2 * Copyright (C) 2005-2009 Atsushi Konno All rights reserved.
3 * Copyright (C) 2005 QSDN,Inc. All rights reserved.
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
9 * http://www.apache.org/licenses/LICENSE-2.0
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.
17 #ifndef __QS_PARSE_STRING_H__
18 #define __QS_PARSE_STRING_H__
20 #include "chxj_apache.h"
21 #include "chxj_buffered_write.h"
29 * Max of memory allocation times.
31 #define QX_ALLOC_MAX (100*1024)
34 * It is judged whether it is the first byte of Japanese Shift_JIS
37 #define is_sjis_kanji(c) ((0x81 <= (unsigned char)(c&0xff) && (unsigned char)(c&0xff) <= 0x9f) \
38 || (0xe0 <= (unsigned char)(c&0xff) && (unsigned char)(c&0xff) <= 0xfc))
40 * It is judged whether it is a byte of Japanese Shift_JIS "HANKAKU KANA".
42 #define is_sjis_kana(c) ((0xa1 <= (unsigned char)(c&0xff) && (unsigned char)(c&0xff) <= 0xdf))
45 * It is judged whether it is a Japanese code part of EUC.
47 #define is_euc_kanji(c) (0xa1 <= (unsigned char)(c&0xff) && (unsigned char)(c&0xff) <= 0xfe)
50 * It is judged whether it is a Japanese code part of EUC "KANA".
52 #define is_euc_kana(c) (0x8e == (unsigned char)(c&0xff))
55 * True is returned for whitespace.
57 #define is_white_space(c) (' ' == (unsigned char)(c&0xff) \
58 || '\t' == (unsigned char)(c&0xff) \
59 || '\n' == (unsigned char)(c&0xff) \
60 || '\r' == (unsigned char)(c&0xff))
63 * True is returned for the quotation mark.
65 #define is_quote(c) ('\'' == (unsigned char)(c&0xff) \
66 ||'"' == (unsigned char)(c&0xff))
69 * It is judged whether the tag of the object has the child element.
70 * The "<option>" tag has the child. Please write </ option >.
72 #define has_child(c) ((strcasecmp(c, "base" ) != 0) \
73 && (strcasecmp(c, "meta" ) != 0) \
74 && (strcasecmp(c, "br" ) != 0) \
75 && (strcasecmp(c, "hr" ) != 0) \
76 && (strcasecmp(c, "img" ) != 0) \
77 && (strcasecmp(c, "input" ) != 0) \
78 && (strcasecmp(c, "?xml" ) != 0) \
79 && (strcasecmp(c, "!doctype" ) != 0) \
80 && (strcasecmp(c, "link" ) != 0) \
81 && (strcasecmp(c, "!--" ) != 0) \
82 && (strncasecmp(c, "![CDATA[",8) != 0))
85 * It is judged whether the tag of the object has the child element.
86 * The "<option>" tag has the child. Please write </ option >.
88 #define has_child_hdml(c) ((strcasecmp(c, "center" ) != 0) \
89 && (strcasecmp(c, "br" ) != 0) \
90 && (strcasecmp(c, "action" ) != 0) \
91 && (strcasecmp(c, "!--" ) != 0))
93 #define QS_PARSE_NL_MARK "CrLf"
96 * The structure of the attribute is defined.
98 typedef struct Attr Attr;
108 * The structure of the element is defined.
110 typedef struct Node Node;
117 struct Node *child_tail;
119 struct Attr *attr_tail;
125 int closed_by_itself;
129 typedef struct pointer_table_t {
130 unsigned int address;
132 struct pointer_table_t *next;
133 struct pointer_table_t *prev;
137 typedef enum chxj_parse_mode_t {
142 typedef struct _doc {
143 Node *now_parent_node;
147 unsigned long alloc_size;
149 Pointer_Table *pointer_table;
150 Pointer_Table *free_list_head;
151 Pointer_Table *free_list_tail;
152 Pointer_Table *allocated_list_head;
153 Pointer_Table *allocated_list_tail;
155 ParseMode_t parse_mode;
157 apr_allocator_t *allocator;
162 #ifndef __NON_MOD_CHXJ__
170 extern Node *qs_init_root_node(
173 extern void qs_add_child_node(
177 extern void qs_free_node(
181 extern Node* qs_get_root(
184 extern Node* qs_parse_string(
189 extern char* qs_get_node_value(
193 extern char* qs_get_node_name(
197 extern int qs_get_node_size(Doc* doc, Node* node) ;
199 extern Node *qs_get_child_node(Doc *doc, Node *node);
200 extern Node *qs_get_next_node(Doc *doc, Node *node);
201 extern Node *qs_get_prev_node(Doc *doc, Node *node);
203 Attr* qs_get_attr(Doc* doc, Node* node) ;
204 Attr* qs_get_next_attr(Doc* doc, Attr* attr) ;
205 char* qs_get_attr_name(Doc* doc, Attr* attr) ;
206 char* qs_get_attr_value(Doc* doc, Attr* attr) ;
208 extern int chxj_cut_tag(const char *s, int len);