X-Git-Url: https://git.ucc.asn.au/?p=frenchie%2Ficalparse.git;a=blobdiff_plain;f=icalparse.py;h=3907da75547cd90c2ad19a2f4f0865423bcf408d;hp=7953420bfdc303c8df466339fe74096a03b39e72;hb=6913e5571252b2ebdaec7f21cf139f17cde24119;hpb=9edbffa51163eef8bac3d125a7cd7320a92df3a8 diff --git a/icalparse.py b/icalparse.py index 7953420..3907da7 100755 --- a/icalparse.py +++ b/icalparse.py @@ -1,6 +1,6 @@ #!/usr/bin/python # -# Copyright (c) 2010 James French +# Copyright (c) 2011 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 @@ -21,120 +21,182 @@ # THE SOFTWARE. import sys -import re import urlparse import os -from optparse import OptionParser - -class InvalidICS(Exception): pass -class notJoined(Exception): pass - -# RFC5545 and RFC5546 iCalendar registries contain upper case letters -# and dashes only and are separated from the value by a colon (:) -icalEntry = re.compile('^[A-Z\-]+:.*') - -def lineJoiner(oldcal): - '''Unfolds a calendar so that items can be parsed''' - - cal = [] - - # Strip newlines ( - for line in oldcal: - line = line.rstrip('\r\n') - - # Reassemble broken Lines - if not line: - if not cal: continue - else: cal[-1] += '\\n' - elif line[0] == ' ': - if not cal: raise InvalidICS, 'First line of ICS must be element' - line = line[1:len(line)] - cal[-1] += line - elif not icalEntry.match(line): - if not cal: raise InvalidICS, 'First line of ICS must be element' - cal[-1] += '\\n' + line - else: - if cal: cal[-1] += '\r\n' - cal.append(line) +import vobject +from cgi import parse_header - cal[-1] += '\r\n' - return cal +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). + ''' -def lineSplitter(oldcal, length=75): - '''Folds content lines to a specified length, returns a list''' + encoding = '' # If we don't populate this, the script will assume UTF-8 - cal = [] - sl = length - 1 + # 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) + if 'http' in parsedURL[0]: return getHTTPContent(url) - for line in oldcal: - # Line & line ending line ending fit inside length, do nothing - if len(line.rstrip()) <= length: - cal.append(line) - else: - brokenline = [line[0:length] + '\r\n'] - ll = length - while ll < len(line.rstrip('\r\n')) + 1: - brokenline.append(' ' + line[ll:sl+ll].rstrip('\r\n') + '\r\n') - ll += sl - cal += brokenline + if stdin: + content = sys.stdin.read() + return (content, encoding) - return cal + if not parsedURL[0]: url = 'file://' + os.path.abspath(url) -if __name__ == '__main__': - # If the user passed us a 'stdin' argument, we'll go with that, - # otherwise we'll try for a url opener + # If we've survived, use python's generic URL opening library to handle it + import urllib2 + try: + res = urllib2.urlopen(url) + content = res.read() + ct = res.info().getplist() + res.close() + except (urllib2.URLError, OSError), e: + sys.stderr.write('%s\n'%e) + sys.exit(1) - parser = OptionParser('usage: %prog [options] url') - parser.add_option('-s', '--stdin', action='store_true', dest='stdin', - default=False, help='Take a calendar from standard input') - parser.add_option('-o', '--output', dest='outfile', default='', - help='Specify output file (defaults to standard output)') + for param in ct: + if 'charset' in param: + encoding = param.split('=')[1] + break - (options, args) = parser.parse_args() + return (content, encoding) - if not args and not options.stdin: - parser.print_usage() - sys.exit(0) - url = args[0] +def getHTTPContent(url='',cache='.httplib2-cache'): + '''This function attempts to play nice when retrieving content from HTTP + services. It's what you should use in a CGI script.''' - # Work out what url parsers we're going to need based on what the user - # gave us on the command line - we do like files after all - parsedURL = urlparse.urlparse(url) - http = 'http' in parsedURL[0] + try: + import httplib2 + except ImportError: + import urllib2 - if not parsedURL[0]: u = False - else: u = True + if not url: return ('','') - if not options.stdin and http: - try: - import httplib2 - except ImportError: - import urllib2 + if not 'http' in urlparse.urlparse(url)[0]: return ('','') - # Try and play nice with HTTP servers unless something goes wrong. We don't - # really care about this cache so it can be somewhere volatile - h = False if 'httplib2' in sys.modules: try: h = httplib2.Http('.httplib2-cache') except OSError: h = httplib2.Http() + else: h = False - if not options.stdin and (not http or not 'httplib2' in sys.modules): - import urllib2 + if h: + try: + req = h.request(url) + except ValueError, e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + + resp, content = req + if 'content-type' in resp: + ct = 'Content-Type: %s'%req[0]['content-type'] + ct = parse_header(ct) + if 'charset' in ct[1]: encoding = ct[1]['charset'] + else: encoding = '' + else: + ct = '' + encoding = '' + + else: + try: + req = urllib2.urlopen(url) + except urllib2.URLError, e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + + content = req.read() + ct = req.info().getplist() + for param in ct: + if 'charset' in param: + encoding = param.split('=')[1] + break + + return (content, encoding) + +def generateRules(ruleConfig): + '''Attempts to load a series of rules into a list''' try: - content = u and (h and h.request(url)[1] or urllib2.urlopen(url).read()) - except (ValueError, urllib2.URLError), e: - sys.stderr.write('%s\n'%e) - sys.exit(1) + 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(cal, rules=[], verbose=False): + 'Runs a series of rules on the lines in ical and mangles its output' + + for rule in rules: + cal = rule(cal) + + return cal + +def writeOutput(cal, outfile=''): + '''Takes a list of lines and outputs to the specified file''' - if not u: - try: content = open(os.path.abspath(url),'r').read() + if not cal: + sys.stderr.write('Refusing to write out an empty file') + sys.exit(0) + + if not outfile: + out = sys.stdout + else: + try: + out = open(outfile, 'w') except (IOError, OSError), e: sys.stderr.write('%s\n'%e) sys.exit(1) - # RFC5545 and RFC5546 Calendars should be generated UTF-8 and we need to - # be able to read ANSI as well. This should take care of us. - content = unicode(content, encoding='utf-8') + cal.serialize(out) + + if not out == sys.stdout: + out.close() + +if __name__ == '__main__': + # Only load options parsing if this script was called directly, skip it + # if it's being called as a module. + from optparse import OptionParser + + parser = OptionParser('usage: %prog [options] url') + parser.add_option('-s', '--stdin', action='store_true', dest='stdin', + default=False, help='Take a calendar from standard input') + parser.add_option('-v', '--verbose', action='store_true', dest='verbose', + default=False, help='Be verbose when rules are being applied') + parser.add_option('-o', '--output', dest='outfile', default='', + help='Specify output file (defaults to standard output)') + 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('-t','--timezone', dest='timezone', default='Australia/Perth', + help='Specify a timezone to use if the remote calendar doesn\'t set it properly') + + (options, args) = parser.parse_args() + + # Ensure the rules process using the desired timezone + ruleConfig = {} + ruleConfig["defaultTZ"] = options.timezone + + # 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) + elif not options.stdin: + url = args[0] + else: + url = '' + + (content, encoding) = getContent(url, options.stdin) + encoding = encoding or options.encoding or 'utf-8' + + cal = vobject.readOne(unicode(content, encoding)) + cal = applyRules(cal, generateRules(ruleConfig), options.verbose) + + writeOutput(cal, options.outfile)