bellingcat/snscrape
1
0
Fork 0
mirror of https://github.com/bellingcat/snscrape.git synced 2026-06-08 02:28:29 +03:00
Code Issues Packages Projects Releases Wiki Activity

Get more data from the place

Browse Source 
Data like the country, place type and the single place name are now also returned on the JSON.
This commit is contained in:
NoeCampos22
2021-04-19 12:01:14 -05:00
parent 39a34a57ac
commit dbf2a2f689
1 changed files with 9 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
snscrape/modules/twitter.py
View File

@@ -43,6 +43,10 @@ class Tweet(snscrape.base.Item):
mentionedUsers: typing.Optional[typing.List['User']] = None mentionedUsers: typing.Optional[typing.List['User']] = None
coordinates: typing.Optional['Coordinates'] = None coordinates: typing.Optional['Coordinates'] = None
place: typing.Optional[str] = None place: typing.Optional[str] = None
place_name: typing.Optional[str] = None
place_type: typing.Optional[str] = None
country: typing.Optional[str] = None
country_code: typing.Optional[str] = None
username = snscrape.base._DeprecatedProperty('username', lambda self: self.user.username, 'user.username') username = snscrape.base._DeprecatedProperty('username', lambda self: self.user.username, 'user.username')
outlinksss = snscrape.base._DeprecatedProperty('outlinksss', lambda self: ' '.join(self.outlinks), 'outlinks') outlinksss = snscrape.base._DeprecatedProperty('outlinksss', lambda self: ' '.join(self.outlinks), 'outlinks')
@@ -378,7 +382,12 @@ class TwitterAPIScraper(snscrape.base.Scraper):
if (coords := tweet['geo']['coordinates']) and len(coords) == 2: if (coords := tweet['geo']['coordinates']) and len(coords) == 2:
kwargs['coordinates'] = Coordinates(coords[1], coords[0]) kwargs['coordinates'] = Coordinates(coords[1], coords[0])
if tweet.get('place'): if tweet.get('place'):
# get place data
kwargs['place'] = tweet['place']['full_name'] kwargs['place'] = tweet['place']['full_name']
kwargs['place_name'] = tweet['place']['name']
kwargs['place_type'] = tweet['place']['place_type']
kwargs['country'] = tweet['place']['country']
kwargs['country_code'] = tweet['place']['country_code']
if 'coordinates' not in kwargs and tweet['place']['bounding_box'] and (coords := tweet['place']['bounding_box']['coordinates']) and coords[0] and len(coords[0][0]) == 2: if 'coordinates' not in kwargs and tweet['place']['bounding_box'] and (coords := tweet['place']['bounding_box']['coordinates']) and coords[0] and len(coords[0][0]) == 2:
# Take the first (longitude, latitude) couple of the "place square" # Take the first (longitude, latitude) couple of the "place square"
kwargs['coordinates'] = Coordinates(coords[0][0][0], coords[0][0][1]) kwargs['coordinates'] = Coordinates(coords[0][0][0], coords[0][0][1])