def parseTrainFile(trainFile):
train_f = open(trainFile, 'r')
queries = []
docs = []
query2docs = {}
for line in train_f:
key = line.split(':', 1)[0].strip()
value = line.split(':', 1)[1].strip()
if (key == 'query'):
query = value
queries.append(query)
query2docs[query] = []
elif (key == 'url'):
doc = Document(query)
docs.append(doc)
query2docs[query].append(doc)
doc.set_url(value)
elif (key == 'title'):
doc.set_title(value)
elif (key == 'header'):
doc.add_header(value)
elif (key == 'body_hits'):
temp = value.split(' ', 1)
doc.add_body_hits(temp[0].strip(), map(int,
temp[1].strip().split()))
elif (key == 'body_length'):
doc.set_body_length(int(value))
elif (key == 'pagerank'):
doc.set_pagerank(int(value))
elif (key == 'anchor_text'):
anchor_text = value
elif (key == 'stanford_anchor_count'):
doc.add_anchor(anchor_text, int(value))
train_f.close()
doc_count = len(docs)
avlen_url = sum(doc.url_length for doc in docs) / doc_count
avlen_title = sum(doc.title_length for doc in docs) / doc_count
avlen_header = sum(doc.header_length for doc in docs) / doc_count
avlen_body = sum(doc.body_length for doc in docs) / doc_count
avlen_anchor = sum(doc.anchor_length for doc in docs) / doc_count
with open('DocFreqDict', 'rb') as doc_freq_f:
doc_freq_dict = marshal.load(doc_freq_f)
for doc in docs:
doc.set_averages(avlen_url, avlen_title, avlen_header, avlen_body,
avlen_anchor)
doc.make_vectors(doc_freq_dict)
return queries, query2docs
def parseTrainFile(trainFile):
train_f = open(trainFile, 'r')
queries = []
docs = []
query2docs = {}
for line in train_f:
key = line.split(':', 1)[0].strip()
value = line.split(':', 1)[1].strip()
if (key == 'query'):
query = value
queries.append(query)
query2docs[query] = []
elif (key == 'url'):
doc = Document(query)
docs.append(doc)
query2docs[query].append(doc)
doc.set_url(value)
elif (key == 'title'):
doc.set_title(value)
elif (key == 'header'):
doc.add_header(value)
elif (key == 'body_hits'):
temp = value.split(' ', 1)
doc.add_body_hits(temp[0].strip(), map(int, temp[1].strip().split()))
elif (key == 'body_length'):
doc.set_body_length(int(value))
elif (key == 'pagerank'):
doc.set_pagerank(int(value))
elif (key == 'anchor_text'):
anchor_text = value
elif (key == 'stanford_anchor_count'):
doc.add_anchor(anchor_text, int(value))
train_f.close()
doc_count = len(docs)
avlen_url = sum(doc.url_length for doc in docs) / doc_count
avlen_title = sum(doc.title_length for doc in docs) / doc_count
avlen_header = sum(doc.header_length for doc in docs) / doc_count
avlen_body = sum(doc.body_length for doc in docs) / doc_count
avlen_anchor = sum(doc.anchor_length for doc in docs) / doc_count
with open('DocFreqDict', 'rb') as doc_freq_f:
doc_freq_dict = marshal.load(doc_freq_f)
for doc in docs:
doc.set_averages(avlen_url, avlen_title, avlen_header, avlen_body, avlen_anchor)
doc.make_vectors(doc_freq_dict)
return queries, docs, query2docs
CAL = "grafics/Kalender.svg"
FN = "grafics/Dateiname.svg"
if __name__ == "__main__":
##############################
# PORTRAIT DOCUMENT
doc = Document(orientation="portrait")
footerText = [["left", "left middle", "right middle", "right"]]
rightHeaderLines = [["top line"], ["middle line"], ["bottom line"]]
headerText = [["", "MAIN TITLE", ""]]
header = create_header(headerText, rightHeaderLines, LOGO)
footer = create_footer(footerText, [USER, CAL, "", FN])
doc.add_footer(footer)
doc.add_header(header)
spacer = create_vertical_spacer(4.8)
# table
titleTableText = [["Title", "Title"], ["Subtitle", "Subtitle"],
["Datum", "Datum"]]
titleTableData = {
"body": titleTableText,
"hTableAlignment": "CENTER",
"colWidths": [8, 8],
"fontsize": 15,
}
table_title = create_table(**titleTableData)
# create a figure
fig, leg = create_figure(hAlign=ar.TA_CENTER)
doc.add_page(bookmark=ar.Bookmark("Main Chapter", 0),
framestyle="single",
