X-Git-Url: https://git.ucc.asn.au/?a=blobdiff_plain;f=icalparse.py;h=293295c62dcf5da892eee740ce01b9088eede732;hb=dd8076b16e4c04a510dc8fe8fe8655f5e350014f;hp=110a19ae703d822afa6fd918672bcf69c158543e;hpb=aace9ea8b2dbcdbdc1b3ed30a672a52b0a619e3f;p=frenchie%2Ficalparse.git diff --git a/icalparse.py b/icalparse.py index 110a19a..293295c 100755 --- a/icalparse.py +++ b/icalparse.py @@ -24,19 +24,32 @@ import sys import urlparse import os + class InvalidICS(Exception): pass -class notJoined(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) - oldcal.replace('\r\n ', '') return oldcal.split('\r\n') @@ -49,25 +62,60 @@ def lineFolder(oldcal, length=75): cal = [] sl = length - 1 - for line in oldcal: + for uline in oldcal: + line = uline.encode('utf-8') + # Line fits inside length, do nothing - if len(line.rstrip()) <= length: + 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 + +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). ''' + 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) @@ -75,60 +123,137 @@ def getContent(url='',stdin=False): if stdin: content = sys.stdin.read() - return content + 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 + 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) content = res.read() + if 'content-type' in res.info(): ct = res.info()['content-type'] + else: ct = '' res.close() - except (urllib2.URLError, ValueError), e: + except (urllib2.URLError, OSError), e: sys.stderr.write('%s\n'%e) sys.exit(1) - return content + + encoding = 'charset' in ct and ct.split(';')[-1].lower().split('=')[-1].strip() or '' + 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. It will (by default) - slurp the first 20 bytes of the file and check that we are indeed looking - at an ICS file before going for broke.''' + services. It's what you should use in a CGI script.''' try: import httplib2 except ImportError: import urllib2 - if not url: return '' + if not url: return ('','') + if not 'http' in parsedURL[0]: return ('','') if 'httplib2' in sys.modules: try: h = httplib2.Http('.httplib2-cache') except OSError: h = httplib2.Http() else: h = False - try: - if h: content = h.request(url)[1] - return content - except ValueError, e: - sys.stderr.write('%s\n'%e) - sys.exit(1) + if h: + try: + req = h.request(url) + except ValueError, e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + + content = req[1] + if 'content-type' in req[0]: ct = req[0]['content-type'] + + else: + try: + req = urllib2.urlopen(url) + except urllib2.URLError, e: + sys.stderr.write('%s\n'%e) + sys.exit(1) + + content = req.read() + info = req.info() + + ct = info['content-type'] + + encoding = 'charset' in ct and ct.split(';')[-1].lower().split('=')[-1].strip() or '' + return (content, encoding) + + +def generateRules(): + '''Attempts to load a series of rules into a list''' try: - content = urllib2.urlopen(url).read() - return content - except urllib2.URLError, e: - sys.stderr.write('%s\n'%e) - sys.exit(1) + 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 cal[-1]: cal.append('') + + out.write('\r\n'.join(cal)) + + if not out == sys.stdout: + out.close() - return '' if __name__ == '__main__': from optparse import OptionParser @@ -138,8 +263,13 @@ if __name__ == '__main__': 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)') (options, args) = parser.parse_args() @@ -151,6 +281,9 @@ if __name__ == '__main__': else: url = '' - content = getContent(url, options.stdin) - cal = lineJoiner(content) - print cal + (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)