#!/usr/bin/python2
# Thomas Heller, 2002年04月04日
#
# A script to look up keywords in the Python manuals index
#
# $Id: pyhelp.py,v 1.9 2002年04月05日 19:45:14 thomas Exp $
#
# $Log: pyhelp.py,v $
# Revision 1.9 2002年04月05日 19:45:14 thomas
# More text in the CGI HTML page.
#
# Revision 1.5 2002年04月05日 19:33:19 thomas
# More HTML code.
#
# Revision 1.4 2002年04月05日 16:15:07 thomas
# Valid HTML 4.01. charset is utf-8, is this correct?
#
# Revision 1.3 2002年04月05日 14:51:26 thomas
# Now also works as CGI script (has even been tested on starship)
#
# Revision 1.2 2002年04月05日 14:49:34 thomas
# Pickles the found links to disk instead of downloading the index pages
# everytime.
#
# Revision 1.1 2002年04月05日 14:48:15 thomas
# First version, posted to python-dev asking for comments.
#
import htmllib, formatter, re
import urllib, webbrowser
import sys, os
if __name__ == '__main__':
__file__ = sys.argv[0]
__version__ = "$Revision: 1.9 $"[11:-2]
DOCMAP = {
"2.0": "http://www.python.org/doc/2.0/",
"2.1": "http://www.python.org/doc/2.1/",
"2.2": "http://www.python.org/doc/2.2/",
"devel": "http://www.python.org/dev/doc/devel/",
"current": "http://www.python.org/doc/current/",
# Can alternatively use local documentation!
## "2.2": "file:c:/python22/doc/",
## "2.1": "file:c:/python21/doc/",
## "2.0": "file:c:/python20/doc/",
}
INDEXPAGE = "genindex.html" # XXX Only valid for 2.0 and above
SECTIONS = "api/ ref/ lib/".split()
# modified from an example in the eff-bot guide to the Python Library...
class Parser(htmllib.HTMLParser):
def __init__(self, url, verbose=0):
self.anchors = {}
f = formatter.NullFormatter()
htmllib.HTMLParser.__init__(self, f, verbose)
self.last_text = ""
self.url = url
def anchor_bgn(self, href, name, type):
self.save_bgn()
self.anchor = self.url + href
def anchor_end(self):
text = self.save_end().strip()
if text == "[Link]" and self.last_text:
text = self.last_text
if self.anchor and text:
self.anchors[text] = self.anchors.get(text, []) + [self.anchor]
self.last_text = text
def get_anchors(version, rebuild):
# returns a list of (topic, url) pairs
# if rebuild is true, the index is rebuilt
# if rebuild is false, the index is rebuild if not present
import cPickle
baseurl = DOCMAP[version]
pathname = baseurl
for char in ":/\\":
pathname = pathname.replace(char, "-")
pathname = pathname + version + ".index"
if not rebuild:
try:
file = open(pathname, "rb")
data = cPickle.load(file)
return data
except (IOError, cPickle.PickleError):
pass
a = []
for sec in SECTIONS:
print "Downloading", baseurl + sec
file = urllib.urlopen(baseurl + sec + INDEXPAGE)
html = file.read()
file.close()
print "Parsing", baseurl + sec
p = Parser(baseurl + sec)
p.feed(html)
p.close()
a.extend(p.anchors.items())
try:
file = open(pathname, "wb")
except IOError, detail:
print detail
print os.path.abspath("index" + version)
else:
cPickle.dump(a, file, 1)
return a
def find_topics(topic, version, regexp, rebuild):
v = []
if regexp:
pat = re.compile(topic)
for key, urls in get_anchors(version, rebuild):
if pat.match(key):
for url in urls:
v.append((key, url))
else:
for key, urls in get_anchors(version, rebuild):
if key.startswith(topic):
for url in urls:
v.append((key, url))
v.sort()
return v
def get_tempdir():
import tempfile
tempfile.mktemp()
return tempfile.tempdir
def help(topic, version="2.2", regexp=0, rebuild=0):
baseurl = DOCMAP[version]
v = find_topics(topic, version, regexp, rebuild)
if len(v) == 0:
print "Not found"
elif len(v) == 1: # only one topic found, display directly
webbrowser.open(v[0][1])
else:
# create a temporary HTML page displaying links to the
# search results. Unfortunately the file cannot be deleted,
# because it may still be needed by the browser.
path = os.path.join(get_tempdir(), "pyhelp-results.html")
print path
file = open(path, "w")
file.write("
\n")
if regexp:
file.write("
Search results for '%s':
\n" % topic)
else:
file.write("
Regexp search results for '%s':
\n" % topic)
for topic, url in v:
file.write("%s \n" % (url, topic))
file.write("
\n")
file.close()
webbrowser.open(path)
def cgi_help():
import cgi, cgitb
cgitb.enable()
form = cgi.FieldStorage()
print "Content-type: text/html; charset=utf-8"
print
print ''
print "
Search Python Manual"
print ""
version = "2.2"
if form.has_key("version"):
version = form["version"].value
regexp = 0
if form.has_key("regexp"):
regexp = form["regexp"].value
if form.has_key("keyword"):
baseurl = DOCMAP[version]
v = find_topics(form["keyword"].value, version=version,
regexp=regexp, rebuild=0)
print ""
topic = form["keyword"].value
print "
%d search results for '%s':
" % (len(v), topic)
for topic, url in v:
print '%s \n' % (url, topic)
print '''
This script looks up keywords in the Python
Library Reference,
Language Reference,
and Python/C API
manuals.
For full text searches better consult the usual web search engines.
'''
print '''Valid HTML 4.01!Powered by Python
'''
print '''
%s, version %s
Please send any comments/suggestions to
Thomas Heller.
''' % (os.path.basename(__file__), __version__)
print ""
def main():
os.chdir(os.path.dirname(os.path.abspath(sys.argv[0])))
if os.environ.has_key("SCRIPT_NAME"):
cgi_help()
sys.exit()
import getopt
try:
opts, args = getopt.getopt(sys.argv[1:], "v:rb", ["version=", "useregexp", "rebuild"])
except getopt.GetoptError:
print "Usage: %s [-b] [-r] [-v version] topic" % sys.argv[0]
sys.exit(1)
version = "2.2"
regexp = 0
rebuild = 0
for o, a in opts:
if o in ("-v", "--version"):
if a not in DOCMAP.keys():
print "version must be one of %s" % ", ".join(DOCMAP.keys())
sys.exit(1)
version = a
if o in ("-r", "--useregexpg"):
regexp = 1
if o in ("-b", "--rebuild"):
rebuild = 1
if len(args) != 1:
print "Usage: %s [-r] [-v version] topic" % sys.argv[0]
sys.exit(1)
os.chdir(os.path.dirname(os.path.abspath(sys.argv[0])))
help(args[0], version, regexp, rebuild)
if __name__ == '__main__':
main()
# -- EOF --