Commit | Line | Data |
---|---|---|
68270f0f JRJ |
1 | /* |
2 | * Copyright (C) 2014 - Jonathan Rajotte <jonathan.r.julien@gmail.com> | |
3 | * | |
4 | * This program is free software; you can redistribute it and/or modify it | |
5 | * under the terms of the GNU General Public License, version 2 only, as | |
6 | * published by the Free Software Foundation. | |
7 | * | |
8 | * This program is distributed in the hope that it will be useful, but WITHOUT | |
9 | * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
10 | * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for | |
11 | * more details. | |
12 | * | |
13 | * You should have received a copy of the GNU General Public License along with | |
14 | * this program; if not, write to the Free Software Foundation, Inc., 51 | |
15 | */ | |
16 | ||
17 | /* | |
71f82f2b | 18 | * Usage: extract_xml [-v|-e] xml_path xpath_expression |
68270f0f JRJ |
19 | * Evaluate XPath expression and prints result node set. |
20 | * args[1] path to the xml file | |
21 | * args[2] xpath expression to extract | |
71f82f2b | 22 | * If -e look if node exist return "true" else nothing |
68270f0f JRJ |
23 | * If -v is set the name of the node will appear with his value delimited by |
24 | * a semicolon(;) | |
25 | * Ex: | |
26 | * Command:extract_xml ../file.xml /test/node/text() | |
27 | * Output: | |
28 | * a | |
29 | * b | |
30 | * c | |
31 | * With -v | |
32 | * node;a; | |
33 | * node;b; | |
34 | * node;c; | |
35 | */ | |
36 | #include <stdlib.h> | |
37 | #include <stdio.h> | |
38 | #include <string.h> | |
39 | #include <assert.h> | |
40 | #include <unistd.h> | |
41 | ||
42 | #include <libxml/tree.h> | |
43 | #include <libxml/parser.h> | |
44 | #include <libxml/xpath.h> | |
45 | #include <libxml/xpathInternals.h> | |
46 | ||
47 | #if defined(LIBXML_XPATH_ENABLED) | |
48 | ||
49 | ||
50 | int opt_verbose; | |
71f82f2b JR |
51 | int node_exist; |
52 | ||
68270f0f JRJ |
53 | /** |
54 | * print_xpath_nodes: | |
55 | * nodes: the nodes set. | |
56 | * output: the output file handle. | |
57 | * | |
58 | * Print the node content to the file | |
59 | */ | |
60 | static int print_xpath_nodes(xmlDocPtr doc, xmlNodeSetPtr nodes, FILE *output) | |
61 | { | |
71f82f2b | 62 | int ret = 0; |
68270f0f JRJ |
63 | int size; |
64 | int i; | |
65 | ||
66 | xmlNodePtr cur; | |
67 | xmlChar *node_child_value_string = NULL; | |
68 | ||
69 | assert(output); | |
70 | size = (nodes) ? nodes->nodeNr : 0; | |
71 | ||
72 | for (i = 0; i < size; ++i) { | |
73 | assert(nodes->nodeTab[i]); | |
74 | ||
75 | if (nodes->nodeTab[i]->type == XML_NAMESPACE_DECL) { | |
76 | fprintf(stderr, "ERR:%s\n", | |
77 | "This executable does not support xml namespacing\n"); | |
78 | ret = -1; | |
79 | goto end; | |
80 | } else if (nodes->nodeTab[i]->type == XML_ELEMENT_NODE) { | |
81 | cur = nodes->nodeTab[i]; | |
82 | ||
83 | if (xmlChildElementCount(cur) == 0) { | |
84 | if (xmlNodeIsText(cur->children)) { | |
85 | node_child_value_string = xmlNodeListGetString(doc, | |
86 | cur->children, 1); | |
71f82f2b JR |
87 | if (node_exist) { |
88 | fprintf(output, "true\n"); | |
89 | } else if (opt_verbose) { | |
68270f0f JRJ |
90 | fprintf(output, "%s;%s;\n", cur->name, |
91 | node_child_value_string); | |
92 | } else { | |
93 | fprintf(output, "%s\n", | |
94 | node_child_value_string); | |
95 | } | |
96 | xmlFree(node_child_value_string); | |
71f82f2b JR |
97 | } else { |
98 | /* We don't want to print non-final element */ | |
99 | if (node_exist) { | |
100 | fprintf(output, "true\n"); | |
101 | } else { | |
102 | fprintf(stderr, "ERR:%s\n", | |
103 | "Xpath expression return non-final xml element"); | |
104 | ret = -1; | |
105 | goto end; | |
106 | } | |
107 | } | |
108 | } else { | |
109 | if (node_exist) { | |
110 | fprintf(output, "true\n"); | |
68270f0f JRJ |
111 | } else { |
112 | /* We don't want to print non-final element */ | |
113 | fprintf(stderr, "ERR:%s\n", | |
114 | "Xpath expression return non-final xml element"); | |
115 | ret = -1; | |
116 | goto end; | |
117 | } | |
68270f0f JRJ |
118 | } |
119 | ||
120 | } else { | |
121 | cur = nodes->nodeTab[i]; | |
71f82f2b JR |
122 | if (node_exist) { |
123 | fprintf(output, "true\n"); | |
124 | } else if (opt_verbose) { | |
68270f0f JRJ |
125 | fprintf(output, "%s;%s;\n", cur->parent->name, cur->content); |
126 | } else { | |
127 | fprintf(output, "%s\n", cur->content); | |
68270f0f JRJ |
128 | } |
129 | } | |
130 | } | |
131 | /* Command Success */ | |
132 | ret = 0; | |
133 | ||
134 | end: | |
135 | return ret; | |
136 | } | |
137 | ||
138 | /* | |
139 | * Extract element corresponding to xpath | |
140 | * xml_path The path to the xml file | |
141 | * xpath: The xpath to evaluate. | |
142 | * | |
143 | * Evaluate an xpath expression onto an xml file. | |
144 | * and print the result one by line. | |
145 | * | |
146 | * Returns 0 on success and a negative value otherwise. | |
147 | */ | |
148 | static int extract_xpath(const char *xml_path, const xmlChar *xpath) | |
149 | { | |
150 | xmlDocPtr doc = NULL; | |
151 | xmlXPathContextPtr xpathCtx = NULL; | |
152 | xmlXPathObjectPtr xpathObj = NULL; | |
153 | ||
154 | assert(xml_path); | |
155 | assert(xpath); | |
156 | ||
157 | /* Parse the xml file */ | |
158 | doc = xmlParseFile(xml_path); | |
159 | if (!doc) { | |
160 | fprintf(stderr, "ERR parsing: xml file invalid \"%s\"\n", xml_path); | |
161 | return -1; | |
162 | } | |
163 | ||
164 | /* Initialize a xpath context */ | |
165 | xpathCtx = xmlXPathNewContext(doc); | |
166 | if (!xpathCtx) { | |
167 | fprintf(stderr, "ERR: XPath context invalid\n"); | |
168 | xmlFreeDoc(doc); | |
169 | return -1; | |
170 | } | |
171 | ||
172 | /* Evaluate xpath expression */ | |
173 | xpathObj = xmlXPathEvalExpression(xpath, xpathCtx); | |
174 | if (!xpathObj) { | |
175 | fprintf(stderr, "ERR: invalid xpath expression \"%s\"\n", xpath); | |
176 | xmlXPathFreeContext(xpathCtx); | |
177 | xmlFreeDoc(doc); | |
178 | return -1; | |
179 | } | |
180 | ||
181 | /* Print results */ | |
182 | if (print_xpath_nodes(doc, xpathObj->nodesetval, stdout)) { | |
183 | xmlXPathFreeObject(xpathObj); | |
184 | xmlXPathFreeContext(xpathCtx); | |
185 | xmlFreeDoc(doc); | |
186 | return -1; | |
187 | } | |
188 | ||
189 | /* Cleanup */ | |
190 | xmlXPathFreeObject(xpathObj); | |
191 | xmlXPathFreeContext(xpathCtx); | |
192 | xmlFreeDoc(doc); | |
193 | ||
194 | return 0; | |
195 | } | |
196 | ||
197 | int main(int argc, char **argv) | |
198 | { | |
199 | int opt; | |
200 | ||
201 | /* Parse command line and process file */ | |
71f82f2b | 202 | while ((opt = getopt(argc, argv, "ve")) != -1) { |
68270f0f JRJ |
203 | switch (opt) { |
204 | case 'v': | |
205 | opt_verbose = 1; | |
206 | break; | |
71f82f2b JR |
207 | case 'e': |
208 | node_exist = 1; | |
209 | break; | |
68270f0f JRJ |
210 | default: |
211 | abort(); | |
212 | } | |
213 | } | |
214 | ||
215 | if (!(optind + 1 < argc)) { | |
216 | fprintf(stderr, "ERR:%s\n", "Arguments missing"); | |
217 | return -1; | |
218 | } | |
219 | ||
220 | /* Init libxml */ | |
221 | xmlInitParser(); | |
222 | xmlKeepBlanksDefault(0); | |
223 | if (access(argv[optind], F_OK)) { | |
224 | fprintf(stderr, "ERR:%s\n", "Xml path not valid"); | |
225 | return -1; | |
226 | } | |
227 | /* Do the main job */ | |
228 | if (extract_xpath(argv[optind], (xmlChar *)argv[optind+1])) { | |
229 | return -1; | |
230 | } | |
231 | ||
232 | /* Shutdown libxml */ | |
233 | xmlCleanupParser(); | |
234 | ||
235 | return 0; | |
236 | } | |
237 | ||
238 | #else | |
239 | int main(void) | |
240 | { | |
241 | fprintf(stderr, "XPath support not compiled in\n"); | |
242 | return -1; | |
243 | } | |
244 | #endif |