Mercurial > hgrepos > Python2 > PyMuPDF

comparison mupdf-source/thirdparty/extract/src/xml.h @ 2:b50eed0cc0ef upstream

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
ADD: MuPDF v1.26.7: the MuPDF source as downloaded by a default build of PyMuPDF 1.26.4. The directory name has changed: no version number in the expanded directory now.
author Franz Glasner <fzglas.hg@dom66.de>
date Mon, 15 Sep 2025 11:43:07 +0200
parents
children
comparison
equal deleted inserted replaced
1:1d09e1dec1d9 2:b50eed0cc0ef
1 #ifndef ARTIFEX_EXTRACT_XML
2 #define ARTIFEX_EXTRACT_XML
3
4 /* Only for internal use by extract code. */
5
6 #include "extract/buffer.h"
7
8 #include "astring.h"
9
10
11 /* Things for representing XML. */
12
13 typedef struct {
14 char *name;
15 char *value;
16 } extract_xml_attribute_t;
17
18 /* Represents a single <...> XML tag plus trailing text. */
19 typedef struct {
20 char *name;
21 extract_xml_attribute_t *attributes;
22 int attributes_num;
23 extract_astring_t text;
24 } extract_xml_tag_t;
25
26
27 /* Initialises tag. Will cause leak if tag contains data - in this case call
28 extract_xml_tag_free(). */
29 void extract_xml_tag_init(extract_xml_tag_t *tag);
30
31 /* Frees tag and then calls extract_xml_tag_init(). */
32 void extract_xml_tag_free(extract_alloc_t *alloc, extract_xml_tag_t *tag);
33
34
35 /* extract_xml_pparse_*(): simple XML 'pull' parser.
36
37 If <first_line> is not NULL, we require that <buffer> starts with the specified
38 text. Usually one would include a final newline in <first_line>.
39
40 extract_xml_pparse_init() merely consumes the initial '<'. Thereafter
41 extract_xml_pparse_next() consumes the next '<' before returning the previous
42 tag. */
43
44 /* Opens specified file.
45
46 If first_line is not NULL, we check that it matches the first line in the file.
47
48 Returns -1 with errno=ESRCH if we fail to read the first '<' due to EOF.
49 */
50 int extract_xml_pparse_init(extract_alloc_t *alloc, extract_buffer_t *buffer, const char *first_line);
51
52
53
54 /* Returns the next XML tag.
55
56 Returns 0 with *out containing next tag; or -1 with errno set if error; or +1
57 with errno=ESRCH if EOF.
58
59 If we return 0, we guarantee that out->name points to valid string and that
60 each item in out->attributes has similarly valid name and value members.
61
62 *out is initially passed to extract_xml_tag_free(), so *out must have been
63 initialised, e.g. by by extract_xml_tag_init(). */
64 int extract_xml_pparse_next(extract_buffer_t *buffer, extract_xml_tag_t *out);
65
66
67 /* Returns pointer to value of specified attribute, or NULL if not found. */
68 char *extract_xml_tag_attributes_find(extract_xml_tag_t *tag, const char *name);
69
70 /* Finds float value of specified attribute, returning error if not found or
71 there is trailing text. */
72 int extract_xml_tag_attributes_find_float(
73 extract_xml_tag_t *tag,
74 const char *name,
75 float *o_out);
76
77 /* Finds double value of specified attribute, returning error if not found or there is
78 trailing text. */
79 int extract_xml_tag_attributes_find_double(
80 extract_xml_tag_t *tag,
81 const char *name,
82 double *o_out);
83
84
85 /* Next few functions write to out-param and return zero on success, else
86 return -1 with errno set.
87
88 An error is returned if value is out of range or there is any trailing text. */
89
90 int extract_xml_str_to_llint(const char *text, long long *o_out);
91
92 int extract_xml_str_to_ullint(const char *text, unsigned long long *o_out);
93
94 int extract_xml_str_to_int(const char *text, int *o_out);
95
96 int extract_xml_str_to_uint(const char *text, unsigned *o_out);
97
98 int extract_xml_str_to_size(const char *text, size_t *o_out);
99
100 int extract_xml_str_to_double(const char *text, double *o_out);
101
102 int extract_xml_str_to_float(const char *text, float *o_out);
103
104
105 /* Finds int value of specified attribute, returning error if not found. */
106 int extract_xml_tag_attributes_find_int(
107 extract_xml_tag_t *tag,
108 const char *name,
109 int *o_out);
110
111 /* Finds unsigned int value of specified attribute, returning error if not
112 found. */
113 int extract_xml_tag_attributes_find_uint(
114 extract_xml_tag_t *tag,
115 const char *name,
116 unsigned *o_out);
117
118 /* Finds unsigned int value of specified attribute, returning error if not
119 found. */
120 int extract_xml_tag_attributes_find_size(
121 extract_xml_tag_t *tag,
122 const char *name,
123 size_t *o_out);
124
125 #endif