date, look for sameas links for the other
elif not self.dates2 and self.dates1:
self.match.append('match-dates--There are no birth or death dates for ' + self.value2label + ' (' + self.value2 + '). Checking for sameas links to other sources ... --')
dates = self.secondpass(self.value2)
if dates:
for d in dates:
if d in self.dates1:
self.match.append(
'match-dates--confidence=1.0--There is an exact match with birth and death dates (' + str(self.dates1) + ' and ' + str(d) + '). These are very likely the same person.--')
self.overallconfidence = 1.0
return self.match
else:
self.match.append(
'match-dates--confidence=0--The dates do not match (' + str(self.dates1) + ' and ' + str(d) +'). These are not likely to be the same person.--')
self.overallconfidence = 0
return self.match
else:
self.match.append(
'match-dates--confidence=0--We only have one date (' + str(self.dates1) + ') and nothing to compare it to.--')
return self.match
# otherwise we have two dates, test them
else:
for dt in self.dates1:
for d in self.dates2:
if dt == d:
self.match.append('match-dates--confidence=' + str(
1.0) + '--There is an exact match with birth and death dates (' + str(self.dates1) + ' and ' + str(self.dates2) + '). These are very likely the same person.--')
self.overallconfidence = 1.0
return self.match
else:
self.match.append('match-dates--confidence=' + str(
0) + '--The dates do not match (' + str(self.dates1) + ' and ' + str(self.dates2) + '). These are not likely to be the same person.--')
self.overallconfidence = 0
return self.match
# if we don't have a name one or both there's not much more we can do
else:
if not names1 and not names2:
self.match.append('match-name--There are no names specified--')
elif not names1:
self.match.append('match-name--There is no name for ' + self.value1label + ' (' + self.value1 + ')--')
elif not names2:
self.match.append('match-name--There is no name for ' + self.value2label + ' (' + self.value2 + ')--')
return self.match
else:
count = self.listcount - 1
if count < 2:
self.match.append('match--insufficient items in the list count=' + str(count))
return self.match
