xmlfile.c 8.1 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276
  1. /*
  2. __ __ _
  3. ___\ \/ /_ __ __ _| |_
  4. / _ \\ /| '_ \ / _` | __|
  5. | __// \| |_) | (_| | |_
  6. \___/_/\_\ .__/ \__,_|\__|
  7. |_| XML parser
  8. Copyright (c) 1997-2000 Thai Open Source Software Center Ltd
  9. Copyright (c) 2000 Clark Cooper <[email protected]>
  10. Copyright (c) 2002-2003 Fred L. Drake, Jr. <[email protected]>
  11. Copyright (c) 2004-2006 Karl Waclawek <[email protected]>
  12. Copyright (c) 2005-2007 Steven Solie <[email protected]>
  13. Copyright (c) 2016-2023 Sebastian Pipping <[email protected]>
  14. Copyright (c) 2017 Rhodri James <[email protected]>
  15. Copyright (c) 2019 David Loffredo <[email protected]>
  16. Copyright (c) 2021 Donghee Na <[email protected]>
  17. Licensed under the MIT license:
  18. Permission is hereby granted, free of charge, to any person obtaining
  19. a copy of this software and associated documentation files (the
  20. "Software"), to deal in the Software without restriction, including
  21. without limitation the rights to use, copy, modify, merge, publish,
  22. distribute, sublicense, and/or sell copies of the Software, and to permit
  23. persons to whom the Software is furnished to do so, subject to the
  24. following conditions:
  25. The above copyright notice and this permission notice shall be included
  26. in all copies or substantial portions of the Software.
  27. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
  28. EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
  29. MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN
  30. NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
  31. DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
  32. OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
  33. USE OR OTHER DEALINGS IN THE SOFTWARE.
  34. */
  35. #include "expat_config.h"
  36. #include <stdio.h>
  37. #include <stdlib.h>
  38. #include <stddef.h>
  39. #include <string.h>
  40. #include <fcntl.h>
  41. #ifdef _WIN32
  42. # include "winconfig.h"
  43. #endif
  44. #include "expat.h"
  45. #include "internal.h" /* for UNUSED_P only */
  46. #include "xmlfile.h"
  47. #include "xmltchar.h"
  48. #include "filemap.h"
  49. #if defined(_MSC_VER)
  50. # include <io.h>
  51. #endif
  52. #ifdef HAVE_UNISTD_H
  53. # include <unistd.h>
  54. #endif
  55. #ifndef O_BINARY
  56. # ifdef _O_BINARY
  57. # define O_BINARY _O_BINARY
  58. # else
  59. # define O_BINARY 0
  60. # endif
  61. #endif
  62. int g_read_size_bytes = 1024 * 8;
  63. typedef struct {
  64. XML_Parser parser;
  65. int *retPtr;
  66. } PROCESS_ARGS;
  67. static int processStream(const XML_Char *filename, XML_Parser parser);
  68. static void
  69. reportError(XML_Parser parser, const XML_Char *filename) {
  70. enum XML_Error code = XML_GetErrorCode(parser);
  71. const XML_Char *message = XML_ErrorString(code);
  72. if (message)
  73. ftprintf(stdout,
  74. T("%s") T(":%") T(XML_FMT_INT_MOD) T("u") T(":%")
  75. T(XML_FMT_INT_MOD) T("u") T(": %s\n"),
  76. filename, XML_GetErrorLineNumber(parser),
  77. XML_GetErrorColumnNumber(parser), 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, const XML_Char *filename,
  84. void *args) {
  85. XML_Parser parser = ((PROCESS_ARGS *)args)->parser;
  86. int *retPtr = ((PROCESS_ARGS *)args)->retPtr;
  87. if (XML_Parse(parser, (const char *)data, (int)size, 1) == XML_STATUS_ERROR) {
  88. reportError(parser, filename);
  89. *retPtr = 0;
  90. } else
  91. *retPtr = 1;
  92. }
  93. #if defined(_WIN32)
  94. static int
  95. isAsciiLetter(XML_Char c) {
  96. return (T('a') <= c && c <= T('z')) || (T('A') <= c && c <= T('Z'));
  97. }
  98. #endif /* _WIN32 */
  99. static const XML_Char *
  100. resolveSystemId(const XML_Char *base, const XML_Char *systemId,
  101. XML_Char **toFree) {
  102. XML_Char *s;
  103. *toFree = 0;
  104. if (! base || *systemId == T('/')
  105. #if defined(_WIN32)
  106. || *systemId == T('\\')
  107. || (isAsciiLetter(systemId[0]) && systemId[1] == T(':'))
  108. #endif
  109. )
  110. return systemId;
  111. *toFree = (XML_Char *)malloc((tcslen(base) + tcslen(systemId) + 2)
  112. * sizeof(XML_Char));
  113. if (! *toFree)
  114. return systemId;
  115. tcscpy(*toFree, base);
  116. s = *toFree;
  117. if (tcsrchr(s, T('/')))
  118. s = tcsrchr(s, T('/')) + 1;
  119. #if defined(_WIN32)
  120. if (tcsrchr(s, T('\\')))
  121. s = tcsrchr(s, T('\\')) + 1;
  122. #endif
  123. tcscpy(s, systemId);
  124. return *toFree;
  125. }
  126. static int
  127. externalEntityRefFilemap(XML_Parser parser, const XML_Char *context,
  128. const XML_Char *base, const XML_Char *systemId,
  129. const XML_Char *publicId) {
  130. int result;
  131. XML_Char *s;
  132. const XML_Char *filename;
  133. XML_Parser entParser = XML_ExternalEntityParserCreate(parser, context, 0);
  134. int filemapRes;
  135. PROCESS_ARGS args;
  136. UNUSED_P(publicId);
  137. args.retPtr = &result;
  138. args.parser = entParser;
  139. filename = resolveSystemId(base, systemId, &s);
  140. XML_SetBase(entParser, filename);
  141. filemapRes = filemap(filename, processFile, &args);
  142. switch (filemapRes) {
  143. case 0:
  144. result = 0;
  145. break;
  146. case 2:
  147. ftprintf(stderr,
  148. T("%s: file too large for memory-mapping")
  149. T(", switching to streaming\n"),
  150. filename);
  151. result = processStream(filename, entParser);
  152. break;
  153. }
  154. free(s);
  155. XML_ParserFree(entParser);
  156. return result;
  157. }
  158. static int
  159. processStream(const XML_Char *filename, XML_Parser parser) {
  160. /* passing NULL for filename means read input from stdin */
  161. int fd = 0; /* 0 is the fileno for stdin */
  162. if (filename != NULL) {
  163. fd = topen(filename, O_BINARY | O_RDONLY);
  164. if (fd < 0) {
  165. tperror(filename);
  166. return 0;
  167. }
  168. }
  169. for (;;) {
  170. int nread;
  171. char *buf = (char *)XML_GetBuffer(parser, g_read_size_bytes);
  172. if (! buf) {
  173. if (filename != NULL)
  174. close(fd);
  175. ftprintf(stderr, T("%s: out of memory\n"),
  176. filename != NULL ? filename : T("xmlwf"));
  177. return 0;
  178. }
  179. nread = read(fd, buf, g_read_size_bytes);
  180. if (nread < 0) {
  181. tperror(filename != NULL ? filename : T("STDIN"));
  182. if (filename != NULL)
  183. close(fd);
  184. return 0;
  185. }
  186. if (XML_ParseBuffer(parser, nread, nread == 0) == XML_STATUS_ERROR) {
  187. reportError(parser, filename != NULL ? filename : T("STDIN"));
  188. if (filename != NULL)
  189. close(fd);
  190. return 0;
  191. }
  192. if (nread == 0) {
  193. if (filename != NULL)
  194. close(fd);
  195. break;
  196. ;
  197. }
  198. }
  199. return 1;
  200. }
  201. static int
  202. externalEntityRefStream(XML_Parser parser, const XML_Char *context,
  203. const XML_Char *base, const XML_Char *systemId,
  204. const XML_Char *publicId) {
  205. XML_Char *s;
  206. const XML_Char *filename;
  207. int ret;
  208. XML_Parser entParser = XML_ExternalEntityParserCreate(parser, context, 0);
  209. UNUSED_P(publicId);
  210. filename = resolveSystemId(base, systemId, &s);
  211. XML_SetBase(entParser, filename);
  212. ret = processStream(filename, entParser);
  213. free(s);
  214. XML_ParserFree(entParser);
  215. return ret;
  216. }
  217. int
  218. XML_ProcessFile(XML_Parser parser, const XML_Char *filename, unsigned flags) {
  219. int result;
  220. if (! XML_SetBase(parser, filename)) {
  221. ftprintf(stderr, T("%s: out of memory"), filename);
  222. exit(1);
  223. }
  224. if (flags & XML_EXTERNAL_ENTITIES)
  225. XML_SetExternalEntityRefHandler(parser, (flags & XML_MAP_FILE)
  226. ? externalEntityRefFilemap
  227. : externalEntityRefStream);
  228. if (flags & XML_MAP_FILE) {
  229. int filemapRes;
  230. PROCESS_ARGS args;
  231. args.retPtr = &result;
  232. args.parser = parser;
  233. filemapRes = filemap(filename, processFile, &args);
  234. switch (filemapRes) {
  235. case 0:
  236. result = 0;
  237. break;
  238. case 2:
  239. ftprintf(stderr,
  240. T("%s: file too large for memory-mapping")
  241. T(", switching to streaming\n"),
  242. filename);
  243. result = processStream(filename, parser);
  244. break;
  245. }
  246. } else
  247. result = processStream(filename, parser);
  248. return result;
  249. }