X-Git-Url: https://git.ucc.asn.au/?a=blobdiff_plain;f=icalparse.py;h=6c3aac67ff4739d01b544462ed279aa3dc805689;hb=18fa3d2d1ba2875bf7c15bd36862fa7368e337c0;hp=3b9091a45f05ceca1e5630f8b06ef64d94814fab;hpb=f1871b1e6866c901ffdfc6e086f25ce786bf0c9b;p=frenchie%2Ficalparse.git diff --git a/icalparse.py b/icalparse.py index 3b9091a..6c3aac6 100755 --- a/icalparse.py +++ b/icalparse.py @@ -21,113 +21,268 @@ # THE SOFTWARE. import sys -import re import urlparse import os -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\-]+:.*') +class InvalidICS(Exception): pass +class IncompleteICS(InvalidICS): pass -def lineJoiner(oldcal): +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) - # RFC2445 This sequence defines a content 'fold' and needs to be stripped - # from the output before writing the file - oldcal.replace('\r\n ', '') 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 line in oldcal: - # Line & line ending line ending fit inside length, do nothing - if len(line.rstrip()) <= length: + for uline in oldcal: + line = uline.encode('utf-8') + + # Line fits inside length, do nothing + if len(line) <= 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 + 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 -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', - 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)') +def splitFields(cal): + '''Takes a list of lines in a calendar file and returns a list of tuples + as (key, value) pairs''' - (options, args) = parser.parse_args() + 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] - if not args and not options.stdin: - parser.print_usage() - sys.exit(0) - elif not options.stdin: - url = args[0] - else: - url = '' - # 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 +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). + ''' + + encoding = '' # If we don't populate this, the script will assume UTF-8 + + # 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) - http = 'http' in parsedURL[0] + if 'http' in parsedURL[0]: return getHTTPContent(url) - if not parsedURL[0]: u = False - else: u = True + if stdin: + content = sys.stdin.read() + return (content, encoding) + + if not parsedURL[0]: + try: content = open(os.path.abspath(url),'r').read() + except (IOError, OSError), e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + return (content, encoding) + + # If we've survived, use python's generic URL opening library to handle it + import urllib2 + try: + res = urllib2.urlopen(url) + content = res.read() + res.close() + except (urllib2.URLError, OSError), e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + return (content, encoding) + + +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.''' + + try: + import httplib2 + except ImportError: + import urllib2 + + if not url: return '' + + encoding = '' # If we don't populate this, the script will assume UTF-8 - if not options.stdin and http: - try: - import httplib2 - except ImportError: - import urllib2 - - # Try and play nice with HTTP servers unless something goes wrong. We don't - # really care about this cache (A lot of ics files seem to be generated with - # php which hates caching with a passion). - h = False if 'httplib2' in sys.modules: try: h = httplib2.Http('.httplib2-cache') except OSError: h = httplib2.Http() + else: h = False - # Load urllib2 if this is not a stdin - if not options.stdin and (not http or not 'httplib2' in sys.modules): - import urllib2 + try: + if h: + req = h.request(url) + content = req[1] + if 'content-type' in req[0]: + for ct in req[0]['content-type'].split(';'): + ct = ct.lower() + if 'charset' in ct: + encoding = ct.split('=')[1].strip() + return (content, encoding) + except ValueError, e: + sys.stderr.write('%s\n'%e) + sys.exit(1) try: - content = u and (h and h.request(url)[1] or urllib2.urlopen(url).read()) - except (ValueError, urllib2.URLError), e: + content = urllib2.urlopen(url).read() + return (content, encoding) + except (urllib2.URLError, OSError), e: sys.stderr.write('%s\n'%e) sys.exit(1) - if not u and not options.stdin: - try: content = open(os.path.abspath(url),'r').read() + return ('', '') + + +def generateRules(): + '''Attempts to load a series of rules into a list''' + try: + import parserrules + except ImportError: + return [] + + rules = [getattr(parserrules, rule) for rule in dir(parserrules) if callable(getattr(parserrules, rule))] + return rules + + +def applyRules(ical, 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 + + +def writeOutput(cal, outfile=''): + '''Takes a list of lines and outputs to the specified file''' + + 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) - if options.stdin: - content = sys.stdin.read() + if cal[-1]: cal.append('') + + out.write('\r\n'.join(cal)) + + if not out == sys.stdout: + out.close() + + +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', + 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)') - # RFC5545 and RFC5546 New 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') + (options, args) = parser.parse_args() + + if not args and not options.stdin: + parser.print_usage() + sys.exit(0) + elif not options.stdin: + url = args[0] + else: + url = '' - #return content + (content, encoding) = getContent(url, options.stdin) + encoding = encoding or options.encoding or 'utf-8' + cal = lineJoiner(content, encoding) + ical = applyRules(splitFields(cal), generateRules(), options.verbose) + output = lineFolder(joinFields(ical)) + writeOutput(output, options.outfile)