library.py 19.8 KB
Newer Older
1
# -*- coding: utf-8 -*-
Janez K's avatar
Janez K committed
2
3
4
5
6
7
'''
Streaming widgets librarby

@author: Janez Kranjc <janez.kranjc@ijs.si>
'''

Janez K's avatar
Janez K committed
8
from workflows.security import safeOpen
Janez K's avatar
Janez K committed
9
10
from workflows.streaming.tripletclient import TripletClient

Janez K's avatar
Janez K committed
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137

def streaming_add_neutral_zone(input_dict):
    import copy
    tweets = copy.deepcopy(input_dict['ltw'])
    neutral_zone = float(input_dict['zone'])

    ltw = []

    for tweet in tweets:
        if tweet['reliability']!=-1.0 and tweet['reliability']<neutral_zone:
            tweet['sentiment']="Neutral"
        ltw.append(tweet)

    output_dict = {}

    output_dict['ltw']=ltw
    return output_dict

def streaming_remove_words_from_tweets(input_dict):
    import copy
    tweets = copy.deepcopy(input_dict['ltw'])
    words = input_dict['words'].encode("utf-8")
    words = words.split("\n")

    ltw = []
    import re

    for tweet in tweets:
        for word in words:
            pattern = re.compile(word, re.IGNORECASE)
            tweet['text']=pattern.sub('',tweet['text'])
        ltw.append(tweet)

    output_dict = {}

    output_dict['ltw']=ltw
    return output_dict

def streaming_simulate_stream_from_text_file(input_dict,widget,stream=None):
    import datetime
    csvfile = safeOpen(input_dict['file'])
    tweet_data = csvfile.read()
    tweet_data = tweet_data.strip()
    tweets = tweet_data.split("\n")
    ltw = []
    i=1
    for tw in tweets:
        tweet = {}
        tweet['id']=i
        tweet['created_at']=datetime.datetime.now()
        tweet['text']=tw
        tweet['user']="dragi"
        tweet['lang']="bg"
        i=i+1
        ltw.append(tweet)
    output_dict = {}
    output_dict['ltw']=ltw
    return output_dict

def streaming_simulate_stream_from_csv(input_dict,widget,stream=None):
    from streams.models import StreamWidgetData
    import datetime
    import csv
    csvfile = safeOpen(input_dict['csv'])
    csvreader = csv.reader(csvfile,delimiter=";",quotechar='"')
    rows = []
    ltw = []
    i=0
    counter = 0
    started = False
    last_id = "not-started-yet"
    if not stream is None:
        try:
            swd = StreamWidgetData.objects.get(stream=stream,widget=widget)
            last_id = swd.value
        except:
            started = True
    else:
        started = True
    for row in csvreader:
        rows.append(row)
        if i!=0:
            rows[i][1] = datetime.datetime.strptime(rows[i][1],"%m/%d/%Y %I:%M:%S %p")
            tweet = {}
            tweet['id'] = rows[i][0]
            tweet['created_at'] = rows[i][1]
            tweet['text'] = rows[i][3].encode('utf-8')
            tweet['user'] = rows[i][5].encode('utf-8')
            tweet['lang'] = rows[i][11]
            if started:
                counter = counter + 1
                ltw.append(tweet)
            if counter == 50 and started:
                started = False
                if not stream is None:
                    try:
                        swd = StreamWidgetData.objects.get(stream=stream,widget=widget)
                        swd.value = tweet['id']
                        swd.save()
                    except:
                        swd = StreamWidgetData()
                        swd.stream = stream
                        swd.widget = widget
                        data = tweet['id']
                        swd.value = data
                        swd.save()
            if tweet['id']==last_id:
                started = True
        i = i + 1
    if counter < 51 and not stream is None and started == True:
        try:
            swd = StreamWidgetData.objects.get(stream=stream,widget=widget)
            swd.value = "done"
            swd.save()
        except:
            swd = StreamWidgetData()
            swd.stream = stream
            swd.widget = widget
            data = "done"
            swd.value = data
            swd.save()
    output_dict = {}
    #print ltw
    #print len(ltw)
    output_dict['ltw']=ltw
    return output_dict

138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
def streaming_split_pos_neg(input_dict):

    tweets = input_dict['ltw']

    positive_tweets = []
    negative_tweets = []

    for tweet in tweets:
        try:
            if tweet['sentiment']=="Positive":
                positive_tweets.append(tweet)
            if tweet['sentiment']=="Negative" and tweet['reliability']!=-1.0:
                negative_tweets.append(tweet)
        except:
            pass

    output_dict = {}

    output_dict['ptw']=positive_tweets
    output_dict['ntw']=negative_tweets
    return output_dict

def streaming_filter_tweets_by_language(input_dict):

    language = input_dict['lang']

    tweets = input_dict['ltw']

    new_tweets = []

    for tweet in tweets:
        if tweet['lang']==language:
            new_tweets.append(tweet)

    output_dict = {}
    output_dict['ltw']=new_tweets
    return output_dict

Janez K's avatar
Janez K committed
176
def streaming_display_tweets(input_dict,widget,stream=None):
Janez K's avatar
Janez K committed
177
178
179
180
181
182
    from streams.models import StreamWidgetData
    if stream is None:
        return {}
    else:
        try:
            swd = StreamWidgetData.objects.get(stream=stream,widget=widget)
Janez K's avatar
Janez K committed
183
184
185
            swd.value = input_dict['ltw']
            swd.save()
        except:
Janez K's avatar
Janez K committed
186
187
188
            swd = StreamWidgetData()
            swd.stream = stream
            swd.widget = widget
Janez K's avatar
Janez K committed
189
            data = input_dict['ltw']
Janez K's avatar
Janez K committed
190
191
            swd.value = data
            swd.save()
Janez K's avatar
Janez K committed
192
193
        return {}

Janez K's avatar
Janez K committed
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
def streaming_triplet_graph(input_dict,widget,stream=None):
    from streams.models import StreamWidgetData
    if stream is None:
        return {}
    else:
        try:
            swd = StreamWidgetData.objects.get(stream=stream,widget=widget)
            swd.value = input_dict['triplets']
            swd.save()
        except:
            swd = StreamWidgetData()
            swd.stream = stream
            swd.widget = widget
            data = input_dict['triplets']
            swd.value = data
            swd.save()
        return {}

Janez K's avatar
Janez K committed
212
213
214
215
216
217
218
219
220
def streaming_collect_and_display_tweets(input_dict,widget,stream=None):
    from streams.models import StreamWidgetData
    if stream is None:
        return {}
    else:
        new_tweets = []
        for tweet in input_dict['ltw']:
            new_tweets.append(StreamWidgetData(stream=stream,widget=widget,value=tweet))
        StreamWidgetData.objects.bulk_create(new_tweets)
Janez K's avatar
Janez K committed
221
222
        return {}

Janez K's avatar
Janez K committed
223
224
225
226
227
def streaming_sentiment_graph(input_dict,widget,stream=None):
    from streams.models import StreamWidgetData
    if stream is None:
        return {}
    else:
Janez K's avatar
Janez K committed
228
229
230
231
        new_tweets = []
        for tweet in input_dict['ltw']:
            new_tweets.append(StreamWidgetData(stream=stream,widget=widget,value=tweet))
        StreamWidgetData.objects.bulk_create(new_tweets)
Janez K's avatar
Janez K committed
232
233
        return {}

234
235
236
def streaming_tweet_sentiment_service(input_dict,widget,stream=None):
    import pickle
    from pysimplesoap.client import SoapClient, SoapFault
Janez K's avatar
Janez K committed
237
    import pysimplesoap
238

Janez K's avatar
Janez K committed
239
    client = SoapClient(location = "http://95.87.154.167:8088/",action = 'http://95.87.154.167:8088/',namespace = "http://example.com/tweetsentiment.wsdl",soap_ns='soap',trace = False,ns = False)
Janez K's avatar
Janez K committed
240
    pysimplesoap.client.TIMEOUT = 60
241
242
243
244
245
246
247
248
249
250
251
252

    list_of_tweets = input_dict['ltw']

    new_list_of_tweets = []

    for tweet in list_of_tweets:
        new_list_of_tweets.append({'id':tweet['id'],'text':tweet['text'],'language':tweet['lang']})

    pickled_list_of_tweets = pickle.dumps(new_list_of_tweets)

    response = client.TweetSentimentService(tweets=pickled_list_of_tweets)

Janez K's avatar
Janez K committed
253
    new_ltw = pickle.loads(str(response.TweetSentimentResult))
254
255
256
257
258
259
260
261

    i=0
    for new_tweet in new_ltw:
        list_of_tweets[i]['sentiment']=new_tweet['sentiment']
        list_of_tweets[i]['lang']=new_tweet['language']
        list_of_tweets[i]['reliability']=new_tweet['reliability']
        i = i + 1

Janez K's avatar
Janez K committed
262
    output_dict = {}
263
264
265

    output_dict['ltw'] = list_of_tweets

Janez K's avatar
Janez K committed
266
267
    return output_dict

Janez K's avatar
Janez K committed
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
def streaming_triplet_porter_stemmer(input_dict,widget,stream=None):
    import nltk
    triplets = input_dict['triplets']
    new_triplets = []
    porter_stemmer = nltk.stem.PorterStemmer()
    for triplet in triplets:
        new_triplet = []
        for word in triplet:
            try:
                new_triplet.append(porter_stemmer.stem(word))
            except:
                new_triplet.append(word)
        new_triplets.append(new_triplet)
    output_dict = {}
    output_dict['triplets']=new_triplets
    return output_dict

def streaming_triplet_wordnet_lemmatizer(input_dict,widget,stream=None):
    import nltk
    triplets = input_dict['triplets']
    new_triplets = []
    lemmatizer = nltk.stem.WordNetLemmatizer()
    for triplet in triplets:
        new_triplet = []
        for word in triplet:
            try:
                new_triplet.append(lemmatizer.lemmatize(word))
            except:
                new_triplet.append(word)
        new_triplets.append(new_triplet)
    output_dict = {}
    output_dict['triplets']=new_triplets
Janez K's avatar
Janez K committed
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
    return output_dict


def streaming_triplet_extraction_2(input_dict, widget, stream=None):
    output_dict = {}
    text = input_dict['text']
    t = TripletClient()
    triplets = t.reverb(text)
    new_triplets = []
    for extraction in triplets['extractions']:
        triplet = [extraction['arg1'].lower(), extraction['relNorm'].lower(), extraction['arg2'].lower()]
        new_triplets.append(triplet)
    output_dict['triplets'] = new_triplets
    return output_dict

Janez K's avatar
Janez K committed
315

Janez K's avatar
Janez K committed
316
317
318
def streaming_triplet_extraction(input_dict,widget,stream=None):
    from pysimplesoap.client import SoapClient, SoapFault
    import pysimplesoap
Janez K's avatar
Janez K committed
319
    import re
Janez K's avatar
Janez K committed
320
    client = SoapClient(location = "http://95.87.154.23:8008/",action = 'http://95.87.154.23:8008/',namespace = "http://example.com/tweetsentiment.wsdl",soap_ns='soap',trace = False,ns = False)
Janez K's avatar
Janez K committed
321
322
323
324
    pysimplesoap.client.TIMEOUT = 36000

    text = input_dict['text']
    response = client.TripletExtraction(text=text)
Janez K's avatar
Janez K committed
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341

    triplets = str(response.TripletExtractionResult)
    triplets = triplets.split("\n")

    new_triplets = []

    for triplet in triplets:
        t = re.sub(r'\([^)]*\)', "",triplet)
        triplet = []
        words = t.strip().split(" ")
        for word in words:
            if len(word)>0:
                triplet.append(word.lower())
        if len(triplet)>0:
            new_triplets.append(triplet)


Janez K's avatar
Janez K committed
342
    output_dict = {}
Janez K's avatar
Janez K committed
343
    output_dict['triplets'] = new_triplets
Janez K's avatar
Janez K committed
344
345
346

    return output_dict

Janez K's avatar
Janez K committed
347
348
def streaming_summarize_url(input_dict,widget,stream=None):
    import pyteaser
Janez K's avatar
revert    
Janez K committed
349
350
351
    summaries = pyteaser.SummarizeUrl(input_dict['url'])
    output_dict = {}
    output_dict['summary']=" ".join(summaries)
Janez K's avatar
Janez K committed
352
    return output_dict
Janez K's avatar
Janez K committed
353

Janez K's avatar
Janez K committed
354
355
356
357
358
359
360
def streaming_get_article_text(input_dict,widget,stream=None):
    import pyteaser
    text = pyteaser.grab_link(input_dict['url'])
    output_dict = {}
    output_dict['text']=text.cleaned_text
    return output_dict

Janez K's avatar
Janez K committed
361
362
363
364
def streaming_active_sentiment_analysis(input_dict,widget,stream=None):
    import pickle
    from pysimplesoap.client import SoapClient, SoapFault
    import pysimplesoap
Janez K's avatar
Janez K committed
365
    client = SoapClient(location = "http://95.87.154.167:8088/",action = 'http://95.87.154.167:8088/',namespace = "http://example.com/tweetsentiment.wsdl",soap_ns='soap',trace = False,ns = False)
Janez K's avatar
Janez K committed
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
    pysimplesoap.client.TIMEOUT = 600

    list_of_tweets = input_dict['ltw']

    new_list_of_tweets = []

    for tweet in list_of_tweets:
        new_list_of_tweets.append(tweet['text'])

    workflow_id = widget.id

    service_input = pickle.dumps((str(workflow_id),new_list_of_tweets))
    #print service_input
    response = client.ActiveClassifyMultiple(workflowtweets=service_input)

    i=0
    new_ltw = pickle.loads(str(response.ActiveClassifyMultipleResult))

Janez K's avatar
Janez K committed
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
    for new_tweet in new_ltw:
        if new_tweet[0]=="True":
            list_of_tweets[i]['sentiment']="Positive"
        elif new_tweet[0]=="False":
            list_of_tweets[i]['sentiment']="Negative"
        list_of_tweets[i]['reliability']=new_tweet[1]
        i=i+1

    output_dict = {}

    output_dict['ltw'] = list_of_tweets

    return output_dict

def streaming_active_sentiment_analysis2(input_dict,widget,stream=None):
    import pickle
    from pysimplesoap.client import SoapClient, SoapFault
    import pysimplesoap
    client = SoapClient(location = "http://95.87.154.167:8098/",action = 'http://95.87.154.167:8098/',namespace = "http://example.com/tweetsentiment.wsdl",soap_ns='soap',trace = False,ns = False)
    pysimplesoap.client.TIMEOUT = 600

    list_of_tweets = input_dict['ltw']

    new_list_of_tweets = []

    for tweet in list_of_tweets:
        new_list_of_tweets.append(tweet['text'])

    workflow_id = widget.id

    service_input = pickle.dumps((str(workflow_id),input_dict['b_size'],input_dict['q_strategy_closest'],input_dict['q_strategy_random'],new_list_of_tweets))
    #print service_input
    response = client.ActiveClassifyMultiple(workflowtweets=service_input)

    i=0
    new_ltw = pickle.loads(str(response.ActiveClassifyMultipleResult))

Janez K's avatar
Janez K committed
421
422
423
424
425
426
427
428
429
430
431
432
433
434
    for new_tweet in new_ltw:
        if new_tweet[0]=="True":
            list_of_tweets[i]['sentiment']="Positive"
        elif new_tweet[0]=="False":
            list_of_tweets[i]['sentiment']="Negative"
        list_of_tweets[i]['reliability']=new_tweet[1]
        i=i+1

    output_dict = {}

    output_dict['ltw'] = list_of_tweets

    return output_dict

Janez K's avatar
Janez K committed
435
436
437
438
439
def streaming_sentiment_analysis(input_dict,widget,stream=None):
    import pickle
    from pysimplesoap.client import SoapClient, SoapFault
    import pysimplesoap

Janez K's avatar
Janez K committed
440
    client = SoapClient(location = "http://95.87.154.167:8088/",action = 'http://95.87.154.167:8088/',namespace = "http://example.com/tweetsentiment.wsdl",soap_ns='soap',trace = False,ns = False)
Janez K's avatar
Janez K committed
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
    pysimplesoap.client.TIMEOUT = 60

    try:
        input_dict['lang']
    except:
        input_dict['lang']="en"

    #list_of_tweets = input_dict['ltw']

    new_list_of_tweets = []

    new_list_of_tweets.append({'id':0,'text':input_dict['text'],'language':input_dict['lang']})

    #for tweet in list_of_tweets:
    #    new_list_of_tweets.append({'id':tweet['id'],'text':tweet['text'],'language':tweet['lang']})

    pickled_list_of_tweets = pickle.dumps(new_list_of_tweets)

    response = client.TweetSentimentService(tweets=pickled_list_of_tweets)

    new_ltw = pickle.loads(str(response.TweetSentimentResult))

    output_dict = {}

    #i=0
    for new_tweet in new_ltw:
        output_dict['sentiment']=new_tweet['sentiment']
        output_dict['language']=new_tweet['language']
        output_dict['reliability']=new_tweet['reliability']
        #list_of_tweets[i]['sentiment']=new_tweet['sentiment']
        #list_of_tweets[i]['lang']=new_tweet['language']
        #list_of_tweets[i]['reliability']=new_tweet['reliability']
        #i = i + 1

    #output_dict = {}

    #output_dict['ltw'] = list_of_tweets

    return output_dict

481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
def streaming_twitter(input_dict,widget,stream=None):
    import tweepy
    from streams.models import StreamWidgetData
    from streams.models import HaltStream

    if input_dict['cfauth']=="true":
        consumer_key="zmK41mqxU3ZNJTFQpYwTdg"
        consumer_secret="9StnKNAe20ebDOREQjsVjAjBEiz5R9feZJTGUYWqLo"
        access_token="45210078-VydgdJMwhWYjZRvlNbrKj6jfqicUIsdMnRbnaPElL"
        access_token_secret="uLvIN3MMxFSxdK4M8P5RYojjUkbc2reqNydYtpT7Ks"
    else:
        consumer_key = input_dict['ck']
        consumer_secret = input_dict['cs']
        access_token = input_dict['at']
        access_token_secret = input_dict['as']

    auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
    auth.set_access_token(access_token, access_token_secret)

    api = tweepy.API(auth)

    query = input_dict['query']

    rate_limit_status = api.rate_limit_status()

    if rate_limit_status['resources']['search']['/search/tweets']['remaining']>0:

        if stream is None:
            try:
                ltw = api.new_search(q=input_dict['query'],geocode=input_dict['geocode'],count=100)
            except Exception as e:
                raise HaltStream("The Twitter API returned an error: "+str(e))
        else:
            try:
                swd = StreamWidgetData.objects.get(stream=stream,widget=widget)
                data = swd.value
            except Exception as e:
                swd = StreamWidgetData()
                swd.stream = stream
                swd.widget = widget
                data = {}
                swd.value = data
                swd.save()
            if data.has_key(query):
                since_id = data[query]
                try:
                    ltw = api.new_search(q=input_dict['query'],geocode=input_dict['geocode'],count=100,since_id=since_id)
                except Exception as e:
                    raise HaltStream("The Twitter API returned an error: "+str(e))
            else:
                try:
                    ltw = api.new_search(q=input_dict['query'],geocode=input_dict['geocode'],count=100)
                except Exception as e:
                    raise HaltStream("The Twitter API returned an error: "+str(e))
            if len(ltw)>0:
                data[query]=ltw[0].id
                swd.value = data
                swd.save()
    else:
        import datetime
        import time
        current_time = int(time.mktime(datetime.datetime.now().timetuple()))
        remaining = rate_limit_status['resources']['search']['/search/tweets']['reset']-current_time
        if input_dict['cfauth']=="true":
            raise HaltStream("The twitter API limit has been reached. Try again in "+str(remaining)+" seconds. Try using your own credentials.")
        else:
            raise HaltStream("The twitter API limit has been reached. Try again in "+str(remaining)+" seconds.")

    output_dict = {}

    tweets = []

    for tw in ltw:
        tweet = {}
        tweet['id'] = tw.id
        tweet['created_at'] = tw.created_at
        tweet['text'] = unicode(tw.text).encode("utf-8")
Janez K's avatar
Janez K committed
558
559
560
561
        try:
            tweet['user'] = tw.user['screen_name']
        except:
            tweet['user'] = ""
562
563
564
565
566
567
568
569
570
        tweet['lang'] = tw.lang
        tweets.append(tweet)

    if len(tweets)>0 or stream is None:
        output_dict['ltw']=tweets
        return output_dict
    else:
        raise HaltStream("No new results, halting stream.")

Janez K's avatar
Janez K committed
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
def streaming_rss_reader(input_dict,widget,stream=None):
    import feedparser
    from streams.models import StreamWidgetData
    feed = feedparser.parse(input_dict['url'])
    output_dict = {}
    if stream is None:
        output_dict['url'] = feed['items'][0]['link']
    else:
        try:
            swd = StreamWidgetData.objects.get(stream=stream,widget=widget)
            data = swd.value
        except:
            swd = StreamWidgetData()
            swd.stream = stream
            swd.widget = widget
            data = []
            swd.value = data
            swd.save()
        feed_length = len(feed['items'])
        feed['items'].reverse()
        for item in feed['items']:
Janez K's avatar
Janez K committed
592
593
            if item['link'] not in data:
                data.append(item['link'])
Janez K's avatar
Janez K committed
594
595
596
597
598
                swd.value = data
                swd.save()
                output_dict['url'] = item['link']
                break
        else:
Janez K's avatar
Janez K committed
599
600
601
            from streams.models import HaltStream
            raise HaltStream("Halting stream.")
    return output_dict
Janez K's avatar
Janez K committed
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629


def streaming_sliding_window(input_dict,widget,stream=None):
    from streams.models import StreamWidgetData
    output_dict = {}
    if stream is None:
        output_dict['list']=input_dict['list'][:int(input_dict['size'])]
    else:
        try:
            swd = StreamWidgetData.objects.get(stream=stream,widget=widget)
            data = swd.value
        except:
            swd = StreamWidgetData()
            swd.stream = stream
            swd.widget = widget
            data = []
            swd.value = data
            swd.save()
        size = int(input_dict['size'])
        if len(input_dict['list'])>=size:
            output_dict['list']=input_dict['list'][:size]
            swd.value=output_dict['list']
            swd.save()
        else:
            current_window = input_dict['list'] + swd.value
            swd.value = current_window[:size]
            output_dict['list']=current_window[:size]
            swd.save()
Janez K's avatar
Janez K committed
630
    return output_dict