def add_bookmark(user, original_language, original_word, translation_language,
translation_word, date, the_context, the_url, the_url_title):
url = Url.find(the_url, the_url_title)
text = Text(the_context, translation_language, url)
if RankedWord.exists(original_word.lower(), original_language):
rank1 = UserWord.find_rank(original_word.lower(), original_language)
w1 = UserWord(original_word, original_language, rank1)
else:
w1 = UserWord(original_word, original_language, None)
if RankedWord.exists(translation_word.lower(), translation_language):
rank2 = UserWord.find_rank(translation_word.lower(),
translation_language)
w2 = UserWord(translation_word, translation_language, rank2)
else:
w2 = UserWord(translation_word, translation_language, None)
zeeguu.db.session.add(url)
zeeguu.db.session.add(text)
zeeguu.db.session.add(w1)
zeeguu.db.session.add(w2)
t1 = Bookmark(w1, w2, user, text, date)
zeeguu.db.session.add(t1)
zeeguu.db.session.commit()
add_probability_to_existing_words_of_user(user, t1, original_language)
def set_know_word_prob():
zeeguu.app.test_request_context().push()
zeeguu.db.session.commit()
enc_probs = EncounterBasedProbability.find_all()
ex_probs = ExerciseBasedProbability.find_all()
for prob in enc_probs:
user = prob.user
word = prob.ranked_word.word
language = prob.ranked_word.language
user_word = None
if UserWord.exists(word, language):
user_word = UserWord.find(word, language)
if ExerciseBasedProbability.exists(user, user_word):
ex_prob = ExerciseBasedProbability.find(user, user_word)
known_word_prob = KnownWordProbability.calculateKnownWordProb(ex_prob.probability, prob.probability)
known_word_probability_obj = KnownWordProbability.find(user, user_word, prob.ranked_word, known_word_prob)
else:
known_word_probability_obj = KnownWordProbability.find(user, None, prob.ranked_word, prob.probability)
zeeguu.db.session.add(known_word_probability_obj)
zeeguu.db.session.commit()
for prob in ex_probs:
user = prob.user
language = prob.user_word.language
word = prob.user_word.word
ranked_word = None
if RankedWord.exists(word, language):
ranked_word = RankedWord.find(word, language)
if not EncounterBasedProbability.exists(user, ranked_word):
if UserWord.exists(word, language):
user_word = UserWord.find(word, language)
known_word_probability_obj = KnownWordProbability(user, user_word, ranked_word, prob.probability)
zeeguu.db.session.add(known_word_probability_obj)
zeeguu.db.session.commit()
print "job3"
def add_bookmark(user, original_language, original_word, translation_language, translation_word, date, the_context, the_url, the_url_title):
url = Url.find (the_url, the_url_title)
text = Text(the_context, translation_language, url)
if RankedWord.exists(original_word.lower(), original_language):
rank1 = UserWord.find_rank(original_word.lower(), original_language)
w1 = UserWord(original_word, original_language,rank1)
else:
w1 = UserWord(original_word, original_language,None)
if RankedWord.exists(translation_word.lower(), translation_language):
rank2 = UserWord.find_rank(translation_word.lower(), translation_language)
w2 = UserWord(translation_word, translation_language,rank2)
else:
w2 = UserWord(translation_word, translation_language,None)
zeeguu.db.session.add(url)
zeeguu.db.session.add(text)
zeeguu.db.session.add(w1)
zeeguu.db.session.add(w2)
t1= Bookmark(w1,w2, user, text, date)
zeeguu.db.session.add(t1)
zeeguu.db.session.commit()
add_probability_to_existing_words_of_user(user,t1,original_language)
def set_default_encounter_based_prob():
zeeguu.app.test_request_context().push()
zeeguu.db.session.commit()
default_probability = 0.5
languages = Language.all()
users = User.find_all()
for user in users:
for lang in languages:
marked_words_of_user_in_text = []
words_of_all_bookmarks_content = []
for bookmark in Bookmark.find_by_specific_user(user):
if bookmark.origin.language == lang:
# bookmark_content_words = re.sub("[^\w]", " ", bookmark.text.content).split()
bookmark_content_words = re.findall(r'(?u)\w+', bookmark.text.content)
words_of_all_bookmarks_content.extend(bookmark_content_words)
marked_words_of_user_in_text.append(bookmark.origin.word)
words_known_from_user= [word for word in words_of_all_bookmarks_content if word not in marked_words_of_user_in_text]
for word_known in words_known_from_user:
if RankedWord.exists(word_known, lang):
rank = RankedWord.find(word_known, lang)
if EncounterBasedProbability.exists(user, rank):
prob = EncounterBasedProbability.find(user,rank, default_probability)
prob.not_looked_up_counter +=1
else:
prob = EncounterBasedProbability.find(user,rank,default_probability)
zeeguu.db.session.add(prob)
zeeguu.db.session.commit()
print 'job2'
def add_ranked_words_to_db(lang_code):
zeeguu.app.test_request_context().push()
zeeguu.db.session.commit()
from_lang = Language.find(lang_code)
initial_line_number = 1
for word in filter_word_list(word_list(lang_code)):
r = RankedWord(word.lower(), from_lang, initial_line_number)
zeeguu.db.session.add(r)
initial_line_number += 1
zeeguu.db.session.commit()
def test_text_difficulty(self):
data = json.dumps(dict(
texts=[dict(content='Der die das warum, wer nicht fragt bleibt bewölkt!', id=1),
dict(content='Dies ist ein Test.', id=2)],
personalized='true'))
RankedWord.cache_ranked_words()
rv = self.api_post('/get_difficulty_for_text/de', data, 'application/json')
difficulties = json.loads(rv.data)['difficulties']
for difficulty in difficulties:
assert 0.0 <= difficulty['score_median'] <= 1.0
assert 0.0 <= difficulty['score_average'] <= 1.0
if difficulty['id'] is 1:
assert difficulty['score_median'] == 1.0
assert round(difficulty['score_average'], 2) == 0.67
elif difficulty['id'] is 2:
assert difficulty['score_median'] == 1.0
assert difficulty['score_average'] == 0.50075
def set_know_word_prob():
zeeguu.app.test_request_context().push()
zeeguu.db.session.commit()
enc_probs = EncounterBasedProbability.find_all()
ex_probs = ExerciseBasedProbability.find_all()
for prob in enc_probs:
user = prob.user
word = prob.ranked_word.word
language = prob.ranked_word.language
user_word = None
if UserWord.exists(word, language):
user_word = UserWord.find(word, language)
if ExerciseBasedProbability.exists(user, user_word):
ex_prob = ExerciseBasedProbability.find(user, user_word)
known_word_prob = KnownWordProbability.calculateKnownWordProb(
ex_prob.probability, prob.probability)
known_word_probability_obj = KnownWordProbability.find(
user, user_word, prob.ranked_word, known_word_prob)
else:
known_word_probability_obj = KnownWordProbability.find(
user, None, prob.ranked_word, prob.probability)
zeeguu.db.session.add(known_word_probability_obj)
zeeguu.db.session.commit()
for prob in ex_probs:
user = prob.user
language = prob.user_word.language
word = prob.user_word.word
ranked_word = None
if RankedWord.exists(word, language):
ranked_word = RankedWord.find(word, language)
if not EncounterBasedProbability.exists(user, ranked_word):
if UserWord.exists(word, language):
user_word = UserWord.find(word, language)
known_word_probability_obj = KnownWordProbability(
user, user_word, ranked_word, prob.probability)
zeeguu.db.session.add(known_word_probability_obj)
zeeguu.db.session.commit()
print 'job3'
def get_known_words(lang_code):
lang_id = Language.find(lang_code)
bookmarks = flask.g.user.all_bookmarks()
known_words = []
filtered_known_words_from_user = []
filtered_known_words_dict_list = []
for bookmark in bookmarks:
if bookmark.check_is_latest_outcome_too_easy():
known_words.append(bookmark.origin.word)
for word_known in known_words:
if RankedWord.exists(word_known, lang_id):
filtered_known_words_from_user.append(word_known)
zeeguu.db.session.commit()
filtered_known_words_from_user = list(set(filtered_known_words_from_user))
for word in filtered_known_words_from_user:
filtered_known_words_dict_list.append({'word': word})
js = json.dumps(filtered_known_words_dict_list)
resp = flask.Response(js, status=200, mimetype='application/json')
return resp
def get_known_words(lang_code):
lang_id = Language.find(lang_code)
bookmarks = flask.g.user.all_bookmarks()
known_words=[]
filtered_known_words_from_user = []
filtered_known_words_dict_list =[]
for bookmark in bookmarks:
if bookmark.check_is_latest_outcome_too_easy():
known_words.append(bookmark.origin.word)
for word_known in known_words:
if RankedWord.exists(word_known, lang_id):
filtered_known_words_from_user.append(word_known)
zeeguu.db.session.commit()
filtered_known_words_from_user = list(set(filtered_known_words_from_user))
for word in filtered_known_words_from_user:
filtered_known_words_dict_list.append( {'word': word} )
js = json.dumps(filtered_known_words_dict_list)
resp = flask.Response(js, status=200, mimetype='application/json')
return resp
def get_difficulty_for_text(lang_code):
"""
URL parameters:
:param lang_code: the language of the text
Json data:
:param texts: json array that contains the texts to calculate the difficulty for. Each text consists of an array
with the text itself as 'content' and an additional 'id' which gets roundtripped unchanged
:param personalized (optional): calculate difficulty score for a specific user? (Enabled by default)
:param rank_boundary (optional): upper boundary for word frequency rank (between 1 and 10'000)
:return difficulties: json array, contains the difficulties as arrays with the key 'score_median' for the median
and 'score_average' for the average difficulty the value (between 0 (easy) and 1 (hard)) and the 'id' parameter
to identify the corresponding text
"""
language = Language.find(lang_code)
if language is None:
return 'FAIL'
data = flask.request.get_json()
texts = []
if 'texts' in data:
for text in data['texts']:
texts.append(text)
else:
return 'FAIL'
personalized = True
if 'personalized' in data:
personalized = data['personalized'].lower()
if personalized == 'false' or personalized == '0':
personalized = False
rank_boundary = 10000.0
if 'rank_boundary' in data:
rank_boundary = float(data['rank_boundary'])
if rank_boundary > 10000.0:
rank_boundary = 10000.0
user = flask.g.user
known_probabilities = KnownWordProbability.find_all_by_user_cached(user)
difficulties = []
for text in texts:
# Calculate difficulty for each word
words = util.split_words_from_text(text['content'])
words_difficulty = []
for word in words:
ranked_word = RankedWord.find_cache(word, language)
word_difficulty = 1.0 # Value between 0 (easy) and 1 (hard)
if ranked_word is not None:
# Check if the user knows the word
try:
known_propability = known_probabilities[
word] # Value between 0 (unknown) and 1 (known)
except KeyError:
known_propability = None
if personalized and known_propability is not None:
word_difficulty -= float(known_propability)
elif ranked_word.rank <= rank_boundary:
word_frequency = (
rank_boundary - (ranked_word.rank - 1)
) / rank_boundary # Value between 0 (rare) and 1 (frequent)
word_difficulty -= word_frequency
words_difficulty.append(word_difficulty)
# Uncomment to print data for histogram generation
#text.generate_histogram(words_difficulty)
# Median difficulty for text
words_difficulty.sort()
center = int(round(len(words_difficulty) / 2, 0))
difficulty_median = words_difficulty[center]
# Average difficulty for text
difficulty_average = sum(words_difficulty) / float(
len(words_difficulty))
difficulties.append(
dict(score_median=difficulty_median,
score_average=difficulty_average,
id=text['id']))
response = json.dumps(dict(difficulties=difficulties))
return flask.Response(response, status=200, mimetype='application/json')
if os.environ.get("ZEEGUU_TESTING"):
db_name = "zeeguu_test"
if os.environ.get("ZEEGUU_PERFORMANCE_TESTING"):
db_name = "zeeguu_performance_test"
db_connection_string += mysql_hostname + "/" + db_name
app.config["SQLALCHEMY_DATABASE_URI"] = db_connection_string
else:
# Ooops: we are not testing, and we don't have a DB configured!
if not "SQLALCHEMY_DATABASE_URI" in app.config:
print "No db configured. You probably have no config file..."
exit()
print "->> DB Connection String: " + app.config["SQLALCHEMY_DATABASE_URI"]
setup_db_connection()
env = flask.ext.assets.Environment(app)
env.cache = app.instance_path
env.directory = os.path.join(app.instance_path, "gen")
env.url = "/gen"
env.append_path(
os.path.join(os.path.dirname(os.path.abspath(__file__)), "static"),
"/static")
db.init_app(app)
db.create_all(app=app)
from zeeguu.model import RankedWord
with app.app_context():
RankedWord.cache_ranked_words()
def get_difficulty_for_text(lang_code):
"""
URL parameters:
:param lang_code: the language of the text
Json data:
:param texts: json array that contains the texts to calculate the difficulty for. Each text consists of an array
with the text itself as 'content' and an additional 'id' which gets roundtripped unchanged
:param personalized (optional): calculate difficulty score for a specific user? (Enabled by default)
:param rank_boundary (optional): upper boundary for word frequency rank (between 1 and 10'000)
:return difficulties: json array, contains the difficulties as arrays with the key 'score_median' for the median
and 'score_average' for the average difficulty the value (between 0 (easy) and 1 (hard)) and the 'id' parameter
to identify the corresponding text
"""
language = Language.find(lang_code)
if language is None:
return 'FAIL'
data = flask.request.get_json()
texts = []
if 'texts' in data:
for text in data['texts']:
texts.append(text)
else:
return 'FAIL'
personalized = True
if 'personalized' in data:
personalized = data['personalized'].lower()
if personalized == 'false' or personalized == '0':
personalized = False
rank_boundary = 10000.0
if 'rank_boundary' in data:
rank_boundary = float(data['rank_boundary'])
if rank_boundary > 10000.0:
rank_boundary = 10000.0
user = flask.g.user
known_probabilities = KnownWordProbability.find_all_by_user_cached(user)
difficulties = []
for text in texts:
# Calculate difficulty for each word
words = util.split_words_from_text(text['content'])
words_difficulty = []
for word in words:
ranked_word = RankedWord.find_cache(word, language)
word_difficulty = 1.0 # Value between 0 (easy) and 1 (hard)
if ranked_word is not None:
# Check if the user knows the word
try:
known_propability = known_probabilities[word] # Value between 0 (unknown) and 1 (known)
except KeyError:
known_propability = None
if personalized and known_propability is not None:
word_difficulty -= float(known_propability)
elif ranked_word.rank <= rank_boundary:
word_frequency = (rank_boundary-(ranked_word.rank-1))/rank_boundary # Value between 0 (rare) and 1 (frequent)
word_difficulty -= word_frequency
words_difficulty.append(word_difficulty)
# Uncomment to print data for histogram generation
#text.generate_histogram(words_difficulty)
# Median difficulty for text
words_difficulty.sort()
center = int(round(len(words_difficulty)/2, 0))
difficulty_median = words_difficulty[center]
# Average difficulty for text
difficulty_average = sum(words_difficulty) / float(len(words_difficulty))
difficulties.append(dict(score_median=difficulty_median, score_average=difficulty_average, id=text['id']))
response = json.dumps(dict(difficulties=difficulties))
return flask.Response(response, status=200, mimetype='application/json')
if os.environ.get("ZEEGUU_TESTING"):
db_name = "zeeguu_test"
if os.environ.get("ZEEGUU_PERFORMANCE_TESTING"):
db_name = "zeeguu_performance_test"
db_connection_string += mysql_hostname+"/"+db_name
app.config["SQLALCHEMY_DATABASE_URI"] = db_connection_string
else:
# Ooops: we are not testing, and we don't have a DB configured!
if not "SQLALCHEMY_DATABASE_URI" in app.config:
print "No db configured. You probably have no config file..."
exit()
print "->> DB Connection String: " + app.config["SQLALCHEMY_DATABASE_URI"]
setup_db_connection()
env = flask.ext.assets.Environment(app)
env.cache = app.instance_path
env.directory = os.path.join(app.instance_path, "gen")
env.url = "/gen"
env.append_path(os.path.join(
os.path.dirname(os.path.abspath(__file__)), "static"
), "/static")
db.init_app(app)
db.create_all(app=app)
from zeeguu.model import RankedWord
with app.app_context():
RankedWord.cache_ranked_words()
