xmlfile.c 7.7 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286
  1. /*
  2. __ __ _
  3. ___\ \/ /_ __ __ _| |_
  4. / _ \\ /| '_ \ / _` | __|
  5. | __// \| |_) | (_| | |_
  6. \___/_/\_\ .__/ \__,_|\__|
  7. |_| XML parser
  8. Copyright (c) 1997-2000 Thai Open Source Software Center Ltd
  9. Copyright (c) 2000-2017 Expat development team
  10. Licensed under the MIT license:
  11. Permission is hereby granted, free of charge, to any person obtaining
  12. a copy of this software and associated documentation files (the
  13. "Software"), to deal in the Software without restriction, including
  14. without limitation the rights to use, copy, modify, merge, publish,
  15. distribute, sublicense, and/or sell copies of the Software, and to permit
  16. persons to whom the Software is furnished to do so, subject to the
  17. following conditions:
  18. The above copyright notice and this permission notice shall be included
  19. in all copies or substantial portions of the Software.
  20. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
  21. EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
  22. MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN
  23. NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
  24. DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
  25. OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
  26. USE OR OTHER DEALINGS IN THE SOFTWARE.
  27. */
  28. #include <stdio.h>
  29. #include <stdlib.h>
  30. #include <stddef.h>
  31. #include <string.h>
  32. #include <fcntl.h>
  33. #ifdef _WIN32
  34. #include "winconfig.h"
  35. #elif defined(HAVE_EXPAT_CONFIG_H)
  36. #include <expat_config.h>
  37. #endif /* ndef _WIN32 */
  38. #include "expat.h"
  39. #include "internal.h" /* for UNUSED_P only */
  40. #include "xmlfile.h"
  41. #include "xmltchar.h"
  42. #include "filemap.h"
  43. #if defined(_MSC_VER)
  44. #include <io.h>
  45. #endif
  46. #ifdef HAVE_UNISTD_H
  47. #include <unistd.h>
  48. #endif
  49. #ifndef O_BINARY
  50. #ifdef _O_BINARY
  51. #define O_BINARY _O_BINARY
  52. #else
  53. #define O_BINARY 0
  54. #endif
  55. #endif
  56. #ifdef _DEBUG
  57. #define READ_SIZE 16
  58. #else
  59. #define READ_SIZE (1024*8)
  60. #endif
  61. typedef struct {
  62. XML_Parser parser;
  63. int *retPtr;
  64. } PROCESS_ARGS;
  65. static int
  66. processStream(const XML_Char *filename, XML_Parser parser);
  67. static void
  68. reportError(XML_Parser parser, const XML_Char *filename)
  69. {
  70. enum XML_Error code = XML_GetErrorCode(parser);
  71. const XML_Char *message = XML_ErrorString(code);
  72. if (message)
  73. ftprintf(stdout, T("%s:%" XML_FMT_INT_MOD "u:%" XML_FMT_INT_MOD "u: %s\n"),
  74. filename,
  75. XML_GetErrorLineNumber(parser),
  76. XML_GetErrorColumnNumber(parser),
  77. message);
  78. else
  79. ftprintf(stderr, T("%s: (unknown message %d)\n"), filename, code);
  80. }
  81. /* This implementation will give problems on files larger than INT_MAX. */
  82. static void
  83. processFile(const void *data, size_t size,
  84. const XML_Char *filename, void *args)
  85. {
  86. XML_Parser parser = ((PROCESS_ARGS *)args)->parser;
  87. int *retPtr = ((PROCESS_ARGS *)args)->retPtr;
  88. if (XML_Parse(parser, (const char *)data, (int)size, 1) == XML_STATUS_ERROR) {
  89. reportError(parser, filename);
  90. *retPtr = 0;
  91. }
  92. else
  93. *retPtr = 1;
  94. }
  95. #if defined(_WIN32)
  96. static int
  97. isAsciiLetter(XML_Char c)
  98. {
  99. return (T('a') <= c && c <= T('z')) || (T('A') <= c && c <= T('Z'));
  100. }
  101. #endif /* _WIN32 */
  102. static const XML_Char *
  103. resolveSystemId(const XML_Char *base, const XML_Char *systemId,
  104. XML_Char **toFree)
  105. {
  106. XML_Char *s;
  107. *toFree = 0;
  108. if (!base
  109. || *systemId == T('/')
  110. #if defined(_WIN32)
  111. || *systemId == T('\\')
  112. || (isAsciiLetter(systemId[0]) && systemId[1] == T(':'))
  113. #endif
  114. )
  115. return systemId;
  116. *toFree = (XML_Char *)malloc((tcslen(base) + tcslen(systemId) + 2)
  117. * sizeof(XML_Char));
  118. if (!*toFree)
  119. return systemId;
  120. tcscpy(*toFree, base);
  121. s = *toFree;
  122. if (tcsrchr(s, T('/')))
  123. s = tcsrchr(s, T('/')) + 1;
  124. #if defined(_WIN32)
  125. if (tcsrchr(s, T('\\')))
  126. s = tcsrchr(s, T('\\')) + 1;
  127. #endif
  128. tcscpy(s, systemId);
  129. return *toFree;
  130. }
  131. static int
  132. externalEntityRefFilemap(XML_Parser parser,
  133. const XML_Char *context,
  134. const XML_Char *base,
  135. const XML_Char *systemId,
  136. const XML_Char *UNUSED_P(publicId))
  137. {
  138. int result;
  139. XML_Char *s;
  140. const XML_Char *filename;
  141. XML_Parser entParser = XML_ExternalEntityParserCreate(parser, context, 0);
  142. int filemapRes;
  143. PROCESS_ARGS args;
  144. args.retPtr = &result;
  145. args.parser = entParser;
  146. filename = resolveSystemId(base, systemId, &s);
  147. XML_SetBase(entParser, filename);
  148. filemapRes = filemap(filename, processFile, &args);
  149. switch (filemapRes) {
  150. case 0:
  151. result = 0;
  152. break;
  153. case 2:
  154. ftprintf(stderr, T("%s: file too large for memory-mapping")
  155. T(", switching to streaming\n"), filename);
  156. result = processStream(filename, entParser);
  157. break;
  158. }
  159. free(s);
  160. XML_ParserFree(entParser);
  161. return result;
  162. }
  163. static int
  164. processStream(const XML_Char *filename, XML_Parser parser)
  165. {
  166. /* passing NULL for filename means read intput from stdin */
  167. int fd = 0; /* 0 is the fileno for stdin */
  168. if (filename != NULL) {
  169. fd = topen(filename, O_BINARY|O_RDONLY);
  170. if (fd < 0) {
  171. tperror(filename);
  172. return 0;
  173. }
  174. }
  175. for (;;) {
  176. int nread;
  177. char *buf = (char *)XML_GetBuffer(parser, READ_SIZE);
  178. if (!buf) {
  179. if (filename != NULL)
  180. close(fd);
  181. ftprintf(stderr, T("%s: out of memory\n"),
  182. filename != NULL ? filename : "xmlwf");
  183. return 0;
  184. }
  185. nread = read(fd, buf, READ_SIZE);
  186. if (nread < 0) {
  187. tperror(filename != NULL ? filename : "STDIN");
  188. if (filename != NULL)
  189. close(fd);
  190. return 0;
  191. }
  192. if (XML_ParseBuffer(parser, nread, nread == 0) == XML_STATUS_ERROR) {
  193. reportError(parser, filename != NULL ? filename : "STDIN");
  194. if (filename != NULL)
  195. close(fd);
  196. return 0;
  197. }
  198. if (nread == 0) {
  199. if (filename != NULL)
  200. close(fd);
  201. break;;
  202. }
  203. }
  204. return 1;
  205. }
  206. static int
  207. externalEntityRefStream(XML_Parser parser,
  208. const XML_Char *context,
  209. const XML_Char *base,
  210. const XML_Char *systemId,
  211. const XML_Char *UNUSED_P(publicId))
  212. {
  213. XML_Char *s;
  214. const XML_Char *filename;
  215. int ret;
  216. XML_Parser entParser = XML_ExternalEntityParserCreate(parser, context, 0);
  217. filename = resolveSystemId(base, systemId, &s);
  218. XML_SetBase(entParser, filename);
  219. ret = processStream(filename, entParser);
  220. free(s);
  221. XML_ParserFree(entParser);
  222. return ret;
  223. }
  224. int
  225. XML_ProcessFile(XML_Parser parser,
  226. const XML_Char *filename,
  227. unsigned flags)
  228. {
  229. int result;
  230. if (!XML_SetBase(parser, filename)) {
  231. ftprintf(stderr, T("%s: out of memory"), filename);
  232. exit(1);
  233. }
  234. if (flags & XML_EXTERNAL_ENTITIES)
  235. XML_SetExternalEntityRefHandler(parser,
  236. (flags & XML_MAP_FILE)
  237. ? externalEntityRefFilemap
  238. : externalEntityRefStream);
  239. if (flags & XML_MAP_FILE) {
  240. int filemapRes;
  241. PROCESS_ARGS args;
  242. args.retPtr = &result;
  243. args.parser = parser;
  244. filemapRes = filemap(filename, processFile, &args);
  245. switch (filemapRes) {
  246. case 0:
  247. result = 0;
  248. break;
  249. case 2:
  250. ftprintf(stderr, T("%s: file too large for memory-mapping")
  251. T(", switching to streaming\n"), filename);
  252. result = processStream(filename, parser);
  253. break;
  254. }
  255. }
  256. else
  257. result = processStream(filename, parser);
  258. return result;
  259. }