1 // Copyright 2011 Google Inc. All Rights Reserved.
2 //
3 // Licensed under the Apache License, Version 2.0 (the "License");
4 // you may not use this file except in compliance with the License.
5 // You may obtain a copy of the License at
6 //
7 //     http://www.apache.org/licenses/LICENSE-2.0
8 //
9 // Unless required by applicable law or agreed to in writing, software
10 // distributed under the License is distributed on an "AS IS" BASIS,
11 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12 // See the License for the specific language governing permissions and
13 // limitations under the License.
14 //
15 // Author: jdtang@google.com (Jonathan Tang)
16 //
17 // This contains the list of states used in the tokenizer.  Although at first
18 // glance it seems like these could be kept internal to the tokenizer, several
19 // of the actions in the parser require that it reach into the tokenizer and
20 // reset the tokenizer state.  For that to work, it needs to have the
21 // definitions of individual states available.
22 //
23 // This may also be useful for providing more detailed error messages for parse
24 // errors, as we can match up states and inputs in a table without having to
25 // clutter the tokenizer code with lots of precise error messages.
26 
27 #ifndef GUMBO_TOKENIZER_STATES_H_
28 #define GUMBO_TOKENIZER_STATES_H_
29 
30 // The ordering of this enum is also used to build the dispatch table for the
31 // tokenizer state machine, so if it is changed, be sure to update that too.
32 typedef enum {
33   GUMBO_LEX_DATA,
34   GUMBO_LEX_CHAR_REF_IN_DATA,
35   GUMBO_LEX_RCDATA,
36   GUMBO_LEX_CHAR_REF_IN_RCDATA,
37   GUMBO_LEX_RAWTEXT,
38   GUMBO_LEX_SCRIPT,
39   GUMBO_LEX_PLAINTEXT,
40   GUMBO_LEX_TAG_OPEN,
41   GUMBO_LEX_END_TAG_OPEN,
42   GUMBO_LEX_TAG_NAME,
43   GUMBO_LEX_RCDATA_LT,
44   GUMBO_LEX_RCDATA_END_TAG_OPEN,
45   GUMBO_LEX_RCDATA_END_TAG_NAME,
46   GUMBO_LEX_RAWTEXT_LT,
47   GUMBO_LEX_RAWTEXT_END_TAG_OPEN,
48   GUMBO_LEX_RAWTEXT_END_TAG_NAME,
49   GUMBO_LEX_SCRIPT_LT,
50   GUMBO_LEX_SCRIPT_END_TAG_OPEN,
51   GUMBO_LEX_SCRIPT_END_TAG_NAME,
52   GUMBO_LEX_SCRIPT_ESCAPED_START,
53   GUMBO_LEX_SCRIPT_ESCAPED_START_DASH,
54   GUMBO_LEX_SCRIPT_ESCAPED,
55   GUMBO_LEX_SCRIPT_ESCAPED_DASH,
56   GUMBO_LEX_SCRIPT_ESCAPED_DASH_DASH,
57   GUMBO_LEX_SCRIPT_ESCAPED_LT,
58   GUMBO_LEX_SCRIPT_ESCAPED_END_TAG_OPEN,
59   GUMBO_LEX_SCRIPT_ESCAPED_END_TAG_NAME,
60   GUMBO_LEX_SCRIPT_DOUBLE_ESCAPED_START,
61   GUMBO_LEX_SCRIPT_DOUBLE_ESCAPED,
62   GUMBO_LEX_SCRIPT_DOUBLE_ESCAPED_DASH,
63   GUMBO_LEX_SCRIPT_DOUBLE_ESCAPED_DASH_DASH,
64   GUMBO_LEX_SCRIPT_DOUBLE_ESCAPED_LT,
65   GUMBO_LEX_SCRIPT_DOUBLE_ESCAPED_END,
66   GUMBO_LEX_BEFORE_ATTR_NAME,
67   GUMBO_LEX_ATTR_NAME,
68   GUMBO_LEX_AFTER_ATTR_NAME,
69   GUMBO_LEX_BEFORE_ATTR_VALUE,
70   GUMBO_LEX_ATTR_VALUE_DOUBLE_QUOTED,
71   GUMBO_LEX_ATTR_VALUE_SINGLE_QUOTED,
72   GUMBO_LEX_ATTR_VALUE_UNQUOTED,
73   GUMBO_LEX_CHAR_REF_IN_ATTR_VALUE,
74   GUMBO_LEX_AFTER_ATTR_VALUE_QUOTED,
75   GUMBO_LEX_SELF_CLOSING_START_TAG,
76   GUMBO_LEX_BOGUS_COMMENT,
77   GUMBO_LEX_MARKUP_DECLARATION,
78   GUMBO_LEX_COMMENT_START,
79   GUMBO_LEX_COMMENT_START_DASH,
80   GUMBO_LEX_COMMENT,
81   GUMBO_LEX_COMMENT_END_DASH,
82   GUMBO_LEX_COMMENT_END,
83   GUMBO_LEX_COMMENT_END_BANG,
84   GUMBO_LEX_DOCTYPE,
85   GUMBO_LEX_BEFORE_DOCTYPE_NAME,
86   GUMBO_LEX_DOCTYPE_NAME,
87   GUMBO_LEX_AFTER_DOCTYPE_NAME,
88   GUMBO_LEX_AFTER_DOCTYPE_PUBLIC_KEYWORD,
89   GUMBO_LEX_BEFORE_DOCTYPE_PUBLIC_ID,
90   GUMBO_LEX_DOCTYPE_PUBLIC_ID_DOUBLE_QUOTED,
91   GUMBO_LEX_DOCTYPE_PUBLIC_ID_SINGLE_QUOTED,
92   GUMBO_LEX_AFTER_DOCTYPE_PUBLIC_ID,
93   GUMBO_LEX_BETWEEN_DOCTYPE_PUBLIC_SYSTEM_ID,
94   GUMBO_LEX_AFTER_DOCTYPE_SYSTEM_KEYWORD,
95   GUMBO_LEX_BEFORE_DOCTYPE_SYSTEM_ID,
96   GUMBO_LEX_DOCTYPE_SYSTEM_ID_DOUBLE_QUOTED,
97   GUMBO_LEX_DOCTYPE_SYSTEM_ID_SINGLE_QUOTED,
98   GUMBO_LEX_AFTER_DOCTYPE_SYSTEM_ID,
99   GUMBO_LEX_BOGUS_DOCTYPE,
100   GUMBO_LEX_CDATA
101 } GumboTokenizerEnum;
102 
103 #endif  // GUMBO_TOKENIZER_STATES_H_
104