X-Git-Url: https://git.ucc.asn.au/?a=blobdiff_plain;f=icalparse.py;h=c5d469e9feb14b2976addac2086c5b3ef04b4a73;hb=d26bdde2386225d1c5dbd2ccc425df1749cecef9;hp=3cdc7f1c6222c68968e31ea71b7f0044a7e888bf;hpb=08ba47d0e811c700411e763559a4e535cc459cf0;p=frenchie%2Ficalparse.git diff --git a/icalparse.py b/icalparse.py index 3cdc7f1..c5d469e 100755 --- a/icalparse.py +++ b/icalparse.py @@ -1,6 +1,6 @@ -#!/usr/bin/python +#!/usr/bin/python3 # -# Copyright (c) 2010 James French +# Copyright (c) 2013 James French # # Permission is hereby granted, free of charge, to any person obtaining a copy # of this software and associated documentation files (the "Software"), to deal @@ -20,96 +20,11 @@ # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN # THE SOFTWARE. -import sys -import urlparse -import os +import sys, os +from urllib import (request, error, parse) +import vobject from cgi import parse_header - -class InvalidICS(Exception): pass -class IncompleteICS(InvalidICS): pass - -def lineJoiner(oldcal, encoding='utf-8'): - '''Takes a string containing a calendar and returns an array of its lines''' - try: - oldcal = unicode(oldcal, encoding) - oldcal = oldcal.replace('\r\n ', '').replace('\r\n\t','') - except UnicodeDecodeError: - # This is probably a file with badly folded lines - oldcal = oldcal.replace('\r\n ', '').replace('\r\n\t','') - try: oldcal = unicode(oldcal, encoding) - except UnicodeDecodeError: - sys.stderr.write('Malformed File') - raise - - if not oldcal[0:15] == 'BEGIN:VCALENDAR': - raise InvalidICS, "Does not appear to be a valid ICS file" - - if not 'END:VCALENDAR' in oldcal[-15:-1]: - raise IncompleteICS, "File appears to be incomplete" - - if list(oldcal) == oldcal: - oldcal = '\r\n'.join(oldcal) - - return oldcal.split('\r\n') - - -def lineFolder(oldcal, length=75): - '''Folds content lines to a specified length, returns a list''' - - if length > 75: - sys.stderr.write('WARN: lines > 75 octets are not RFC compliant\n') - - cal = [] - sl = length - 1 - - for uline in oldcal: - line = uline.encode('utf-8') - - # Line fits inside length, do nothing - if len(line) <= length: - cal.append(line) - - else: - ll = length - foldedline = [] - while uline: - ufold = unicode(line[0:ll], 'utf-8', 'ignore') - fold = ufold.encode('utf-8') - uline = uline.replace(ufold,u'',1) - line = uline.encode('utf-8') - foldedline.append(fold) - - # Subsequent lines are shorter as they include a space - ll = length - 1 - cal.append('\r\n '.join(foldedline)) - - return cal - - -def splitFields(cal): - '''Takes a list of lines in a calendar file and returns a list of tuples - as (key, value) pairs''' - - ical = [] - - # Check that we got 2 items on every line - for line in [tuple(x.split(':',1)) for x in cal]: - if not len(line) == 2 and line[0]: - raise InvalidICS, 'Unusual content line: %s'%line - elif line[0]: - ical.append(line) - - return ical - - -def joinFields(ical): - '''Takes a list of tuples that make up a calendar file and returns it to a - list of lines''' - - return [':'.join(x) for x in ical] - - def getContent(url='',stdin=False): '''Generic content retriever, DO NOT use this function in a CGI script as it can read from the local disk (which you probably don't want it to). @@ -119,7 +34,7 @@ def getContent(url='',stdin=False): # Special case, if this is a HTTP url, return the data from it using # the HTTP functions which attempt to play a bit nicer. - parsedURL = urlparse.urlparse(url) + parsedURL = parse.urlparse(url) if 'http' in parsedURL[0]: return getHTTPContent(url) if stdin: @@ -129,13 +44,12 @@ def getContent(url='',stdin=False): if not parsedURL[0]: url = 'file://' + os.path.abspath(url) # If we've survived, use python's generic URL opening library to handle it - import urllib2 try: - res = urllib2.urlopen(url) + res = request.urlopen(url) content = res.read() ct = res.info().getplist() res.close() - except (urllib2.URLError, OSError), e: + except (error.URLError, OSError) as e: sys.stderr.write('%s\n'%e) sys.exit(1) @@ -147,28 +61,28 @@ def getContent(url='',stdin=False): return (content, encoding) -def getHTTPContent(url='',cache='.httplib2-cache'): +def getHTTPContent(url=''): '''This function attempts to play nice when retrieving content from HTTP services. It's what you should use in a CGI script.''' try: import httplib2 except ImportError: - import urllib2 + pass if not url: return ('','') - if not 'http' in urlparse.urlparse(url)[0]: return ('','') + if not 'http' in parse.urlparse(url)[0]: return ('','') if 'httplib2' in sys.modules: - try: h = httplib2.Http('.httplib2-cache') + try: h = httplib2.Http() except OSError: h = httplib2.Http() else: h = False if h: try: req = h.request(url) - except ValueError, e: + except ValueError as e: sys.stderr.write('%s\n'%e) sys.exit(1) @@ -184,8 +98,8 @@ def getHTTPContent(url='',cache='.httplib2-cache'): else: try: - req = urllib2.urlopen(url) - except urllib2.URLError, e: + req = request.urlopen(url) + except error.URLError as e: sys.stderr.write('%s\n'%e) sys.exit(1) @@ -199,48 +113,27 @@ def getHTTPContent(url='',cache='.httplib2-cache'): return (content, encoding) -def generateRules(): +def generateRules(ruleConfig): '''Attempts to load a series of rules into a list''' try: import parserrules except ImportError: return [] + for conf in ruleConfig: + parserrules.ruleConfig[conf] = ruleConfig[conf] + rules = [getattr(parserrules, rule) for rule in dir(parserrules) if callable(getattr(parserrules, rule))] return rules -def applyRules(ical, rules=[], verbose=False): +def applyRules(cal, rules=[], verbose=False): 'Runs a series of rules on the lines in ical and mangles its output' for rule in rules: - output = [] - if rule.__doc__ and verbose: - print(rule.__doc__) - for line in ical: - try: - out = rule(line[0],line[1]) - except TypeError, e: - output.append(line) - print(e) - continue - - # Drop lines that are boolean False - if not out and not out == None: continue - - # If the rule did something and is a tuple or a list we'll accept it - # otherwise, pay no attention to the man behind the curtain - try: - if tuple(out) == out or list(out) == out and len(out) == 2: - output.append(tuple(out)) - else: - output.append(line) - except TypeError, e: - output.append(line) - - ical = output - - return ical + cal = rule(cal) + + return cal def writeOutput(cal, outfile=''): @@ -255,37 +148,20 @@ def writeOutput(cal, outfile=''): else: try: out = open(outfile, 'w') - except (IOError, OSError), e: + except (IOError, OSError) as e: sys.stderr.write('%s\n'%e) sys.exit(1) - if cal[-1]: cal.append('') - - out.write('\r\n'.join(cal)) + cal.serialize(out) if not out == sys.stdout: out.close() -def vobjectRules(ics): - '''Applies rules to the ICS file before we have our way with it''' +def runLocal(): + '''Main run function if this script is called locally''' - try: - import vobjectRules - except ImportError: - sys.stderr.write('Vobject rules file could not be imported\n') - return ics - - for rule in vobjectRules.runRules: - ics = rule(ics) - - return ics - - -if __name__ == '__main__': from optparse import OptionParser - # If the user passed us a 'stdin' argument, we'll go with that, - # otherwise we'll try for a url opener parser = OptionParser('usage: %prog [options] url') parser.add_option('-s', '--stdin', action='store_true', dest='stdin', @@ -297,12 +173,14 @@ if __name__ == '__main__': parser.add_option('-m','--encoding', dest='encoding', default='', help='Specify a different character encoding' '(ignored if the remote server also specifies one)') - parser.add_option('-r','--vobject-rules', - action='store_true', dest='vobject', - help='Run rules written for vobject stored in vobjectRules.py') + parser.add_option('-t','--timezone', dest='timezone', default='', + help='Specify a timezone to use if the remote calendar doesn\'t set it properly') (options, args) = parser.parse_args() + ruleConfig["defaultTZ"] = options.timezone or ruleConfig["defaultTZ"] + # If the user passed us a 'stdin' argument, we'll go with that, + # otherwise we'll try for a url opener if not args and not options.stdin: parser.print_usage() sys.exit(0) @@ -313,8 +191,75 @@ if __name__ == '__main__': (content, encoding) = getContent(url, options.stdin) encoding = encoding or options.encoding or 'utf-8' - if options.vobject: content = vobjectRules(content) - cal = lineJoiner(content, encoding) - ical = applyRules(splitFields(cal), generateRules(), options.verbose) - output = lineFolder(joinFields(ical)) - writeOutput(output, options.outfile) + + cal = vobject.readOne(str(content, encoding)) + cal = applyRules(cal, generateRules(ruleConfig), options.verbose) + + writeOutput(cal, options.outfile) + + +def exitQuiet(exitstate=0): + '''When called as a CGI script, exit quietly if theres any errors''' + print('Content-Type: text/html\n') + sys.exit(exitstate) + + +def runCGI(): + '''Main run function if this script is called as a CGI script + to process facebook ical files''' + import cgi + import re + import cgitb; cgitb.enable() + + ruleConfig["facebook"] = True + + form = cgi.FieldStorage() + if "uid" not in form or "key" not in form: + print('Content-Type: text/calendar\n') + sys.exit(0) + try: + # UID should be numeric, if it's not we have someone playing games + uid = int(form['uid'].value) + except: + exitQuiet() + + # The user's key will be a 16 character string + key = form['key'].value + re.search('[&?]+', key) and exitQuiet() + len(key) == 16 or exitQuiet() + + # Historically facebook has been notoriously bad at setting timzeones + # in their stuff so this should be a user setting. If it is set in + # their calendar it'll be used otherwise if the user feeds crap or + # nothing just assume they want Australia/Perth + tz = "" + if "tz" in form: + from pytz import timezone + try: + timezone(form['tz'].value) + tz = form['tz'].value + except: pass + + ruleConfig["defaultTZ"] = tz or ruleConfig["defaultTZ"] + + # Okay, we're happy that the input is sane, lets serve up some data + url = 'https://www.facebook.com/events/ical/upcoming/?uid=%s&key=%s'%(uid,key) + (content, encoding) = getHTTPContent(url) + + cal = vobject.readOne(str(content, encoding)) + cal = applyRules(cal, generateRules(ruleConfig), False) + + print(('Content-Type: text/calendar; charset=%s\n'%encoding)) + writeOutput(cal) + + +if __name__ == '__main__': + # Ensure the rules process using the desired timezone + ruleConfig = {} + ruleConfig["defaultTZ"] = 'Australia/Perth' + + # Detect if this script has been called by CGI and proceed accordingly + if 'REQUEST_METHOD' in os.environ: + runCGI() + else: + runLocal()