1 #!/usr/bin/env python
2 # -*- coding: utf-8 -*-
4 # yocto-css, an extremely bare minimum CSS parser
5 #
6 # Copyright 2009 Jeff Schiller
7 #
8 # This file is part of Scour, http://www.codedread.com/scour/
9 #
10 # Licensed under the Apache License, Version 2.0 (the "License");
11 # you may not use this file except in compliance with the License.
12 # You may obtain a copy of the License at
13 #
14 # http://www.apache.org/licenses/LICENSE-2.0
15 #
16 # Unless required by applicable law or agreed to in writing, software
17 # distributed under the License is distributed on an "AS IS" BASIS,
18 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
19 # See the License for the specific language governing permissions and
20 # limitations under the License.
22 # In order to resolve Bug 368716 (https://bugs.launchpad.net/scour/+bug/368716)
23 # scour needed a bare-minimum CSS parser in order to determine if some elements
24 # were still referenced by CSS properties.
26 # I looked at css-py (a CSS parser built in Python), but that library
27 # is about 35k of Python and requires ply to be installed. I just need
28 # something very basic to suit scour's needs.
30 # yocto-css takes a string of CSS and tries to spit out a list of rules
31 # A rule is an associative array (dictionary) with the following keys:
32 # - selector: contains the string of the selector (see CSS grammar)
33 # - properties: contains an associative array of CSS properties for this rule
35 # TODO: need to build up some unit tests for yocto_css
37 # stylesheet : [ CDO | CDC | S | statement ]*;
38 # statement : ruleset | at-rule;
39 # at-rule : ATKEYWORD S* any* [ block | ';' S* ];
40 # block : '{' S* [ any | block | ATKEYWORD S* | ';' S* ]* '}' S*;
41 # ruleset : selector? '{' S* declaration? [ ';' S* declaration? ]* '}' S*;
42 # selector : any+;
43 # declaration : property S* ':' S* value;
44 # property : IDENT;
45 # value : [ any | block | ATKEYWORD S* ]+;
46 # any : [ IDENT | NUMBER | PERCENTAGE | DIMENSION | STRING
47 # | DELIM | URI | HASH | UNICODE-RANGE | INCLUDES
48 # | DASHMATCH | FUNCTION S* any* ')'
49 # | '(' S* any* ')' | '[' S* any* ']' ] S*;
51 def parseCssString(str):
52 rules = []
53 # first, split on } to get the rule chunks
54 chunks = str.split('}')
55 for chunk in chunks:
56 # second, split on { to get the selector and the list of properties
57 bits = chunk.split('{')
58 if len(bits) != 2: continue
59 rule = {}
60 rule['selector'] = bits[0].strip()
61 # third, split on ; to get the property declarations
62 bites = bits[1].strip().split(';')
63 if len(bites) < 1: continue
64 props = {}
65 for bite in bites:
66 # fourth, split on : to get the property name and value
67 nibbles = bite.strip().split(':')
68 if len(nibbles) != 2: continue
69 props[nibbles[0].strip()] = nibbles[1].strip()
70 rule['properties'] = props
71 rules.append(rule)
72 return rules