script/utils/export_twitter_alchemy.py
changeset 464 b9243ade95e2
parent 411 0471e6eb8a1b
child 467 266c861f293f
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/script/utils/export_twitter_alchemy.py	Sat Jan 07 16:12:44 2012 +0100
@@ -0,0 +1,362 @@
+#!/usr/bin/env python
+# coding=utf-8
+
+from lxml import etree
+from iri_tweet.models import setup_database
+from optparse import OptionParser #@UnresolvedImport
+from sqlalchemy import Table, Column, BigInteger
+from iri_tweet.utils import (set_logging_options, set_logging, get_filter_query, 
+    get_logger)
+import anyjson
+import datetime
+import httplib2
+import os.path
+import re
+import sys
+import time
+import uuid #@UnresolvedImport
+from dateutil.parser import parse as parse_date
+
+#class TweetExclude(object):
+#    def __init__(self, id):
+#        self.id = id
+#        
+#    def __repr__(self):
+#        return "<TweetExclude(id=%d)>" % (self.id)
+
+
+def parse_polemics(tw, extended_mode):
+    """
+    parse polemics in text and return a list of polemic code. None if not polemic found
+    """
+    polemics = {} 
+    for m in re.finditer("(\+\+|\-\-|\?\?|\=\=)",tw.text):
+        pol_link = {
+            '++' : u'OK',
+            '--' : u'KO',
+            '??' : u'Q',
+            '==' : u'REF'}[m.group(1)]
+        polemics[pol_link] = pol_link
+    
+    if extended_mode:
+        if "?" in tw.text:
+            polemics["Q"] = "Q"
+        
+        for entity in tw.entity_list:
+            if entity.type == "entity_url":
+                polemics["REF"] = "REF" 
+    
+    if len(polemics) > 0:
+        return polemics.keys()
+    else:
+        return None
+
+def get_options():
+    parser = OptionParser()
+    parser.add_option("-f", "--file", dest="filename",
+                      help="write export to file", metavar="FILE", default="project.ldt")
+    parser.add_option("-d", "--database", dest="database",
+                      help="Input database", metavar="DATABASE")
+    parser.add_option("-s", "--start-date", dest="start_date",
+                      help="start date", metavar="START_DATE", default=None)
+    parser.add_option("-e", "--end-date", dest="end_date",
+                      help="end date", metavar="END_DATE", default=None)
+    parser.add_option("-I", "--content-file", dest="content_file",
+                      help="Content file", metavar="CONTENT_FILE")
+    parser.add_option("-c", "--content", dest="content",
+                      help="Content url", metavar="CONTENT")
+    parser.add_option("-V", "--video-url", dest="video",
+                      help="video url", metavar="VIDEO")
+    parser.add_option("-i", "--content-id", dest="content_id",
+                      help="Content id", metavar="CONTENT_ID")
+    parser.add_option("-x", "--exclude", dest="exclude",
+                      help="file containing the id to exclude", metavar="EXCLUDE")
+    parser.add_option("-C", "--color", dest="color",
+                      help="Color code", metavar="COLOR", default="16763904")
+    parser.add_option("-H", "--hashtag", dest="hashtag",
+                      help="Hashtag", metavar="HASHTAG", default=[], action="append")                      
+    parser.add_option("-D", "--duration", dest="duration", type="int",
+                      help="Duration", metavar="DURATION", default=None)
+    parser.add_option("-n", "--name", dest="name",
+                      help="Cutting name", metavar="NAME", default=u"Tweets")
+    parser.add_option("-R", "--replace", dest="replace", action="store_true",
+                      help="Replace tweet ensemble", metavar="REPLACE", default=False)
+    parser.add_option("-m", "--merge", dest="merge", action="store_true",
+                      help="merge tweet ensemble, choose the first ensemble", metavar="MERGE", default=False)
+    parser.add_option("-L", "--list-conf", dest="listconf",
+                      help="list of file to process", metavar="LIST_CONF", default=None)
+    parser.add_option("-E", "--extended", dest="extended_mode", action="store_true",
+                      help="Trigger polemic extended mode", metavar="EXTENDED", default=False)
+    parser.add_option("--user-whitelist", dest="user_whitelist", action="store",
+                      help="A list of user screen name", metavar="USER_WHITELIST",default=None)
+    
+    
+    set_logging_options(parser)
+
+    
+    return parser.parse_args() + (parser,)
+
+
+if __name__ == "__main__" :
+
+    (options, args, parser) = get_options()
+    
+    set_logging(options)
+        
+    get_logger().debug("OPTIONS : " + repr(options)) #@UndefinedVariable
+    
+    if len(sys.argv) == 1 or options.database is None:
+        parser.print_help()
+        sys.exit(1)
+    
+    conn_str = options.database.strip()
+    if not re.match("^\w+://.+", conn_str):
+        conn_str = 'sqlite:///' + conn_str
+
+    engine, metadata, Session = setup_database(conn_str, echo=((options.verbose-options.quiet)>0), create_all = False)        
+    conn = None
+    try :
+        conn = engine.connect()    
+        session = None
+        try :
+            session = Session(bind=conn)         
+            tweet_exclude_table = Table("tweet_exclude", metadata, Column('id', BigInteger, primary_key=True), prefixes=['TEMPORARY'])
+            #mapper(TweetExclude, tweet_exclude_table)
+            metadata.create_all(bind=conn, tables=[tweet_exclude_table])
+            
+            if options.exclude and os.path.exists(options.exclude):
+                with open(options.exclude, 'r+') as f:
+                    tei = tweet_exclude_table.insert()
+                    for line in f:
+                        conn.execute(tei.values(id=long(line.strip())))
+            user_whitelist_file = options.user_whitelist
+            user_whitelist = None
+            
+            if options.listconf:
+                
+                parameters = []
+                confdoc = etree.parse(options.listconf)
+                for node in confdoc.xpath("/twitter_export/file"):
+                    params = {}
+                    for snode in node:
+                        if snode.tag == "path":
+                            params['content_file'] = snode.text
+                        elif snode.tag == "start_date":
+                            params['start_date'] = snode.text
+                        elif snode.tag == "end_date":
+                            params['end_date'] = snode.text
+                        elif snode.tag == "duration":
+                            params['duration'] = int(snode.text)
+                        elif snode.tag == "hashtags":
+                            params['hashtags'] = [snode.text]
+                    if options.hashtag or 'hashtags' not in params :
+                        params['hashtags'] = options.hashtag
+                    parameters.append(params)
+            else:                        
+                parameters = [{
+                    'start_date': options.start_date,
+                    'end_date' : options.end_date,
+                    'duration' : options.duration,
+                    'content_file' : options.content_file,
+                    'hashtags' : options.hashtag
+                }]
+            
+            for params in parameters:
+                
+                get_logger().debug("PARAMETERS " + repr(params)) #@UndefinedVariable
+                
+                start_date_str = params.get("start_date",None)
+                end_date_str = params.get("end_date", None)
+                duration = params.get("duration", None)
+                content_file = params.get("content_file", None)
+                hashtags = params.get('hashtags', [])
+                  
+                if user_whitelist_file:
+                    with open(user_whitelist_file, 'r+') as f:
+                        user_whitelist = list(set([s.strip() for s in f]))
+                
+                start_date = None
+                ts = None
+                if start_date_str:
+                    start_date = parse_date(start_date_str) 
+                    ts = time.mktime(start_date.timetuple())
+            
+                end_date = None
+                if end_date_str:
+                    end_date = parse_date(end_date_str)
+                elif start_date and duration:
+                    end_date = start_date + datetime.timedelta(seconds=duration)
+                
+                query = get_filter_query(session, start_date, end_date, hashtags, tweet_exclude_table, user_whitelist)
+                    
+                query_res = query.all()
+                                 
+                root = None
+                ensemble_parent = None
+                
+                #to do : analyse situation ldt or iri ? filename set or not ?
+                
+                if content_file and content_file.find("http") == 0:
+                    
+                    get_logger().debug("url : " + content_file) #@UndefinedVariable
+                    
+                    h = httplib2.Http()
+                    resp, content = h.request(content_file)
+                    
+                    get_logger().debug("url response " + repr(resp) + " content " + repr(content)) #@UndefinedVariable
+                    
+                    project = anyjson.deserialize(content)
+                    root = etree.fromstring(project["ldt"])
+                
+                elif content_file and os.path.exists(content_file):
+
+                    doc = etree.parse(content_file)
+                    root = doc.getroot()
+                    
+                
+                if root is None:
+                
+                    root = etree.Element(u"iri")
+                        
+                    project = etree.SubElement(root, u"project", {u"abstract":u"Polemics Tweets",u"title":u"Polemic Tweets", u"user":u"IRI Web", u"id":unicode(uuid.uuid4())})
+                
+                    medias = etree.SubElement(root, u"medias")
+                    media = etree.SubElement(medias, u"media", {u"pict":u"", u"src":unicode(options.content), u"video":unicode(options.video), u"id":unicode(options.content_id), u"extra":u""})
+                    
+                    annotations = etree.SubElement(root, u"annotations")    
+                    content = etree.SubElement(annotations, u"content", {u"id":unicode(options.content_id)})
+                    ensemble_parent = content
+                    
+                
+                if ensemble_parent is None:
+                    file_type = None
+                    for node in root:
+                        if node.tag == "project":
+                            file_type = "ldt"
+                            break
+                        elif node.tag == "head":
+                            file_type = "iri"
+                            break
+                    
+                    if file_type == "ldt":
+                        media_nodes = root.xpath("//media")
+                        if len(media_nodes) > 0:
+                            media = media_nodes[0]
+                        annotations_node = root.find(u"annotations")
+                        if annotations_node is None:
+                            annotations_node = etree.SubElement(root, u"annotations")
+                        content_node = annotations_node.find(u"content")
+                        if content_node is None:
+                            content_node = etree.SubElement(annotations_node,u"content", id=media.get(u"id"))
+                        ensemble_parent = content_node
+                    elif file_type == "iri":
+                        body_node = root.find(u"body")
+                        if body_node is None:
+                            body_node = etree.SubElement(root, u"body")
+                        ensembles_node = body_node.find(u"ensembles")
+                        if ensembles_node is None:
+                            ensembles_node = etree.SubElement(body_node, u"ensembles")
+                        ensemble_parent = ensembles_node
+                    
+                    
+                if ensemble_parent is None:
+                    get_logger().error("Can not process file") #@UndefinedVariable
+                    sys.exit()
+            
+                if options.replace:
+                    for ens in ensemble_parent.iterchildren(tag=u"ensemble"):
+                        if ens.get("id","").startswith("tweet_"):
+                            ensemble_parent.remove(ens)
+                
+                ensemble = None
+                elements = None
+                
+                if options.merge:
+                    ensemble = ensemble_parent.find(u"ensemble")
+                    if ensemble is not None:
+                        elements = ensemble.find(u".//elements")                
+                    
+                if ensemble is None or elements is None:
+                    ensemble = etree.SubElement(ensemble_parent, u"ensemble", {u"id":u"tweet_" + unicode(uuid.uuid4()), u"title":u"Ensemble Twitter", u"author":u"IRI Web", u"abstract":u"Ensemble Twitter"})
+                    decoupage = etree.SubElement(ensemble, u"decoupage", {u"id": unicode(uuid.uuid4()), u"author": u"IRI Web"})
+                
+                    etree.SubElement(decoupage, u"title").text = unicode(options.name)
+                    etree.SubElement(decoupage, u"abstract").text = unicode(options.name)
+                
+                    elements = etree.SubElement(decoupage, u"elements")
+
+                
+                for tw in query_res:
+                    tweet_ts_dt = tw.created_at
+                    tweet_ts = int(time.mktime(tweet_ts_dt.timetuple()))
+                    if ts is None:
+                        ts = tweet_ts
+                    tweet_ts_rel = (tweet_ts-ts) * 1000
+                    username = None
+                    profile_url = ""
+                    if tw.user is not None:
+                        username = tw.user.name                    
+                        profile_url = tw.user.profile_image_url if tw.user.profile_image_url is not None else ""
+                    if not username:
+                        username = "anon."
+                    
+                    element = etree.SubElement(elements, u"element" , {u"id":unicode(uuid.uuid4())+u"-"+unicode(tw.id), u"color":unicode(options.color), u"author":unicode(username), u"date":unicode(tweet_ts_dt.strftime("%Y/%m/%d")), u"begin": unicode(tweet_ts_rel), u"dur":u"0", u"src":unicode(profile_url)})
+                    etree.SubElement(element, u"title").text = unicode(username) + u": " + unicode(tw.text)
+                    etree.SubElement(element, u"abstract").text = unicode(tw.text)
+            
+                    tags_node = etree.SubElement(element, u"tags")
+                    
+                    for entity in tw.entity_list:
+                        if entity.type == u'entity_hashtag': 
+                            etree.SubElement(tags_node,u"tag").text = entity.hashtag.text
+                            
+                    meta_element = etree.SubElement(element, u'meta')
+                    
+                    polemics_list = parse_polemics(tw, options.extended_mode)
+                    if polemics_list:
+                        polemics_element = etree.Element(u'polemics')
+                        for pol in polemics_list:
+                            etree.SubElement(polemics_element, u'polemic').text = pol
+                        meta_element.append(polemics_element)
+
+                    etree.SubElement(meta_element, u"source", attrib={"url":u"http://dev.twitter.com", "mimetype":u"application/json"}).text = etree.CDATA(unicode(tw.tweet_source.original_json))
+                    
+                # sort by tc in
+                if options.merge :
+                    # remove all elements and put them in a array
+                    # sort them with tc
+                    #put them back
+                    elements[:] = sorted(elements,key=lambda n: int(n.get('begin')))
+                    
+                    
+                
+                
+                output_data = etree.tostring(root, encoding="utf-8", method="xml", pretty_print=True, xml_declaration=True)  
+                
+                if content_file and content_file.find("http") == 0:
+                    
+                    project["ldt"] = output_data
+                    body = anyjson.serialize(project)
+                    get_logger().debug("write http " + content_file) #@UndefinedVariable
+                    get_logger().debug("write http " + repr(body)) #@UndefinedVariable
+                    h = httplib2.Http()
+                    resp, content = h.request(content_file, "PUT", headers={'content-type':'application/json'}, body=body)
+                    get_logger().debug("write http " + repr(resp) + " content " + content) #@UndefinedVariable
+                else:
+                    if content_file and os.path.exists(content_file):
+                        dest_file_name = content_file 
+                    else:
+                        dest_file_name = options.filename
+            
+                    get_logger().debug("WRITE : " + dest_file_name) #@UndefinedVariable
+                    output = open(dest_file_name, "w")
+                    output.write(output_data)
+                    output.flush()
+                    output.close()
+                
+        finally:
+            if session:
+                session.close()
+    finally:
+        if conn:
+            conn.close()