def download(doc_id):
try:
logger.debug("Download %s", doc_id)
doc = mongo.find_document('samediff', doc_id)
headers = [
_('word'),
_('uses in') + ' ' + doc['filenames'][0],
_('uses in') + ' ' + doc['filenames'][1],
_('total uses')
]
rows = []
for f, w in doc['sameWords']:
doc1Count = next(f2 for f2, w2 in doc['mostFrequentDoc1']
if w == w2)
doc2Count = next(f2 for f2, w2 in doc['mostFrequentDoc2']
if w == w2)
rows.append([w, doc1Count, doc2Count, f])
for f, w in doc['diffWordsDoc1']:
rows.append([w, f, 0, f])
for f, w in doc['diffWordsDoc1']:
rows.append([w, 0, f, f])
# TODO: clean up file name
file_path = filehandler.write_to_csv(
headers, rows,
filehandler.generate_filename('csv', '', doc['filenames'][0],
doc['filenames'][1]), False)
logger.debug(' created csv to download at %s', file_path)
return filehandler.generate_csv(file_path)
except Exception as e:
logging.exception(e)
abort(400)
def download_gexf(doc_id):
"""
Download GEXF file
"""
logger.info('[CTD] Requesting GEXF for doc: %s', doc_id)
doc = mongo.find_document('connectthedots', doc_id)
return Response(doc.get('results')['gexf'], mimetype='application/xml')
def results(doc_id):
remaining_days = None
try:
job = mongo.find_document('samediff', doc_id)
if job['sample_id'] == u'':
remaining_days = mongo.get_remaining_days('samediff', doc_id)
except:
logger.warning("Unable to find doc '%s'", doc_id)
return render_template('no_results.html', tool_name='samediff')
whatnext = {}
whatnext['most_common_word'] = job['sameWords'][0][1] if len(job['sameWords']) > 0 else ''
whatnext['second_most_common_word'] = job['sameWords'][1][1] if len(job['sameWords']) > 1 else ''
whatnext['doc2_most_common_word'] = job['diffWordsDoc2'][0][1] if len(job['diffWordsDoc2']) > 0 else ''
if(job['totalWordsDoc1'] > job['totalWordsDoc2']):
pct_length_diff = float(job['totalWordsDoc1'] - job['totalWordsDoc2']) / float(job['totalWordsDoc1'])
else:
pct_length_diff = float(job['totalWordsDoc2'] - job['totalWordsDoc1']) / float(job['totalWordsDoc2'])
return render_template('samediff/results.html', results=job,
pct_length_diff = pct_length_diff,
cosine_similarity= {'score':job['cosineSimilarity'],'description':interpretCosineSimilarity(job['cosineSimilarity'])},
whatnext=whatnext, tool_name='samediff', doc_id=doc_id,
remaining_days=remaining_days)
def results(doc_id):
remaining_days = None
try:
job = mongo.find_document('samediff', doc_id)
if job['sample_id'] == u'':
remaining_days = mongo.get_remaining_days('samediff', doc_id)
except:
logger.warning("Unable to find doc '%s'", doc_id)
return render_template('no_results.html', tool_name='samediff')
whatnext = {}
whatnext['most_common_word'] = job['sameWords'][0][1] if len(job['sameWords']) > 0 else ''
whatnext['second_most_common_word'] = job['sameWords'][1][1] if len(job['sameWords']) > 1 else ''
whatnext['doc2_most_common_word'] = job['diffWordsDoc2'][0][1] if len(job['diffWordsDoc2']) > 0 else ''
if(job['totalWordsDoc1'] > job['totalWordsDoc2']):
pct_length_diff = float(job['totalWordsDoc1'] - job['totalWordsDoc2']) / float(job['totalWordsDoc1'])
else:
pct_length_diff = float(job['totalWordsDoc2'] - job['totalWordsDoc1']) / float(job['totalWordsDoc2'])
return render_template('samediff/results.html', results=job,
pct_length_diff = pct_length_diff,
cosine_similarity= {'score':job['cosineSimilarity'],'description':interpretCosineSimilarity(job['cosineSimilarity'])},
whatnext=whatnext, tool_name='samediff', doc_id=doc_id,
remaining_days=remaining_days)
def download_gexf(doc_id):
"""
Download GEXF file
"""
logger.info('[CTD] Requesting GEXF for doc: %s', doc_id)
doc = mongo.find_document('connectthedots', doc_id)
return Response(doc.get('results')['gexf'], mimetype='application/xml')
def results(doc_id):
"""
Lookup results for a given document
"""
try:
results = mongo.find_document('connectthedots', doc_id).get('results')
logger.info('[CTD] Showing results for doc: %s', doc_id)
return render_results(doc_id)
except Exception as e:
logger.warning('[CTD] Unable to find doc: %s', doc_id)
logger.warning('[CTD] Error: %s', str(e))
return render_template('no_results.html', tool_name='connectthedots')
def download_table(doc_id):
"""
Download CSV of degree/centrality scores
"""
logger.info('[CTD] Requesting CSV of table for doc: %s', doc_id)
doc = mongo.find_document('connectthedots', doc_id)
def as_csv(rows, headers):
yield ','.join(headers) + '\n'
for r in rows:
yield ','.join(map(str, [r['id'], r['degree'], r['centrality'], r['community']])) + '\n'
return Response(
as_csv(doc.get('results')['table'], ['node', 'degree', 'betweenness centrality', 'community']),
mimetype='text/csv')
def render_results(doc_id):
"""
Render results page
"""
doc = mongo.find_document('connectthedots', doc_id)
results = doc.get('results')
if doc.get('source') != 'sample':
remaining_days = mongo.get_remaining_days('connectthedots', doc_id)
else:
remaining_days = None
first_mismatch = None # get first centrality/degree mismatch
degree_index = 0
centrality_index = 0
table_by_degree = sorted(results['table'],
key=operator.itemgetter('degree'),
reverse=True)
table_by_centrality = results['table']
for i, row in enumerate(table_by_degree):
if row['id'] != table_by_centrality[i]['id']:
first_mismatch = row['id']
degree_index = i
break
if first_mismatch is not None:
for i, row in enumerate(
table_by_centrality[degree_index +
1:]): # start from where we left off
if row['id'] == first_mismatch:
centrality_index = i + degree_index + 1
break
what_next = {
'mismatch_id': first_mismatch,
'mismatch_degree': ordinal(degree_index + 1),
'mismatch_centrality': ordinal(centrality_index + 1),
'lowest_degree': table_by_degree[-1]['id']
}
biography = results['biography'] if 'biography' in results else None
return render_template('connectthedots/results.html',
results=results,
whatnext=what_next,
tool_name='connectthedots',
source=doc['source'],
has_multiple_sheets=results['has_multiple_sheets'],
remaining_days=remaining_days,
biography=biography)
def download_csv(doc_id, analysis_type):
logger.debug("Download %s", analysis_type)
if analysis_type not in ['words','bigrams','trigrams']:
logger.warning("Requested unknown csv type: %s",analysis_type)
abort(400)
try:
doc = mongo.find_document('wordcounter', doc_id)
except:
logger.warning("Unable to find doc '%s'", doc_id)
abort(400)
file_path = create_csv_file(doc.get('counts'),analysis_type)
logger.debug(' created %s csv to download at %s', analysis_type, file_path)
if file_path is None:
abort(500)
return filehandler.generate_csv(file_path)
def download_csv(doc_id, analysis_type):
logger.debug("Download %s", analysis_type)
if analysis_type not in ['words','bigrams','trigrams']:
logger.warning("Requested unknown csv type: %s",analysis_type)
abort(400)
try:
doc = mongo.find_document('wordcounter', doc_id)
except:
logger.warning("Unable to find doc '%s'", doc_id)
abort(400)
file_path = create_csv_file(doc.get('counts'),analysis_type)
logger.debug(' created %s csv to download at %s', analysis_type, file_path)
if file_path is None:
abort(500)
return filehandler.generate_csv(file_path)
def results_page(doc_id):
try:
results = mongo.find_document('wtfcsv', doc_id).get('results')
# Doc has more than one sheet to analyze
if len(results) > 1:
logger.info("Showing results %s (sheet 0)", doc_id)
submit = request.args.get('submit', '')
param = '?submit=true' if 'true' in submit else ''
return redirect(g.current_lang + '/wtfcsv/results/' + doc_id +
'/sheets/0' + param)
else:
logger.info("Showing results %s", doc_id)
return render_results(doc_id, 0)
except Exception as e:
logger.warning("Unable to find doc '%s'", doc_id)
logger.exception(e)
return render_template('no_results.html', tool_name='wtfcsv')
def render_results(doc_id):
"""
Render results page
"""
doc = mongo.find_document('connectthedots', doc_id)
results = doc.get('results')
if doc.get('source') != 'sample':
remaining_days = mongo.get_remaining_days('connectthedots', doc_id)
else:
remaining_days = None
first_mismatch = None # get first centrality/degree mismatch
degree_index = 0
centrality_index = 0
table_by_degree = sorted(results['table'], key=operator.itemgetter('degree'), reverse=True)
table_by_centrality = results['table']
for i, row in enumerate(table_by_degree):
if row['id'] != table_by_centrality[i]['id']:
first_mismatch = row['id']
degree_index = i
break
if first_mismatch is not None:
for i, row in enumerate(table_by_centrality[degree_index + 1:]): # start from where we left off
if row['id'] == first_mismatch:
centrality_index = i + degree_index + 1
break
whatnext = {}
whatnext['mismatch_id'] = first_mismatch
whatnext['mismatch_degree'] = ordinal(degree_index + 1)
whatnext['mismatch_centrality'] = ordinal(centrality_index + 1)
whatnext['lowest_degree'] = table_by_degree[-1]['id']
return render_template('connectthedots/results.html',
results=results,
whatnext=whatnext,
tool_name='connectthedots',
source=doc['source'],
has_multiple_sheets=results['has_multiple_sheets'],
remaining_days=remaining_days)
def download_table(doc_id):
"""
Download CSV of degree/centrality scores
"""
logger.info('[CTD] Requesting CSV of table for doc: %s', doc_id)
doc = mongo.find_document('connectthedots', doc_id)
def as_csv(rows, headers):
yield ','.join(headers) + '\n'
for r in rows:
yield ','.join(
map(str,
[r['id'], r['degree'], r['centrality'], r['community']
])) + '\n'
return Response(as_csv(
doc.get('results')['table'],
['node', 'degree', 'betweenness centrality', 'community']),
mimetype='text/csv')
def results(doc_id):
try:
results = mongo.find_document('wtfcsv', doc_id).get('results')
#Doc has more than one sheet to analyze
if len(results) > 1:
logger.info("Showing results %s (sheet 0)", doc_id)
submit = request.args.get('submit', '')
param = '?submit=true' if 'true' in submit else ''
return redirect(g.current_lang + '/wtfcsv/results/' + doc_id + '/sheets/0' + param)
else:
logger.info("Showing results %s", doc_id)
return render_results(doc_id, 0)
except:
#more robust exception logging
logger.warning("Unable to find doc '%s'", doc_id)
logger.warning("Unexpected error:", sys.exc_info()[0])
logger.warning(traceback.format_exc())
return render_template('no_results.html', tool_name='wtfcsv')
def results(doc_id):
try:
results = mongo.find_document('wtfcsv', doc_id).get('results')
#Doc has more than one sheet to analyze
if len(results) > 1:
logger.info("Showing results %s (sheet 0)", doc_id)
submit = request.args.get('submit', '')
param = '?submit=true' if 'true' in submit else ''
return redirect(g.current_lang + '/wtfcsv/results/' + doc_id +
'/sheets/0' + param)
else:
logger.info("Showing results %s", doc_id)
return render_results(doc_id, 0)
except:
#more robust exception logging
logger.warning("Unable to find doc '%s'", doc_id)
logger.warning("Unexpected error:", sys.exc_info()[0])
logger.warning(traceback.format_exc())
return render_template('no_results.html', tool_name='wtfcsv')
def download(doc_id):
try:
logger.debug("Download %s", doc_id)
doc = mongo.find_document('samediff', doc_id)
headers = [_('word'), _('uses in') +' ' + doc['filenames'][0], _('uses in') + ' ' + doc['filenames'][1], _('total uses')]
rows = []
for f, w in doc['sameWords']:
doc1Count = next(f2 for f2, w2 in doc['mostFrequentDoc1'] if w == w2)
doc2Count = next(f2 for f2, w2 in doc['mostFrequentDoc2'] if w == w2)
rows.append([w, doc1Count, doc2Count, f])
for f, w in doc['diffWordsDoc1']:
rows.append([w, f, 0, f])
for f, w in doc['diffWordsDoc1']:
rows.append([w, 0, f, f])
# TODO: clean up file name
file_path = filehandler.write_to_csv(headers, rows,
filehandler.generate_filename('csv', '', doc['filenames'][0], doc['filenames'][1]), False)
logger.debug(' created csv to download at %s', file_path)
return filehandler.generate_csv(file_path)
except Exception as e:
logging.exception(e)
abort(400)
def render_results(doc_id, sheet_idx):
doc = mongo.find_document('wtfcsv', doc_id)
results = doc.get('results')
if doc['sample_id'] == u'':
remaining_days = mongo.get_remaining_days('wtfcsv', doc_id)
else:
remaining_days = None
if 'bad_formatting' in results:
return render_template('wtfcsv/results.html',
results=results,
tool_name='wtfcsv',
index=0)
def get_random_column():
return random.choice(results[int(sheet_idx)]['columns'])
columns = results[int(sheet_idx)]['columns']
if len(columns) < 1:
whatnext = 'no_data'
else:
random_column = get_random_column()
random_column2 = get_random_column()
random_column3 = get_random_column()
if len(columns) > 0 and next(
(c for c in columns if 'most_freq_values' in c), None) is not None:
while 'most_freq_values' not in random_column:
random_column = get_random_column()
if len(columns) > 1:
while random_column2 == random_column:
random_column2 = get_random_column()
else:
random_column2 = random_column
if len(columns) > 2:
while random_column3 == random_column or random_column3 == random_column2:
random_column3 = get_random_column()
else:
random_column3 = random_column
whatnext = {}
if 'most_freq_values' in random_column and len(
random_column['most_freq_values']) > 0:
whatnext[
'random_column_top_value'] = random_column['most_freq_values'][
0]['value'] if 'most_freq_values' in random_column else ''
else:
whatnext['random_column_top_value'] = 0
whatnext['random_column_name'] = random_column['name']
whatnext['random_column_name2'] = random_column2['name']
whatnext['random_column_name3'] = random_column3['name']
# build a list of summary result data for the chart
for col in columns:
is_string = 'text' in col['display_type_name']
data_to_use = []
# pick the right results to summarize
if 'deciles' in col:
data_to_use = col['deciles']
elif 'most_freq_values' in col:
data_to_use = col['most_freq_values']
elif 'word_counts' in col:
#for word in col['word_counts']['unique_words'][:20]:
# print str(word[0]) + " is " + str(word[1])
data_to_use = [{
'value': word[0],
'count': word[1]
} for word in col['word_counts']['unique_words'][:20]]
# stitch together the overview
overview_data = {'categories': [], 'values': []}
for d in data_to_use:
key = str(d['value']) if is_string else str(d['value']).replace(
'_', '.')
overview_data['categories'].append(key)
overview_data['values'].append(d['count'])
if 'others' in col:
overview_data['categories'].append(gettext('Other'))
overview_data['values'].append(int(col['others']))
col['overview'] = overview_data
return render_template('wtfcsv/results.html',
results=results,
whatnext=whatnext,
tool_name='wtfcsv',
index=int(sheet_idx),
source=doc['source'],
remaining_days=remaining_days)
def results_for_doc(doc_id):
results = {}
remaining_days = None
try:
doc = mongo.find_document('wordcounter', doc_id)
if doc['sample_id'] == '':
remaining_days = mongo.get_remaining_days('wordcounter', doc_id)
except:
logger.warning("Unable to find doc '%s'", doc_id)
return render_template('no_results.html', tool_name='wordcounter')
counts = doc.get('counts')
# only render the top 40 results on the page (the csv contains all results)
results['unique_words'] = counts['unique_words'][:40]
results['bigrams'] = counts['bigrams'][:40]
results['trigrams'] = counts['trigrams'][:40]
max_index = min(20, len(results['unique_words']))
min_index = max(0, max_index - 5)
random_unpopular_word = ['', '']
top_word = ''
word_in_bigrams_count = 0
word_in_trigrams_count = 0
if len(results['unique_words']) > 0:
random_unpopular_word = results['unique_words'][random.randrange(min_index, max_index+1)]\
if len(results['unique_words']) > 1 else results['unique_words'][0]
'''
Find the most popular word that is also present in bigrams and trigrams.
If none can be found, just get the most popular word.
'''
if results['unique_words'] and results['bigrams'] and results[
'trigrams']:
for word in results['unique_words']:
top_word = word[0]
word_in_bigrams_count = 0
word_in_trigrams_count = 0
for b in results['bigrams']:
if top_word in b[0]:
word_in_bigrams_count += 1
for t in results['trigrams']:
if top_word in t[0]:
word_in_trigrams_count += 1
if word_in_bigrams_count > 0 and word_in_trigrams_count > 0:
break
if word_in_bigrams_count == 0 and word_in_trigrams_count == 0:
top_word = results['unique_words'][0][0]
whatnext = {}
whatnext['top_word'] = top_word
whatnext['word_in_bigrams_count'] = word_in_bigrams_count
whatnext['word_in_trigrams_count'] = word_in_trigrams_count
whatnext['random_unpopular_word'] = random_unpopular_word[0]
whatnext['random_unpopular_word_count'] = random_unpopular_word[1]
biography = doc['biography'] if 'biography' in doc else None
return render_template('wordcounter/results.html',
results=results,
whatnext=whatnext,
tool_name='wordcounter',
title=doc['title'],
doc_id=doc_id,
source=doc['source'],
remaining_days=remaining_days,
total_words=counts['total_word_count'],
biography=biography)
def render_results(doc_id, sheet_idx):
doc = mongo.find_document('wtfcsv', doc_id)
results = doc.get('results')
if doc['sample_id'] == u'':
remaining_days = mongo.get_remaining_days('wtfcsv', doc_id)
else:
remaining_days = None
if 'bad_formatting' in results:
return render_template('wtfcsv/results.html', results=results, tool_name='wtfcsv', index=0)
def get_random_column():
return random.choice(results[int(sheet_idx)]['columns'])
columns = results[int(sheet_idx)]['columns']
if len(columns) < 1:
whatnext = 'no_data'
else:
random_column = get_random_column()
random_column2 = get_random_column()
random_column3 = get_random_column()
if len(columns) > 0 and next((c for c in columns if 'most_freq_values' in c), None) is not None:
while 'most_freq_values' not in random_column:
random_column = get_random_column()
if len(columns) > 1:
while random_column2 == random_column:
random_column2 = get_random_column()
else:
random_column2 = random_column
if len(columns) > 2:
while random_column3 == random_column or random_column3 == random_column2:
random_column3 = get_random_column()
else:
random_column3 = random_column
whatnext = {}
if 'most_freq_values' in random_column and len(random_column['most_freq_values']) > 0:
whatnext['random_column_top_value'] = random_column['most_freq_values'][0]['value'] if 'most_freq_values' in random_column else ''
else:
whatnext['random_column_top_value'] = 0
whatnext['random_column_name'] = random_column['name']
whatnext['random_column_name2'] = random_column2['name']
whatnext['random_column_name3'] = random_column3['name']
# build a list of summary result data for the chart
for col in columns:
is_string = 'text' in col['display_type_name']
data_to_use = []
# pick the right results to summarize
if 'deciles' in col:
data_to_use = col['deciles']
elif 'most_freq_values' in col:
data_to_use = col['most_freq_values']
elif 'word_counts' in col:
for word in col['word_counts']['unique_words'][:20]:
print str(word[0]) + " is " + str(word[1])
data_to_use = [ {'value':word[0], 'count':word[1]} for word in col['word_counts']['unique_words'][:20] ]
# stitch together the overview
overview_data = {'categories':[],'values':[]}
for d in data_to_use:
key = str(d['value']) if is_string else str(d['value']).replace('_', '.')
overview_data['categories'].append(key)
overview_data['values'].append(d['count'])
if 'others' in col:
overview_data['categories'].append(gettext('Other'))
overview_data['values'].append(int(col['others']))
col['overview'] = overview_data
return render_template('wtfcsv/results.html',
results=results,
whatnext=whatnext,
tool_name='wtfcsv',
index=int(sheet_idx),
source=doc['source'],
remaining_days=remaining_days)
def results(doc_id):
counts = None
results = {}
remaining_days = None
try:
doc = mongo.find_document('wordcounter', doc_id)
if doc['sample_id'] == u'':
remaining_days = mongo.get_remaining_days('wordcounter', doc_id)
except:
logger.warning("Unable to find doc '%s'", doc_id)
return render_template('no_results.html', tool_name='wordcounter')
counts = doc.get('counts')
# only render the top 40 results on the page (the csv contains all results)
results['unique_words'] = counts['unique_words'][:40]
results['bigrams'] = counts['bigrams'][:40]
results['trigrams'] = counts['trigrams'][:40]
max_index = min(20, len(results['unique_words']))
min_index = max(0, max_index-5)
random_unpopular_word = ['','']
top_word = ''
word_in_bigrams_count = 0
word_in_trigrams_count = 0
if len(results['unique_words']) > 0:
random_unpopular_word = results['unique_words'][random.randrange(min_index, max_index+1)] if len(results['unique_words']) > 1 else results['unique_words'][0]
'''
Find the most popular word that is also present in bigrams and trigrams.
If none can be found, just get the most popular word.
'''
if results['unique_words'] and results['bigrams'] and results['trigrams']:
for word in results['unique_words']:
top_word = word[0]
word_in_bigrams_count = 0
word_in_trigrams_count = 0
for b in results['bigrams']:
if top_word in b[0]:
word_in_bigrams_count += 1
for t in results['trigrams']:
if top_word in t[0]:
word_in_trigrams_count += 1
if word_in_bigrams_count > 0 and word_in_trigrams_count > 0:
break
if word_in_bigrams_count == 0 and word_in_trigrams_count == 0:
top_word = results['unique_words'][0][0]
whatnext = {}
whatnext['top_word'] = top_word
whatnext['word_in_bigrams_count'] = word_in_bigrams_count
whatnext['word_in_trigrams_count'] = word_in_trigrams_count
whatnext['random_unpopular_word'] = random_unpopular_word[0]
whatnext['random_unpopular_word_count'] = random_unpopular_word[1]
return render_template('wordcounter/results.html',
results=results,
whatnext=whatnext,
tool_name='wordcounter',
title=doc['title'],
doc_id=doc_id,
source=doc['source'],
remaining_days=remaining_days,
total_words=counts['total_word_count'])
