def load_watchlist():
global watchlist, wl_hash, has_keywords
wl = []
for x in open(args.watchlist, 'r').readlines():
x = x.rstrip().lower()
if x.startswith(';'):
username = x[1:]
disabled_users[username] = True
else:
username = x
if username[0] == '#' and not has_keywords:
has_keywords = True
if not username[0] == '#' and not os.path.exists(
paths.get_user_json(username)):
new_accounts.append(username)
if not os.path.exists(paths.get_user(username)):
retry_makedirs(paths.get_user(username))
wl.append(username)
newhash = hashlib.md5(''.join(wl)).hexdigest()
if newhash != wl_hash:
print('reloading watchlist')
wl_hash = newhash
watchlist = wl
json_loads()
if has_keywords and os.path.exists('users'):
for file in os.listdir('users'):
d = os.path.join('users', file)
if os.path.isdir(d): load_user_json(file)
def load_watchlist():
global watchlist, wl_hash
wl = []
for x in open(args.watchlist, 'r').readlines():
x = x.rstrip()
if x.startswith(';'):
username = x[1:]
disabled_users[username] = True
else:
username = x
if not os.path.exists(paths.get_user_json(username)):
new_accounts.append(username)
if not os.path.exists(paths.get_user(username)):
retry_makedirs(paths.get_user(username))
wl.append(username)
newhash = hashlib.md5(''.join(wl)).hexdigest()
if newhash != wl_hash:
print('reloading watchlist')
wl_hash = newhash
watchlist = wl
json_loads()
def scrape(user, http, host):
global nitters
if user in new_accounts:
count = args.count
checkfn = None
new_accounts.remove(user)
else:
checkfn = fetch_more_tweets_callback
count = -1
elapsed_time = time.time()
insert_pos = 0
sys.stdout.write('\r[%s] scraping %s... ' % (get_timestamp("%Y-%m-%d %H:%M:%S", elapsed_time), user))
sys.stdout.flush()
twats, nitters, host, http = get_twats(user, proxies=args.proxy, count=count, http=http, checkfn=checkfn, nitters=nitters, host=host)
new = False
for t in twats:
if not in_twatlist(user, t):
new = True
if args.unshorten: t = unshorten_urls(t, proxies=args.proxy, shorteners=shorteners)
add_twatlist(user, t, insert_pos)
insert_pos += 1
if 'quote_tweet' in t:
if not os.path.isdir(paths.get_user(t[quote_tweet]['user'])): retry_makedirs(paths.get_user(t[quote_tweet]['user']))
fetch_profile_picture(t[quote_tweet]['user'], args.proxy, twhttp=nitter_rshttp, nitters=nitters)
if 'user' in t:
if not os.path.isdir(paths.get_user(t['user'])): retry_makedirs(paths.get_user(t['user']))
fetch_profile_picture(t['user'], args.proxy, twhttp=nitter_rshttp, nitters=nitters)
if args.mirror: mirror_twat(t, args=args)
sys.stdout.write('\r[%s] scraping %s... +%d ' % (get_timestamp("%Y-%m-%d %H:%M:%S", elapsed_time), user, insert_pos))
sys.stdout.flush()
if new: write_user_tweets(user)
elapsed_time = (time.time() - elapsed_time)
sys.stdout.write('done (%s)\n' % get_timestamp("%H:%M:%S", elapsed_time))
sys.stdout.flush()
return http, host
def htmlize_twat(twat, variables, quoted=False):
tw = '<div class="twat-container">'
tweet_pic = None
retweet_pic = None
if not 'rid' in twat:
retweet_str = ""
if paths.has_profile_pic(twat['owner']): tweet_pic = paths.get_profile_pic(twat['owner'])
else:
if paths.has_profile_pic(twat['user']): tweet_pic = paths.get_profile_pic(twat['user'])
else: tweet_pic = "data:image/gif;base64,R0lGODdhAQABAIAAAP///////ywAAAAAAQABAAACAkQBADs="
if paths.has_profile_pic(twat['owner']): retweet_pic = paths.get_profile_pic(twat['owner'])
retweet_str = " (RT %s<a target='_blank' href='https://twitter.com/%s/status/%s'>%s</a>)" % \
(user_at_link(twat['user']), twat['user'], twat['id'], twat['user'])
if tweet_pic: tw += '<div class="profile_picture"><img width="100%%" height="100%%" src="%s"></div>' % tweet_pic
if retweet_pic: tw += '<div class="profile_picture_retweet"><img width="100%%" height="100%%" src="%s"></div>' % retweet_pic
user_str = user_at_link(twat["owner"])
user_str += "<a target='_blank' href='https://twitter.com/%s/status/%s'>%s</a>%s" % \
(twat["owner"], get_effective_twat_id(twat), twat["owner"], retweet_str)
tw += '\n<div class="twat-title">'
## add icon bar
if args.iconbar: tw += build_iconbar(twat, variables, quoted)
time_str = 'unknown' if twat["time"] == 0 else format_time(twat["time"])
tw += '%s - %s' % (user_str, time_str)
tw += '\n</div>\n'
## replace urls in twats
twat['text'] = replace_url_in_twat(twat, args=args)
## strip html ?
if args.nohtml: twat['text']= strip_tags(twat['text'])
tw += '<p class="twat-text">%s</p>\n' % (replace_twat_text(twat['text']))
if 'curl' in twat and args.iframe > 0:
user = twat['user'].lower()
ifu = paths.get_user(user) + '/%s-%s' % (twat['id'], "card.html")
if (not 'c' in args.mirror) or (not file_exists(ifu)):
ifu = twat['curl']
tw += '<span class="twat-iframe"><iframe src="%s"></iframe></span>\n'%ifu
if 'images' in twat:
tw += '<p class="twat-image">'
if len(twat['images']) > 1: wdth = (100/len(twat['images'])) - 1
else: wdth = 100
for i in twat['images']:
if args.images <= 0:
tw += '<a href="%s">%s</a>'%(i, i)
else:
img_path = paths.get_user(twat['user']) + "/%s-%s" % (twat['id'], i.split('/')[-1])
if not file_exists(img_path): img_path = i
span_or_div = "span"
img_class = "img"
div_class = ""
if args.upstream_img:
href = i
title = "view remote image"
elif 'video' in twat or 'ext_tw_video_thumb' in i:
mp4_path = paths.get_user(twat['user']) + '/%s.mp4' % str(twat['id'])
if os.path.exists(mp4_path):
href = mp4_path
title = "view local video"
else:
href = "https://twitter.com/i/status/" + twat['id']
title = "view remote video"
img_class = ""
div_class = "video-thumbnail"
span_or_div = "div"
else:
href = img_path
title = "view local image"
tw += '<a href="%s" title="%s"><%s class="%s"><img class="%s" src="%s" width="%d%%"></%s></a>' % (href, title, span_or_div, div_class, img_class, img_path, wdth, span_or_div)
tw += '</p>\n'
if 'quote' in twat:
pseudo_twat = {
'user' : twat['quote']['user'],
'owner' : twat['quote']['user'],
'id' : twat['quote']['id'],
'text' : twat['quote']['text'],
'time' : 0
}
tw += htmlize_twat(pseudo_twat, variables, quoted=True)
tw += '</div>\n'
return tw
def scrape(item, http, host, search, user_agent):
global nitters
global mastodon_rshttp
item = item.lower()
if item in new_accounts:
count = args.count
checkfn = None
new_accounts.remove(item)
else:
checkfn = fetch_more_tweets_callback
count = args.count if item[0] == '#' else -1
if item.count('@') < 2:
platform = 'twitter'
twats, nitters, host, http, page = get_twats(item,
proxies=args.proxy,
count=count,
http=http,
checkfn=checkfn,
nitters=nitters,
host=host,
search=search,
user_agent=user_agent,
blacklist=blacklist,
whitelist=whitelist)
else:
platform = 'mastodon'
twats, http = get_toots(item,
proxies=args.proxy,
count=count,
http=http,
checkfn=checkfn,
user_agent=user_agent,
blacklist=args.blacklist,
whitelist=args.whitelist)
mastodon_rshttp[host] = http
insert_pos = dict()
new = False
user = None if item[0] == '#' else item
insert_pos_total = 0
elapsed_time = time.time()
for t in twats:
if item[0] == '#': user = t['user'].lower()
if not user in insert_pos: insert_pos[user] = 0
if not in_twatlist(user, t):
new = True
if args.unshorten:
t = unshorten_urls(t,
proxies=args.proxy,
shorteners=shorteners)
add_twatlist(user, t, insert_pos[user])
insert_pos[user] += 1
insert_pos_total += 1
if 'quote_tweet' in t:
if '@' in t['quote_tweet']['user']:
_, foo, bar = t['quote_tweet']['user'].split('@')
http = None if not bar in mastodon_rshttp else mastodon_rshttp[
bar]
if not os.path.isdir(paths.get_user(t[quote_tweet]['user'])):
retry_makedirs(paths.get_user(t[quote_tweet]['user']))
fetch_profile_picture(t[quote_tweet]['user'],
args.proxy,
twhttp=http,
nitters=nitters,
platform=platform)
if 'user' in t:
if '@' in t['user']:
_, foo, bar = t['user'].split('@')
http = None if not bar in mastodon_rshttp else mastodon_rshttp[
bar]
if not os.path.isdir(paths.get_user(t['user'])):
retry_makedirs(paths.get_user(t['user']))
fetch_profile_picture(t['user'],
args.proxy,
twhttp=http,
nitters=nitters,
platform=platform)
if args.mirror: mirror_twat(t, args=args)
sys.stdout.write(
'\r[%s] %s: extracting from %d page(s): +%d twat(s)' %
(misc.get_timestamp("%Y-%m-%d %H:%M:%S", elapsed_time), item,
page, insert_pos_total))
sys.stdout.flush()
if new:
if item[0] == '#':
for user in insert_pos.keys():
write_user_tweets(user)
else:
write_user_tweets(item)
elapsed_time = (time.time() - elapsed_time)
sys.stdout.write('done (%s)\n' %
misc.get_timestamp("%H:%M:%S", elapsed_time))
sys.stdout.flush()
return http, host
def _mirror_file(url_components,
user,
tid,
args=None,
content_type=None,
force=False):
outname = paths.get_user(user) + '/%s-%s' % (tid,
url_components['filename'])
if not force and os.path.exists(outname):
return
http = RsHttp(url_components['host'],
ssl=url_components['ssl'],
port=url_components['port'],
keep_alive=True,
follow_redirects=True,
auto_set_cookies=True,
proxies=args.proxy,
user_agent="curl/7.60.0")
## do nothing if we cannot connect
if not http.connect(): return None
ext = url_components['filename'].split('.')[-1]
if content_type:
if args.ext: filtre = str(args.ext).split(',')
else: filtre = []
hdr = http.head(url_components['uri'])
## max mirror size
if args.mirror_size:
# extract second part of the Content-Length: line
value = [
str(i.split(':')[1]).strip() for i in hdr.split('\n')
if i.lower().startswith('content-length:')
]
if not len(value) or int(value[0]) > args.mirror_size: return
# extract second part of the Content-Type: line
value = [
str(i.split(':')[1]).strip() for i in hdr.split('\n')
if i.lower().startswith('content-type:')
]
## server does not provide Content-Type info
if not len(value):
return
# content type contains ';' (usually when html)
elif ';' in value[0]:
value[0] = value[0].split(';')[0]
value = value[0].split('/')
## when filtering extensions (--ext)
## if unset, everything is mirrored
if len(filtre):
## values don't match anything
if len(value) < 2 or (not value[0] in filtre
and not value[1] in filtre):
return
# XXX : mirror html files
## we actually don't save html files
## what about making automated save
## thru the wayback machine ?
if 'html' in value: return
## previous http object cannot be re-used
http = RsHttp(url_components['host'],
ssl=url_components['ssl'],
port=url_components['port'],
keep_alive=True,
follow_redirects=True,
auto_set_cookies=True,
proxies=args.proxy,
user_agent="curl/7.60.0")
## do nothing if we cannot connect
if not http.connect(): return
extras = []
if url_components[
'filename'] == 'card.html' and 'twitter.com' in url_components[
'host']:
extras.append("Referer: https://twitter.com/")
hdr, res = http.get(url_components['uri'], extras=extras)
if res == '' and hdr != "":
# print http error code when things go wrong
print "%s%s : %s" % (url_components['host'], url_components['uri'],
hdr.split('\n')[0])
return
res_bytes = res.encode('utf-8') if isinstance(res, unicode) else res
filehash = _hash(res_bytes)
out_fn = 'data/%s.%s' % (filehash, ext)
if not os.path.exists(out_fn):
retry_write(out_fn, res_bytes)
if os.path.lexists(outname): os.unlink(outname)
os.symlink('../../data/%s.%s' % (filehash, ext), outname)
def mirror_twat(twat, args=None):
if 'owner' in twat:
user = twat['owner'].lower()
else:
user = twat['user'].lower()
if not os.path.isdir('data'): retry_makedirs('data')
## soupify user's text
soup = soupify(twat["text"])
## try to automatically mirror links posted by the user,
## if it matches the extension list.
if 'c' in args.mirror and 'curl' in twat:
url = twat['curl']
# XXX: unsupported nitter feature
# this displays fine when loading from twitter in a regular browser,
# which is probably converted using some js code
# TODO: check if nitter handles card:// stuff..
unsuported_shemes = ['card://']
for _us in unsuported_shemes:
if url.startswith(_us): continue
url_components = _split_url(url)
url_components['filename'] = 'card.html' #% twat['id']
_mirror_file(url_components, user, twat['id'], args)
if 'f' in args.mirror:
for a in soup.body.find_all('a'):
if 'data-expanded-url' in a.attrs:
url_components = _split_url(a.attrs['data-expanded-url'])
if 'filename' in url_components:
_mirror_file(url_components,
user,
twat['id'],
args,
content_type=True)
## mirror videos
if 'v' in args.mirror and 'video' in twat:
tid = str(twat['id'])
url = 'https://twitter.com/%s/status/%s' % (twat['user'], tid)
outname = paths.get_user(twat['user']) + '/%s.mp4' % tid
if not os.path.exists('data/%s.mp4' % tid):
if args.proxy:
os.system('%s --proxy %s -o data/%s.mp4 %s > /dev/null 2>&1' %
(args.ytdl, args.rawproxy, tid, url))
else:
os.system('%s -o data/%s.mp4 %s > /dev/null 2>&1' %
(args.ytdl, tid, url))
if not os.path.exists('%s' % outname) and os.path.exists(
'data/%s.mp4' % tid):
os.symlink('../../data/%s.mp4' % tid, outname)
## mirror posted pictures
if 'images' in twat and 'i' in args.mirror:
for x in xrange(0, len(twat['images'])):
i = twat['images'][x]
if '?format=' in i:
i = i.split('&')[0]
fmt = i.split('=')[1]
i = '%s.%s' % (i.split('?')[0], fmt)
url_components = _split_url(i)
if 'filename' in url_components:
_mirror_file(url_components, user, twat['id'], args)
## deal with emojis
if 'e' in args.mirror:
for img in soup.body.find_all('img'):
if 'class' in img.attrs and 'Emoji' in img.attrs['class']:
src = img.attrs['src']
src = src.encode('utf-8') if isinstance(src, unicode) else src
split = src.split('/')
host = split[2]
emodir = '/'.join(split[3:len(split) - 1])
filename = split[-1]
uri = '%s/%s' % (emodir, filename)
if not os.path.isdir(emodir):
retry_makedirs(emodir)
if not os.path.exists('%s/%s' % (emodir, filename)):
http = RsHttp(host=host,
port=443,
timeout=30,
ssl=True,
keep_alive=True,
follow_redirects=True,
auto_set_cookies=True,
proxies=args.proxy,
user_agent="curl/7.60.0")
while not http.connect():
# FIXME : what should happen on connect error ?
pass
hdr, res = http.get('/%s' % uri)
res = res.encode('utf-8') if isinstance(res,
unicode) else res
retry_write('%s/%s' % (emodir, filename), res)
