def voteArticle(upOrDown, item, topic="General"):
"""
voteArticle function, takes arguments upOrDown vote, topic to vote,
and the text to vote for
"""
text = purify.cleanText(item.description)
title = purify.cleanText(item.title)
c = Classifier(classifierDir, [topic,"not"+topic])
c2 = Classifier(classifierDir, [topic+"Title", "not"+topic+"Title"])
try:
if upOrDown is "up":
c.learn(topic, text)
c2.learn(topic+"Title", title)
elif upOrDown is "down":
c.learn("not"+topic, text)
c2.learn("not"+topic+"Title", title)
if topic is not "General":
#always add the upvote to the General interest category as well.
d = Classifier(classifierDir, ["General","notGeneral"])
d2 = Classifier(classifierDir, ["GeneralTitle","notGeneralTitle"])
if upOrDown is "up":
d.learn("General", text)
d2.learn("GeneralTitle", title)
elif upOrDown is "down":
d.learn("notGeneral", text)
d2.learn("notGeneralTitle", title)
except UnicodeEncodeError:
print "Article content contains invalid characters!"
def calcScore(self, article):
"""
when an article is passed this function calculates its overall score, by
adding content score + feed score + feed update frequency(should be
caluculated by taking the deviation from the mean update friquency.)
"""
# Get the Article title and content in plain text.
text = purify.cleanText(article.description)
titleText = purify.cleanText(article.title)
#Calculate the Score for the texual content of the article
(textTopic, textScore) = classifier.classifyArticleText(self.topic.title, text)
textTopic = textTopic.replace("_", " ")
# print "text : %s, %s" % (textTopic, textScore)
#Calculate the Score for the title of the article.
(titleTopic,titleScore) = classifier.classifyArticleTitle(self.topic.title, titleText)
titleTopic = titleTopic.replace("_", " ")
# print "Title : %s, %s" % (titleTopic,titleScore)
#Now set the textual scores to minus values if the article "notTopic"
if textTopic ==self.topic.title:
textScore = textScore * 10000
else:
textScore = textScore * (-100)
if titleTopic ==self.topic.title+"Title":
titleScore = titleScore * 10000
else:
titleScore = titleScore * (-100)
#Get the Score for the feed from the db
scoreFeed = ScoreFeed.query.filter_by(feed = article.feed, topic = self.topic).first()
feedScore = scoreFeed.score * 100
#updateFrequencyScore - score based on the feeds update frequncy.
#less frequently updated content would get fairly better scores.
# Set weights to be given for the calculated individual scores.
#TODO: Give an option for the user to set the weights of these scores from GUI.
textScoreWeight = TEXT_SCORE_WEIGHT
titleScoreWeight = TITLE_SCORE_WEIGHT
feedScoreWeight = FEED_SCORE_WEIGHT
# updateFrequencyWeight = UPDATE_FREQUENCY_WEIGHT
finalScore = ( ( textScoreWeight * textScore ) +
( titleScoreWeight * titleScore ) +
( feedScoreWeight * feedScore ) )
# ( updateFrequencyWeight * updateFrequencyScore ) )
return finalScore
def calcScore(self, article):
"""
when an article is passed this function calculates its overall score, by
adding content score + feed score + feed update frequency(should be
caluculated by taking the deviation from the mean update friquency.)
"""
# Get the Article title and content in plain text.
text = purify.cleanText(article.description)
titleText = purify.cleanText(article.title)
#Calculate the Score for the texual content of the article
(textTopic, textScore) = classifier.classifyArticleText(self.topic.title, text)
# print "text : %s, %s" % (textTopic, textScore)
#Calculate the Score for the title of the article.
(titleTopic,titleScore) = classifier.classifyArticleTitle(self.topic.title, titleText)
# print "Title : %s, %s" % (titleTopic,titleScore)
#Now set the textual scores to minus values if the article "notTopic"
if textTopic ==self.topic.title:
textScore = textScore * 10000
else:
textScore = textScore * (-100)
if titleTopic ==self.topic.title+"Title":
titleScore = titleScore * 10000
else:
titleScore = titleScore * (-100)
#Get the Score for the feed from the db
feedScore = article.feed.numVotes * 200
#updateFrequencyScore - score based on the feeds update frequncy.
#less frequently updated content would get fairly better scores.
# Set weights to be given for the calculated individual scores.
#TODO: Give an option for the user to set the weights of these scores from GUI.
textScoreWeight = 0.55
titleScoreWeight = 0.35
feedScoreWeight = 0.1
# updateFrequencyWeight = 0.1
finalScore = ( ( textScoreWeight * textScore ) +
( titleScoreWeight * titleScore ) +
( feedScoreWeight * feedScore ) )
# ( updateFrequencyWeight * updateFrequencyScore ) )
return finalScore
def voteArticle(upOrDown, item, topic):
"""
voteArticle function, takes arguments upOrDown vote, topic to vote,
and the text to vote for
"""
# vote for the feed that the article belongs.
_voteFeed(upOrDown, item.feed, topic)
# we'll be working with strings from now on, so get the topic title.
topic = topic.title
# calls to CRM breaks when a topic title contains spaces, replace them with underscores.
topic = topic.replace(" ", "_")
text = purify.cleanText(item.description)
title = purify.cleanText(item.title)
c = Classifier(classifierDir, [topic, "not" + topic])
c2 = Classifier(classifierDir, [topic + "Title", "not" + topic + "Title"])
try:
if upOrDown is "up":
c.learn(topic, text)
c2.learn(topic + "Title", title)
elif upOrDown is "down":
c.learn("not" + topic, text)
c2.learn("not" + topic + "Title", title)
if topic is not "General":
#always add the upvote to the General interest category as well.
d = Classifier(classifierDir, ["General", "notGeneral"])
d2 = Classifier(classifierDir, ["GeneralTitle", "notGeneralTitle"])
if upOrDown is "up":
d.learn("General", text)
d2.learn("GeneralTitle", title)
# We sould only get the Upvote for general. So disabled the DownVote.
# elif upOrDown is "down":
# d.learn("notGeneral", text)
# d2.learn("notGeneralTitle", title)
except UnicodeEncodeError:
print "Article content contains invalid characters!"
def voteArticle(upOrDown, item, topic):
"""
voteArticle function, takes arguments upOrDown vote, topic to vote,
and the text to vote for
"""
# vote for the feed that the article belongs.
_voteFeed(upOrDown, item.feed, topic)
# we'll be working with strings from now on, so get the topic title.
topic = topic.title
# calls to CRM breaks when a topic title contains spaces, replace them with underscores.
topic = topic.replace(" ", "_")
text = purify.cleanText(item.description)
title = purify.cleanText(item.title)
c = Classifier(classifierDir, [topic,"not"+topic])
c2 = Classifier(classifierDir, [topic+"Title", "not"+topic+"Title"])
try:
if upOrDown is "up":
c.learn(topic, text)
c2.learn(topic+"Title", title)
elif upOrDown is "down":
c.learn("not"+topic, text)
c2.learn("not"+topic+"Title", title)
if topic is not "General":
#always add the upvote to the General interest category as well.
d = Classifier(classifierDir, ["General","notGeneral"])
d2 = Classifier(classifierDir, ["GeneralTitle","notGeneralTitle"])
if upOrDown is "up":
d.learn("General", text)
d2.learn("GeneralTitle", title)
# We sould only get the Upvote for general. So disabled the DownVote.
# elif upOrDown is "down":
# d.learn("notGeneral", text)
# d2.learn("notGeneralTitle", title)
except UnicodeEncodeError:
print "Article content contains invalid characters!"
def on_actionRead_activated(self, i = None):
"""
Read article implementataion.Can play or stop the selected article.
"""
if i is None: return
selected = self.currentItem
if self.parent.playerState =='standby':
self.parent.playerState = 'playing'
self.parent.sp.say(purify.cleanText(str(selected.article.title + "....." + selected.article.description)))
#self.sp.say(selected.article.description)
else:
self.parent.sp.stop()
self.parent.playerState='standby'
def on_actionRead_activated(self, i = None):
"""
Read article implementataion.Can play or stop the selected article.
"""
if i is None: return
selected = self.currentItem
if self.parent.playerState =='standby':
self.parent.playerState = 'playing'
self.parent.sp.say(purify.cleanText(str(selected.article.title + "....." + selected.article.description)))
#self.sp.say(selected.article.description)
else:
self.parent.sp.stop()
self.parent.playerState='standby'
def on_actionTranslate_the_article_activated(self, i = None):
"""
Translate the given article in a different language into English language.
"""
if i is None: return
try:
toTranslate = self.currentItem
selectedItem = toTranslate.article
#thisOne = selectedItem.description
v1 = purify.cleanText(str(selectedItem.title))
vx = purify.cleanText(str(selectedItem.description))
v2 = ' '.join(vx.split())
translate = Translator().translate
translatedTitle = translate(v1, lang_to="en")
translatedDescription = translate(v2, lang_to="en")
selectedItem.title = translatedTitle
selectedItem.description = translatedDescription
self.displayArticle()
except:
text = "Error Translating the article"
def addTopic(self):
topic = unicode(self.ui.addTopicLinedit.text())
topic = purify.cleanText(topic)
classifier.addTopic(topic)
self.close()
