generatenice.py

# DEPRECATED. DO NOT USE. LEAVING THIS HERE JUST IN CASE PEOPLE FIND IT USEFUL SOMEHOW
# AS IT IS LESS COMPLICATED THAN generatenicelda.py

# creates the nice .html page
# assumes that pdftowordcloud.py, pdftothumbs.py and scrape.py were already run

import cPickle as pickle

# load the pickle of papers scraped from the HTML page (result of scrape.py)
paperdict = pickle.load(open( "papers.p", "rb" ))
print "Loaded %d papers from papers.p (generated by scrape.py)" % (len(paperdict), )

# load the top word frequencies (result of pdftowordcloud.py)
topdict = pickle.load(open("topwords.p", "rb"))
print "Loaded %d entries from topwords.p (generated by pdftowordcloud.py)" % (len(topdict), )

# build up the string
html = open("nipsnice_template.html", "r").read()
s = ""
for p in paperdict:

	# get title, author
	title, author = paperdict[p]

	# get top words
	topwords = topdict.get(p, [])
	t = [x[0] for x in topwords]
	tcat = ", ".join(t)

	# get path to thumbnails for this paper
	thumbpath = "thumbs/NIPS2012_%s.pdf.jpg" % (p, )

	# get links to PDF, supplementary and bibtex on NIPS servers
	pdflink = "http://books.nips.cc/papers/files/nips25/NIPS2012_%s.pdf" % (p, )
	bibtexlink = "http://books.nips.cc/papers/files/nips25/bibhtml/NIPS2012_%s.html" % (p, )
	supplink = "http://books.nips.cc/papers/files/nips25/NIPS2012_%s.extra.zip" % (p, )

	s += """

	<div class="apaper">
	<div class="paperdesc">
		<span class="ts">%s</span><br />
		<span class="as">%s</span><br /><br />
	</div>
	<div class="dllinks">
		<a href="%s">[pdf] </a>
		<a href="%s">[bibtex] </a>
		<a href="%s">[supplementary]<br /></a>
	</div>
	<img src = "%s"><br />
	<span class="tt">TOP 100 words: %s</span>
	</div>

	""" % (title, author, pdflink, bibtexlink, supplink, thumbpath, tcat)

newhtml = html.replace("RESULTTABLE", s)

f = open("nipsnice.html", "w")
f.write(newhtml)
f.close()