python/google/appengine/api/pagespeedinfo.py

   1 #!/usr/bin/env python
   2 #
   3 # Copyright 2007 Google Inc.
   4 #
   5 # Licensed under the Apache License, Version 2.0 (the "License");
   6 # you may not use this file except in compliance with the License.
   7 # You may obtain a copy of the License at
   8 #
   9 #     http://www.apache.org/licenses/LICENSE-2.0
  10 #
  11 # Unless required by applicable law or agreed to in writing, software
  12 # distributed under the License is distributed on an "AS IS" BASIS,
  13 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  14 # See the License for the specific language governing permissions and
  15 # limitations under the License.
  16 #
  17
  18
  19
  20
  21 """PageSpeed configuration tools.
  22
  23 Library for parsing pagespeed configuration data from app.yaml and working
  24 with these in memory.
  25 """
  26
  27
  28
  29
  30
  31
  32
  33 import google
  34
  35 from google.appengine.api import validation
  36 from google.appengine.api import yaml_builder
  37 from google.appengine.api import yaml_listener
  38 from google.appengine.api import yaml_object
  39
  40 _URL_BLACKLIST_REGEX = r'http(s)?://\S{0,499}'
  41 _REWRITER_NAME_REGEX = r'[a-zA-Z0-9_]+'
  42 _DOMAINS_TO_REWRITE_REGEX = r'(http(s)?://)?[-a-zA-Z0-9_.*]+(:\d+)?'
  43
  44 URL_BLACKLIST = 'url_blacklist'
  45 ENABLED_REWRITERS = 'enabled_rewriters'
  46 DISABLED_REWRITERS = 'disabled_rewriters'
  47 DOMAINS_TO_REWRITE = 'domains_to_rewrite'
  48
  49
  50 class MalformedPagespeedConfiguration(Exception):
  51   """Configuration file for PageSpeed API is malformed."""
  52
  53
  54
  55
  56
  57
  58 class PagespeedEntry(validation.Validated):
  59   """Describes the format of a pagespeed configuration from a yaml file.
  60
  61   URL blacklist entries are patterns (with '?' and '*' as wildcards).  Any URLs
  62   that match a pattern on the blacklist will not be optimized by PageSpeed.
  63
  64   Rewriter names are strings (like 'CombineCss' or 'RemoveComments') describing
  65   individual PageSpeed rewriters.  A full list of valid rewriter names can be
  66   found in the PageSpeed documentation.
  67
  68   The domains-to-rewrite list is a whitelist of domain name patterns with '*' as
  69   a wildcard, optionally starting with 'http://' or 'https://'.  If no protocol
  70   is given, 'http://' is assumed.  A resource will only be rewritten if it is on
  71   the same domain as the HTML that references it, or if its domain is on the
  72   domains-to-rewrite list.
  73   """
  74   ATTRIBUTES = {
  75       URL_BLACKLIST: validation.Optional(
  76           validation.Repeated(validation.Regex(_URL_BLACKLIST_REGEX))),
  77       ENABLED_REWRITERS: validation.Optional(
  78           validation.Repeated(validation.Regex(_REWRITER_NAME_REGEX))),
  79       DISABLED_REWRITERS: validation.Optional(
  80           validation.Repeated(validation.Regex(_REWRITER_NAME_REGEX))),
  81       DOMAINS_TO_REWRITE: validation.Optional(
  82           validation.Repeated(validation.Regex(_DOMAINS_TO_REWRITE_REGEX))),
  83   }
  84
  85
  86 def LoadPagespeedEntry(pagespeed_entry, open_fn=None):
  87   """Load a yaml file or string and return a PagespeedEntry.
  88
  89   Args:
  90     pagespeed_entry: The contents of a pagespeed entry from a yaml file
  91       as a string, or an open file object.
  92     open_fn: Function for opening files. Unused.
  93
  94   Returns:
  95     A PagespeedEntry instance which represents the contents of the parsed yaml.
  96
  97   Raises:
  98     yaml_errors.EventError: An error occured while parsing the yaml.
  99     MalformedPagespeedConfiguration: The configuration is parseable but invalid.
 100   """
 101   builder = yaml_object.ObjectBuilder(PagespeedEntry)
 102   handler = yaml_builder.BuilderHandler(builder)
 103   listener = yaml_listener.EventListener(handler)
 104   listener.Parse(pagespeed_entry)
 105
 106   parsed_yaml = handler.GetResults()
 107   if not parsed_yaml:
 108     return PagespeedEntry()
 109
 110   if len(parsed_yaml) > 1:
 111     raise MalformedPagespeedConfiguration(
 112         'Multiple configuration sections in the yaml')
 113
 114   return parsed_yaml[0]