if word.startswith("#"):
count = 0
if page in pagedict:
pagedict[page] += count
else:
pagedict[page] = count
htid = filename[0:-7]
filedict[htid] = pagedict
with open(
"/Users/tunder/Dropbox/pagedata/seventhfeatures/pagelevelwordcounts.tsv",
mode="w",
encoding="utf-8") as f:
f.write("htid\tpage\twordcount\n")
for htid, pagedict in filedict.items():
tuplelist = utils.sortvaluesbykey(pagedict)
counter = 0
for twotuple in tuplelist:
pagenum, count = twotuple
if pagenum < 0:
continue
elif pagenum != counter:
print("pagination anomaly")
else:
counter += 1
outline = htid + "\t" + str(pagenum) + "\t" + str(count) + "\n"
f.write(outline)
if word.startswith("#"):
count = 0
if page in pagedict:
pagedict[page] += count
else:
pagedict[page] = count
htid = filename[0:-7]
filedict[htid] = pagedict
with open("/Users/tunder/Dropbox/pagedata/seventhfeatures/pagelevelwordcounts.tsv", mode="w", encoding="utf-8") as f:
f.write("htid\tpage\twordcount\n")
for htid, pagedict in filedict.items():
tuplelist = utils.sortvaluesbykey(pagedict)
counter = 0
for twotuple in tuplelist:
pagenum, count = twotuple
if pagenum < 0:
continue
elif pagenum != counter:
print("pagination anomaly")
else:
counter += 1
outline = htid + "\t" + str(pagenum) + "\t" + str(count) + "\n"
f.write(outline)
