You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.

test_parse.c 31 kB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735
  1. #ifdef NDEBUG
  2. #undef NDEBUG
  3. #endif
  4. #include <assert.h>
  5. #include <stddef.h>
  6. #include <stdio.h>
  7. #include <stdlib.h>
  8. #include <string.h>
  9. #include "json.h"
  10. #include "json_tokener.h"
  11. #include "json_visit.h"
  12. static void test_basic_parse(void);
  13. static void test_utf8_parse(void);
  14. static void test_verbose_parse(void);
  15. static void test_incremental_parse(void);
  16. int main(void)
  17. {
  18. MC_SET_DEBUG(1);
  19. static const char separator[] = "==================================";
  20. test_basic_parse();
  21. puts(separator);
  22. test_utf8_parse();
  23. puts(separator);
  24. test_verbose_parse();
  25. puts(separator);
  26. test_incremental_parse();
  27. puts(separator);
  28. return 0;
  29. }
  30. static json_c_visit_userfunc clear_serializer;
  31. static void do_clear_serializer(json_object *jso);
  32. static void single_incremental_parse(const char *test_string, int clear_serializer)
  33. {
  34. size_t ii;
  35. int chunksize = atoi(getenv("TEST_PARSE_CHUNKSIZE"));
  36. struct json_tokener *tok;
  37. enum json_tokener_error jerr;
  38. json_object *all_at_once_obj, *new_obj;
  39. const char *all_at_once_str, *new_str;
  40. new_obj = NULL;
  41. assert(chunksize > 0);
  42. all_at_once_obj = json_tokener_parse(test_string);
  43. if (clear_serializer)
  44. do_clear_serializer(all_at_once_obj);
  45. all_at_once_str = json_object_to_json_string(all_at_once_obj);
  46. tok = json_tokener_new();
  47. size_t test_string_len = strlen(test_string) + 1; // Including '\0' !
  48. for (ii = 0; ii < test_string_len; ii += chunksize)
  49. {
  50. int len_to_parse = chunksize;
  51. if (ii + chunksize > test_string_len)
  52. len_to_parse = test_string_len - ii;
  53. if (getenv("TEST_PARSE_DEBUG") != NULL)
  54. printf(" chunk: %.*s\n", len_to_parse, &test_string[ii]);
  55. new_obj = json_tokener_parse_ex(tok, &test_string[ii], len_to_parse);
  56. jerr = json_tokener_get_error(tok);
  57. if (jerr != json_tokener_continue || new_obj)
  58. break;
  59. }
  60. if (clear_serializer && new_obj)
  61. do_clear_serializer(new_obj);
  62. new_str = json_object_to_json_string(new_obj);
  63. if (strcmp(all_at_once_str, new_str) != 0)
  64. {
  65. printf("ERROR: failed to parse (%s) in %d byte chunks: %s != %s\n", test_string,
  66. chunksize, all_at_once_str, new_str);
  67. }
  68. json_tokener_free(tok);
  69. if (all_at_once_obj)
  70. json_object_put(all_at_once_obj);
  71. if (new_obj)
  72. json_object_put(new_obj);
  73. }
  74. static void single_basic_parse(const char *test_string, int clear_serializer)
  75. {
  76. json_object *new_obj;
  77. new_obj = json_tokener_parse(test_string);
  78. if (clear_serializer)
  79. do_clear_serializer(new_obj);
  80. printf("new_obj.to_string(%s)=%s\n", test_string, json_object_to_json_string(new_obj));
  81. json_object_put(new_obj);
  82. if (getenv("TEST_PARSE_CHUNKSIZE") != NULL)
  83. single_incremental_parse(test_string, clear_serializer);
  84. }
  85. static void test_basic_parse(void)
  86. {
  87. single_basic_parse("\"\003\"", 0);
  88. single_basic_parse("/* hello */\"foo\"", 0);
  89. single_basic_parse("// hello\n\"foo\"", 0);
  90. single_basic_parse("\"foo\"blue", 0);
  91. single_basic_parse("\'foo\'", 0);
  92. single_basic_parse("\"\\u0041\\u0042\\u0043\"", 0);
  93. single_basic_parse("\"\\u4e16\\u754c\\u00df\"", 0);
  94. single_basic_parse("\"\\u4E16\"", 0);
  95. single_basic_parse("\"\\u4e1\"", 0);
  96. single_basic_parse("\"\\u4e1@\"", 0);
  97. single_basic_parse("\"\\ud840\\u4e16\"", 0);
  98. single_basic_parse("\"\\ud840\"", 0);
  99. single_basic_parse("\"\\udd27\"", 0);
  100. // Test with a "short" high surrogate
  101. single_basic_parse("[9,'\\uDAD", 0);
  102. single_basic_parse("null", 0);
  103. single_basic_parse("NaN", 0);
  104. single_basic_parse("-NaN", 0); /* non-sensical, returns null */
  105. single_basic_parse("Inf", 0); /* must use full string, returns null */
  106. single_basic_parse("inf", 0); /* must use full string, returns null */
  107. single_basic_parse("Infinity", 0);
  108. single_basic_parse("infinity", 0);
  109. single_basic_parse("-Infinity", 0);
  110. single_basic_parse("-infinity", 0);
  111. single_basic_parse("{ \"min\": Infinity, \"max\": -Infinity}", 0);
  112. single_basic_parse("Infinity!", 0);
  113. single_basic_parse("Infinitynull", 0);
  114. single_basic_parse("InfinityXXXX", 0);
  115. single_basic_parse("-Infinitynull", 0);
  116. single_basic_parse("-InfinityXXXX", 0);
  117. single_basic_parse("Infinoodle", 0);
  118. single_basic_parse("InfinAAA", 0);
  119. single_basic_parse("-Infinoodle", 0);
  120. single_basic_parse("-InfinAAA", 0);
  121. single_basic_parse("True", 0);
  122. single_basic_parse("False", 0);
  123. /* not case sensitive */
  124. single_basic_parse("tRue", 0);
  125. single_basic_parse("fAlse", 0);
  126. single_basic_parse("nAn", 0);
  127. single_basic_parse("iNfinity", 0);
  128. single_basic_parse("12", 0);
  129. single_basic_parse("12.3", 0);
  130. /* Even though, when using json_tokener_parse() there's no way to
  131. * know when there is more data after the parsed object,
  132. * an object is successfully returned anyway (in some cases)
  133. */
  134. single_basic_parse("12.3.4", 0);
  135. single_basic_parse("2015-01-15", 0);
  136. single_basic_parse("12.3xxx", 0);
  137. single_basic_parse("12.3{\"a\":123}", 0);
  138. single_basic_parse("12.3\n", 0);
  139. single_basic_parse("12.3 ", 0);
  140. single_basic_parse("{\"FoO\" : -12.3E512}", 0);
  141. single_basic_parse("{\"FoO\" : -12.3e512}", 0);
  142. single_basic_parse("{\"FoO\" : -12.3E51.2}", 0); /* non-sensical, returns null */
  143. single_basic_parse("{\"FoO\" : -12.3E512E12}", 0); /* non-sensical, returns null */
  144. single_basic_parse("[\"\\n\"]", 0);
  145. single_basic_parse("[\"\\nabc\\n\"]", 0);
  146. single_basic_parse("[null]", 0);
  147. single_basic_parse("[]", 0);
  148. single_basic_parse("[false]", 0);
  149. single_basic_parse("[\"abc\",null,\"def\",12]", 0);
  150. single_basic_parse("{}", 0);
  151. single_basic_parse("{ \"foo\": \"bar\" }", 0);
  152. single_basic_parse("{ \'foo\': \'bar\' }", 0);
  153. single_basic_parse("{ \"foo\": \"bar\", \"baz\": null, \"bool0\": true }", 0);
  154. single_basic_parse("{ \"foo\": [null, \"foo\"] }", 0);
  155. single_basic_parse("{ \"abc\": 12, \"foo\": \"bar\", \"bool0\": false, \"bool1\": true, "
  156. "\"arr\": [ 1, 2, 3, null, 5 ] }",
  157. 0);
  158. single_basic_parse("{ \"abc\": \"blue\nred\\ngreen\" }", 0);
  159. // Clear serializer for these tests so we see the actual parsed value.
  160. single_basic_parse("null", 1);
  161. single_basic_parse("false", 1);
  162. single_basic_parse("[0e]", 1);
  163. single_basic_parse("[0e+]", 1);
  164. single_basic_parse("[0e+-1]", 1);
  165. single_basic_parse("\"hello world!\"", 1);
  166. // uint64/int64 range test
  167. single_basic_parse("[9223372036854775806]", 1);
  168. single_basic_parse("[9223372036854775807]", 1);
  169. single_basic_parse("[9223372036854775808]", 1);
  170. single_basic_parse("[-9223372036854775807]", 1);
  171. single_basic_parse("[-9223372036854775808]", 1);
  172. single_basic_parse("[-9223372036854775809]", 1);
  173. single_basic_parse("[18446744073709551614]", 1);
  174. single_basic_parse("[18446744073709551615]", 1);
  175. single_basic_parse("[18446744073709551616]", 1);
  176. }
  177. static void test_utf8_parse(void)
  178. {
  179. // json_tokener_parse doesn't support checking for byte order marks.
  180. // It's the responsibility of the caller to detect and skip a BOM.
  181. // Both of these checks return null.
  182. const char *utf8_bom = "\xEF\xBB\xBF";
  183. const char *utf8_bom_and_chars = "\xEF\xBB\xBF{}";
  184. single_basic_parse(utf8_bom, 0);
  185. single_basic_parse(utf8_bom_and_chars, 0);
  186. }
  187. // Clear the re-serialization information that the tokener
  188. // saves to ensure that the output reflects the actual
  189. // values we parsed, rather than just the original input.
  190. static void do_clear_serializer(json_object *jso)
  191. {
  192. json_c_visit(jso, 0, clear_serializer, NULL);
  193. }
  194. static int clear_serializer(json_object *jso, int flags, json_object *parent_jso,
  195. const char *jso_key, size_t *jso_index, void *userarg)
  196. {
  197. if (jso)
  198. json_object_set_serializer(jso, NULL, NULL, NULL);
  199. return JSON_C_VISIT_RETURN_CONTINUE;
  200. }
  201. static void test_verbose_parse(void)
  202. {
  203. json_object *new_obj;
  204. enum json_tokener_error error = json_tokener_success;
  205. new_obj = json_tokener_parse_verbose("{ foo }", &error);
  206. assert(error == json_tokener_error_parse_object_key_name);
  207. assert(new_obj == NULL);
  208. new_obj = json_tokener_parse("{ foo }");
  209. assert(new_obj == NULL);
  210. new_obj = json_tokener_parse("foo");
  211. assert(new_obj == NULL);
  212. new_obj = json_tokener_parse_verbose("foo", &error);
  213. assert(new_obj == NULL);
  214. /* b/c the string starts with 'f' parsing return a boolean error */
  215. assert(error == json_tokener_error_parse_boolean);
  216. puts("json_tokener_parse_verbose() OK");
  217. }
  218. struct incremental_step
  219. {
  220. const char *string_to_parse;
  221. int length;
  222. int char_offset;
  223. enum json_tokener_error expected_error;
  224. int reset_tokener; /* Set to 1 to call json_tokener_reset() after parsing */
  225. int tok_flags; /* JSON_TOKENER_* flags to pass to json_tokener_set_flags() */
  226. } incremental_steps[] = {
  227. /* Check that full json messages can be parsed, both w/ and w/o a reset */
  228. {"{ \"foo\": 123 }", -1, -1, json_tokener_success, 0, 0},
  229. {"{ \"foo\": 456 }", -1, -1, json_tokener_success, 1, 0},
  230. {"{ \"foo\": 789 }", -1, -1, json_tokener_success, 1, 0},
  231. /* Check the comment parse*/
  232. {"/* hello */{ \"foo\"", -1, -1, json_tokener_continue, 0, 0},
  233. {"/* hello */:/* hello */", -1, -1, json_tokener_continue, 0, 0},
  234. {"\"bar\"/* hello */", -1, -1, json_tokener_continue, 0, 0},
  235. {"}/* hello */", -1, -1, json_tokener_success, 1, 0},
  236. {"/ hello ", -1, 1, json_tokener_error_parse_comment, 1, 0},
  237. {"/* hello\"foo\"", -1, -1, json_tokener_continue, 1, 0},
  238. {"/* hello*\"foo\"", -1, -1, json_tokener_continue, 1, 0},
  239. {"// hello\"foo\"", -1, -1, json_tokener_continue, 1, 0},
  240. /* Check a basic incremental parse */
  241. {"{ \"foo", -1, -1, json_tokener_continue, 0, 0},
  242. {"\": {\"bar", -1, -1, json_tokener_continue, 0, 0},
  243. {"\":13}}", -1, -1, json_tokener_success, 1, 0},
  244. /* Check the UTF-16 surrogate pair handling in various ways.
  245. * Note: \ud843\udd1e is u+1D11E, Musical Symbol G Clef
  246. * Your terminal may not display these correctly, in particular
  247. * PuTTY doesn't currently show this character.
  248. */
  249. /* parse one char at every time */
  250. {"\"\\", -1, -1, json_tokener_continue, 0, 0},
  251. {"u", -1, -1, json_tokener_continue, 0, 0},
  252. {"d", -1, -1, json_tokener_continue, 0, 0},
  253. {"8", -1, -1, json_tokener_continue, 0, 0},
  254. {"3", -1, -1, json_tokener_continue, 0, 0},
  255. {"4", -1, -1, json_tokener_continue, 0, 0},
  256. {"\\", -1, -1, json_tokener_continue, 0, 0},
  257. {"u", -1, -1, json_tokener_continue, 0, 0},
  258. {"d", -1, -1, json_tokener_continue, 0, 0},
  259. {"d", -1, -1, json_tokener_continue, 0, 0},
  260. {"1", -1, -1, json_tokener_continue, 0, 0},
  261. {"e\"", -1, -1, json_tokener_success, 1, 0},
  262. /* parse two char at every time */
  263. {"\"\\u", -1, -1, json_tokener_continue, 0, 0},
  264. {"d8", -1, -1, json_tokener_continue, 0, 0},
  265. {"34", -1, -1, json_tokener_continue, 0, 0},
  266. {"\\u", -1, -1, json_tokener_continue, 0, 0},
  267. {"dd", -1, -1, json_tokener_continue, 0, 0},
  268. {"1e\"", -1, -1, json_tokener_success, 1, 0},
  269. /* check the low surrogate pair */
  270. {"\"\\ud834", -1, -1, json_tokener_continue, 0, 0},
  271. {"\\udd1e\"", -1, -1, json_tokener_success, 1, 0},
  272. {"\"\\ud834\\", -1, -1, json_tokener_continue, 0, 0},
  273. {"udd1e\"", -1, -1, json_tokener_success, 1, 0},
  274. {"\"\\ud834\\u", -1, -1, json_tokener_continue, 0, 0},
  275. {"dd1e\"", -1, -1, json_tokener_success, 1, 0},
  276. {"\"fff \\ud834\\ud", -1, -1, json_tokener_continue, 0, 0},
  277. {"d1e bar\"", -1, -1, json_tokener_success, 1, 0},
  278. {"\"fff \\ud834\\udd", -1, -1, json_tokener_continue, 0, 0},
  279. {"1e bar\"", -1, -1, json_tokener_success, 1, 0},
  280. /* \ud83d\ude00 is U+1F600, Grinning Face
  281. * Displays fine in PuTTY, though you may need "less -r"
  282. */
  283. {"\"fff \\ud83d\\ude", -1, -1, json_tokener_continue, 0, 0},
  284. {"00 bar\"", -1, -1, json_tokener_success, 1, 0},
  285. /* Check that json_tokener_reset actually resets */
  286. {"{ \"foo", -1, -1, json_tokener_continue, 1, 0},
  287. {": \"bar\"}", -1, 0, json_tokener_error_parse_unexpected, 1, 0},
  288. /* Check incremental parsing with trailing characters */
  289. {"{ \"foo", -1, -1, json_tokener_continue, 0, 0},
  290. {"\": {\"bar", -1, -1, json_tokener_continue, 0, 0},
  291. {"\":13}}XXXX", 10, 6, json_tokener_success, 0, 0},
  292. {"XXXX", 4, 0, json_tokener_error_parse_unexpected, 1, 0},
  293. /* Check that trailing characters can change w/o a reset */
  294. {"{\"x\": 123 }\"X\"", -1, 11, json_tokener_success, 0, 0},
  295. {"\"Y\"", -1, -1, json_tokener_success, 1, 0},
  296. /* Trailing characters should cause a failure in strict mode */
  297. {"{\"foo\":9}{\"bar\":8}", -1, 9, json_tokener_error_parse_unexpected, 1, JSON_TOKENER_STRICT},
  298. /* ... unless explicitly allowed. */
  299. {"{\"foo\":9}{\"bar\":8}", -1, 9, json_tokener_success, 0,
  300. JSON_TOKENER_STRICT | JSON_TOKENER_ALLOW_TRAILING_CHARS},
  301. {"{\"b\":8}ignored garbage", -1, 7, json_tokener_success, 1,
  302. JSON_TOKENER_STRICT | JSON_TOKENER_ALLOW_TRAILING_CHARS},
  303. /* To stop parsing a number we need to reach a non-digit, e.g. a \0 */
  304. {"1", 1, 1, json_tokener_continue, 0, 0},
  305. /* This should parse as the number 12, since it continues the "1" */
  306. {"2", 2, 1, json_tokener_success, 0, 0},
  307. {"12{", 3, 2, json_tokener_success, 1, 0},
  308. /* Parse number in strict mode */
  309. {"[02]", -1, 3, json_tokener_error_parse_number, 1, JSON_TOKENER_STRICT},
  310. {"0e+0", 5, 4, json_tokener_success, 1, 0},
  311. {"[0e+0]", -1, -1, json_tokener_success, 1, 0},
  312. /* The behavior when missing the exponent varies slightly */
  313. {"0e", 2, 2, json_tokener_continue, 1, 0},
  314. {"0e", 3, 2, json_tokener_success, 1, 0},
  315. {"0e", 3, 2, json_tokener_error_parse_eof, 1, JSON_TOKENER_STRICT},
  316. {"[0e]", -1, -1, json_tokener_success, 1, 0},
  317. {"[0e]", -1, 3, json_tokener_error_parse_number, 1, JSON_TOKENER_STRICT},
  318. {"0e+", 3, 3, json_tokener_continue, 1, 0},
  319. {"0e+", 4, 3, json_tokener_success, 1, 0},
  320. {"0e+", 4, 3, json_tokener_error_parse_eof, 1, JSON_TOKENER_STRICT},
  321. {"[0e+]", -1, -1, json_tokener_success, 1, 0},
  322. {"[0e+]", -1, 4, json_tokener_error_parse_number, 1, JSON_TOKENER_STRICT},
  323. {"0e-", 3, 3, json_tokener_continue, 1, 0},
  324. {"0e-", 4, 3, json_tokener_success, 1, 0},
  325. {"0e-", 4, 3, json_tokener_error_parse_eof, 1, JSON_TOKENER_STRICT},
  326. {"[0e-]", -1, -1, json_tokener_success, 1, 0},
  327. {"[0e-]", -1, 4, json_tokener_error_parse_number, 1, JSON_TOKENER_STRICT},
  328. /* You might expect this to fail, but it won't because
  329. it's a valid partial parse; note the char_offset: */
  330. {"0e+-", 5, 3, json_tokener_success, 1, 0},
  331. {"0e+-", 5, 3, json_tokener_error_parse_number, 1, JSON_TOKENER_STRICT},
  332. {"[0e+-]", -1, 4, json_tokener_error_parse_number, 1, 0},
  333. /* Similar tests for other kinds of objects: */
  334. /* These could all return success immediately, since regardless of
  335. what follows the false/true/null token we *will* return a json object,
  336. but it currently doesn't work that way. hmm... */
  337. {"false", 5, 5, json_tokener_continue, 1, 0},
  338. {"false", 6, 5, json_tokener_success, 1, 0},
  339. {"true", 4, 4, json_tokener_continue, 1, 0},
  340. {"true", 5, 4, json_tokener_success, 1, 0},
  341. {"null", 4, 4, json_tokener_continue, 1, 0},
  342. {"null", 5, 4, json_tokener_success, 1, 0},
  343. {"Infinity", 9, 8, json_tokener_success, 1, 0},
  344. {"infinity", 9, 8, json_tokener_success, 1, 0},
  345. {"-infinity", 10, 9, json_tokener_success, 1, 0},
  346. {"infinity", 9, 0, json_tokener_error_parse_unexpected, 1, JSON_TOKENER_STRICT},
  347. {"-infinity", 10, 1, json_tokener_error_parse_unexpected, 1, JSON_TOKENER_STRICT},
  348. {"inf", 3, 3, json_tokener_continue, 0, 0},
  349. {"inity", 6, 5, json_tokener_success, 1, 0},
  350. {"-inf", 4, 4, json_tokener_continue, 0, 0},
  351. {"inity", 6, 5, json_tokener_success, 1, 0},
  352. {"i", 1, 1, json_tokener_continue, 0, 0},
  353. {"n", 1, 1, json_tokener_continue, 0, 0},
  354. {"f", 1, 1, json_tokener_continue, 0, 0},
  355. {"i", 1, 1, json_tokener_continue, 0, 0},
  356. {"n", 1, 1, json_tokener_continue, 0, 0},
  357. {"i", 1, 1, json_tokener_continue, 0, 0},
  358. {"t", 1, 1, json_tokener_continue, 0, 0},
  359. {"y", 1, 1, json_tokener_continue, 0, 0},
  360. {"", 1, 0, json_tokener_success, 1, 0},
  361. {"-", 1, 1, json_tokener_continue, 0, 0},
  362. {"inf", 3, 3, json_tokener_continue, 0, 0},
  363. {"ini", 3, 3, json_tokener_continue, 0, 0},
  364. {"ty", 3, 2, json_tokener_success, 1, 0},
  365. {"-", 1, 1, json_tokener_continue, 0, 0},
  366. {"i", 1, 1, json_tokener_continue, 0, 0},
  367. {"nfini", 5, 5, json_tokener_continue, 0, 0},
  368. {"ty", 3, 2, json_tokener_success, 1, 0},
  369. {"-i", 2, 2, json_tokener_continue, 0, 0},
  370. {"nfinity", 8, 7, json_tokener_success, 1, 0},
  371. {"InfinityX", 10, 8, json_tokener_success, 0, 0},
  372. {"X", 1, 0, json_tokener_error_parse_unexpected, 1, 0},
  373. {"Infinity1234", 13, 8, json_tokener_success, 0, 0},
  374. {"1234", 5, 4, json_tokener_success, 1, 0},
  375. {"Infinity9999", 8, 8, json_tokener_continue, 0, 0},
  376. /* returns the Infinity loaded up by the previous call: */
  377. {"1234", 5, 0, json_tokener_success, 0, 0},
  378. {"1234", 5, 4, json_tokener_success, 1, 0},
  379. /* INT64_MAX */
  380. {"[9223372036854775807]", 22, 21, json_tokener_success, 1, 0},
  381. /* INT64_MAX+1 => parsed as uint64 */
  382. {"[9223372036854775808]", 22, 21, json_tokener_success, 1, 0},
  383. /* INT64_MIN */
  384. {"[-9223372036854775808]", 23, 22, json_tokener_success, 1, 0},
  385. /* INT64_MIN-1 => success, but value ends up capped */
  386. {"[-9223372036854775809]", 23, 22, json_tokener_success, 1, 0},
  387. /* INT64_MIN-1 => failure due to underflow detected */
  388. {"[-9223372036854775809]", 23, 21, json_tokener_error_parse_number, 1, JSON_TOKENER_STRICT},
  389. /* UINT64_MAX */
  390. {"[18446744073709551615]", 23, 22, json_tokener_success, 1, 0},
  391. /* UINT64_MAX+1 => success, but value ends up capped */
  392. {"[18446744073709551616]", 23, 22, json_tokener_success, 1, 0},
  393. /* UINT64_MAX+1 => failure due to overflow detected */
  394. {"[18446744073709551616]", 23, 21, json_tokener_error_parse_number, 1, JSON_TOKENER_STRICT},
  395. /* XXX this seems like a bug, should fail with _error_parse_number instead */
  396. {"18446744073709551616", 21, 20, json_tokener_error_parse_eof, 1, JSON_TOKENER_STRICT},
  397. /* Exceeding integer limits as double parse OK */
  398. {"[9223372036854775808.0]", 24, 23, json_tokener_success, 1, 0},
  399. {"[-9223372036854775809.0]", 25, 24, json_tokener_success, 1, JSON_TOKENER_STRICT},
  400. {"[18446744073709551615.0]", 25, 24, json_tokener_success, 1, 0},
  401. {"[18446744073709551616.0]", 25, 24, json_tokener_success, 1, JSON_TOKENER_STRICT},
  402. /* offset=1 because "n" is the start of "null". hmm... */
  403. {"noodle", 7, 1, json_tokener_error_parse_null, 1, 0},
  404. /* offset=2 because "na" is the start of "nan". hmm... */
  405. {"naodle", 7, 2, json_tokener_error_parse_null, 1, 0},
  406. /* offset=2 because "tr" is the start of "true". hmm... */
  407. {"track", 6, 2, json_tokener_error_parse_boolean, 1, 0},
  408. {"fail", 5, 2, json_tokener_error_parse_boolean, 1, 0},
  409. /* Although they may initially look like they should fail,
  410. * the next few tests check that parsing multiple sequential
  411. * json objects in the input works as expected
  412. */
  413. {"null123", 8, 4, json_tokener_success, 0, 0},
  414. {&"null123"[4], 4, 3, json_tokener_success, 1, 0},
  415. {"nullx", 6, 4, json_tokener_success, 0, 0},
  416. {&"nullx"[4], 2, 0, json_tokener_error_parse_unexpected, 1, 0},
  417. {"{\"a\":1}{\"b\":2}", 15, 7, json_tokener_success, 0, 0},
  418. {&"{\"a\":1}{\"b\":2}"[7], 8, 7, json_tokener_success, 1, 0},
  419. /*
  420. * Though this may seem invalid at first glance, it
  421. * parses as three separate numbers, 2015, -1 and -15
  422. * Of course, simply pasting together a stream of arbitrary
  423. * positive numbers won't work, since there'll be no way to
  424. * tell where in e.g. "2015015" the next number stats, so
  425. * a reliably parsable stream must not include json_type_int
  426. * or json_type_double objects without some other delimiter.
  427. * e.g. whitespace
  428. */
  429. {&"2015-01-15"[0], 11, 4, json_tokener_success, 1, 0},
  430. {&"2015-01-15"[4], 7, 3, json_tokener_success, 1, 0},
  431. {&"2015-01-15"[7], 4, 3, json_tokener_success, 1, 0},
  432. {&"2015 01 15"[0], 11, 5, json_tokener_success, 1, 0},
  433. {&"2015 01 15"[4], 7, 4, json_tokener_success, 1, 0},
  434. {&"2015 01 15"[7], 4, 3, json_tokener_success, 1, 0},
  435. /* Strings have a well defined end point, so we can stop at the quote */
  436. {"\"blue\"", -1, -1, json_tokener_success, 0, 0},
  437. /* Check each of the escape sequences defined by the spec */
  438. {"\"\\\"\"", -1, -1, json_tokener_success, 0, 0},
  439. {"\"\\\\\"", -1, -1, json_tokener_success, 0, 0},
  440. {"\"\\b\"", -1, -1, json_tokener_success, 0, 0},
  441. {"\"\\f\"", -1, -1, json_tokener_success, 0, 0},
  442. {"\"\\n\"", -1, -1, json_tokener_success, 0, 0},
  443. {"\"\\r\"", -1, -1, json_tokener_success, 0, 0},
  444. {"\"\\t\"", -1, -1, json_tokener_success, 0, 0},
  445. {"\"\\/\"", -1, -1, json_tokener_success, 0, 0},
  446. // Escaping a forward slash is optional
  447. {"\"/\"", -1, -1, json_tokener_success, 0, 0},
  448. /* Check wrong escape sequences */
  449. {"\"\\a\"", -1, 2, json_tokener_error_parse_string, 1, 0},
  450. /* Check '\'' in strict model */
  451. {"\'foo\'", -1, 0, json_tokener_error_parse_unexpected, 1, JSON_TOKENER_STRICT},
  452. /* Parse array/object */
  453. {"[1,2,3]", -1, -1, json_tokener_success, 0, 0},
  454. {"[1,2,3}", -1, 6, json_tokener_error_parse_array, 1, 0},
  455. {"{\"a\"}", -1, 4, json_tokener_error_parse_object_key_sep, 1, 0},
  456. {"{\"a\":1]", -1, 6, json_tokener_error_parse_object_value_sep, 1, 0},
  457. {"{\"a\"::1}", -1, 5, json_tokener_error_parse_unexpected, 1, 0},
  458. {"{\"a\":}", -1, 5, json_tokener_error_parse_unexpected, 1, 0},
  459. {"{\"a\":1,\"a\":2}", -1, -1, json_tokener_success, 1, 0},
  460. {"\"a\":1}", -1, 3, json_tokener_success, 1, 0},
  461. {"{\"a\":1", -1, -1, json_tokener_continue, 1, 0}, //}
  462. {"[,]", -1, 1, json_tokener_error_parse_unexpected, 1, 0},
  463. {"[,1]", -1, 1, json_tokener_error_parse_unexpected, 1, 0},
  464. /* This behaviour doesn't entirely follow the json spec, but until we have
  465. * a way to specify how strict to be we follow Postel's Law and be liberal
  466. * in what we accept (up to a point).
  467. */
  468. {"[1,2,3,]", -1, -1, json_tokener_success, 0, 0},
  469. {"[1,2,,3,]", -1, 5, json_tokener_error_parse_unexpected, 0, 0},
  470. {"[1,2,3,]", -1, 7, json_tokener_error_parse_unexpected, 1, JSON_TOKENER_STRICT},
  471. {"{\"a\":1,}", -1, 7, json_tokener_error_parse_unexpected, 1, JSON_TOKENER_STRICT},
  472. // utf-8 test
  473. // acsll encoding
  474. {"\x22\x31\x32\x33\x61\x73\x63\x24\x25\x26\x22", -1, -1, json_tokener_success, 1,
  475. JSON_TOKENER_VALIDATE_UTF8},
  476. {"\x22\x31\x32\x33\x61\x73\x63\x24\x25\x26\x22", -1, -1, json_tokener_success, 1, 0},
  477. // utf-8 encoding
  478. {"\x22\xe4\xb8\x96\xe7\x95\x8c\x22", -1, -1, json_tokener_success, 1,
  479. JSON_TOKENER_VALIDATE_UTF8},
  480. {"\x22\xe4\xb8", -1, 3, json_tokener_error_parse_utf8_string, 0, JSON_TOKENER_VALIDATE_UTF8},
  481. {"\x96\xe7\x95\x8c\x22", -1, 0, json_tokener_error_parse_utf8_string, 1,
  482. JSON_TOKENER_VALIDATE_UTF8},
  483. {"\x22\xe4\xb8\x96\xe7\x95\x8c\x22", -1, -1, json_tokener_success, 1, 0},
  484. {"\x22\xcf\x80\xcf\x86\x22", -1, -1, json_tokener_success, 1, JSON_TOKENER_VALIDATE_UTF8},
  485. {"\x22\xf0\xa5\x91\x95\x22", -1, -1, json_tokener_success, 1, JSON_TOKENER_VALIDATE_UTF8},
  486. // wrong utf-8 encoding
  487. {"\x22\xe6\x9d\x4e\x22", -1, 3, json_tokener_error_parse_utf8_string, 1,
  488. JSON_TOKENER_VALIDATE_UTF8},
  489. {"\x22\xe6\x9d\x4e\x22", -1, 5, json_tokener_success, 1, 0},
  490. // GBK encoding
  491. {"\x22\xc0\xee\xc5\xf4\x22", -1, 2, json_tokener_error_parse_utf8_string, 1,
  492. JSON_TOKENER_VALIDATE_UTF8},
  493. {"\x22\xc0\xee\xc5\xf4\x22", -1, 6, json_tokener_success, 1, 0},
  494. // char after space
  495. {"\x20\x20\x22\xe4\xb8\x96\x22", -1, -1, json_tokener_success, 1, JSON_TOKENER_VALIDATE_UTF8},
  496. {"\x20\x20\x81\x22\xe4\xb8\x96\x22", -1, 2, json_tokener_error_parse_utf8_string, 1,
  497. JSON_TOKENER_VALIDATE_UTF8},
  498. {"\x5b\x20\x81\x31\x5d", -1, 2, json_tokener_error_parse_utf8_string, 1,
  499. JSON_TOKENER_VALIDATE_UTF8},
  500. // char in state inf
  501. {"\x49\x6e\x66\x69\x6e\x69\x74\x79", 9, 8, json_tokener_success, 1, 0},
  502. {"\x49\x6e\x66\x81\x6e\x69\x74\x79", -1, 3, json_tokener_error_parse_utf8_string, 1,
  503. JSON_TOKENER_VALIDATE_UTF8},
  504. // char in escape unicode
  505. {"\x22\x5c\x75\x64\x38\x35\x35\x5c\x75\x64\x63\x35\x35\x22", 15, 14, json_tokener_success, 1,
  506. JSON_TOKENER_VALIDATE_UTF8},
  507. {"\x22\x5c\x75\x64\x38\x35\x35\xc0\x75\x64\x63\x35\x35\x22", -1, 8,
  508. json_tokener_error_parse_utf8_string, 1, JSON_TOKENER_VALIDATE_UTF8},
  509. {"\x22\x5c\x75\x64\x30\x30\x33\x31\xc0\x22", -1, 9, json_tokener_error_parse_utf8_string, 1,
  510. JSON_TOKENER_VALIDATE_UTF8},
  511. // char in number
  512. {"\x31\x31\x81\x31\x31", -1, 2, json_tokener_error_parse_utf8_string, 1,
  513. JSON_TOKENER_VALIDATE_UTF8},
  514. // char in object
  515. {"\x7b\x22\x31\x81\x22\x3a\x31\x7d", -1, 3, json_tokener_error_parse_utf8_string, 1,
  516. JSON_TOKENER_VALIDATE_UTF8},
  517. // Note, current asciiz APIs can't parse \x00, skip it
  518. { "\"0\x01\x02\x02\x03\x04\x05\x06\x07\x08\x09\x0a\x0b\x0c\x0d\x0e\x0f" \
  519. "\x10\x11\x12\x12\x13\x14\x15\x16\x17\x18\x19\x1a\x1b\x1c\x1d\x1e\x1f\"",
  520. -1, -1, json_tokener_success, 1, 0 },
  521. // Test control chars again, this time in strict mode, which should fail
  522. { "\"\x01\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  523. { "\"\x02\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  524. { "\"\x03\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  525. { "\"\x04\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  526. { "\"\x05\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  527. { "\"\x06\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  528. { "\"\x07\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  529. { "\"\x08\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  530. { "\"\x09\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  531. { "\"\x0a\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  532. { "\"\x0b\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  533. { "\"\x0c\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  534. { "\"\x0d\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  535. { "\"\x0e\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  536. { "\"\x0f\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  537. { "\"\x10\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  538. { "\"\x11\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  539. { "\"\x12\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  540. { "\"\x13\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  541. { "\"\x14\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  542. { "\"\x15\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  543. { "\"\x16\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  544. { "\"\x17\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  545. { "\"\x18\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  546. { "\"\x19\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  547. { "\"\x1a\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  548. { "\"\x1b\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  549. { "\"\x1c\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  550. { "\"\x1d\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  551. { "\"\x1e\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  552. { "\"\x1f\"", -1, 1, json_tokener_error_parse_string, 1, JSON_TOKENER_STRICT },
  553. {NULL, -1, -1, json_tokener_success, 0, 0},
  554. };
  555. static void test_incremental_parse(void)
  556. {
  557. json_object *new_obj;
  558. enum json_tokener_error jerr;
  559. struct json_tokener *tok;
  560. const char *string_to_parse;
  561. int ii;
  562. int num_ok, num_error;
  563. num_ok = 0;
  564. num_error = 0;
  565. printf("Starting incremental tests.\n");
  566. printf("Note: quotes and backslashes seen in the output here are literal values passed\n");
  567. printf(" to the parse functions. e.g. this is 4 characters: \"\\f\"\n");
  568. string_to_parse = "{ \"foo"; /* } */
  569. printf("json_tokener_parse(%s) ... ", string_to_parse);
  570. new_obj = json_tokener_parse(string_to_parse);
  571. if (new_obj == NULL)
  572. puts("got error as expected");
  573. /* test incremental parsing in various forms */
  574. tok = json_tokener_new();
  575. for (ii = 0; incremental_steps[ii].string_to_parse != NULL; ii++)
  576. {
  577. int this_step_ok = 0;
  578. struct incremental_step *step = &incremental_steps[ii];
  579. int length = step->length;
  580. size_t expected_char_offset;
  581. json_tokener_set_flags(tok, step->tok_flags);
  582. if (length == -1)
  583. length = (int)strlen(step->string_to_parse);
  584. if (step->char_offset == -1)
  585. expected_char_offset = length;
  586. else
  587. expected_char_offset = step->char_offset;
  588. printf("json_tokener_parse_ex(tok, %-12s, %3d) ... ", step->string_to_parse,
  589. length);
  590. new_obj = json_tokener_parse_ex(tok, step->string_to_parse, length);
  591. jerr = json_tokener_get_error(tok);
  592. if (step->expected_error != json_tokener_success)
  593. {
  594. if (new_obj != NULL)
  595. printf("ERROR: invalid object returned: %s\n",
  596. json_object_to_json_string(new_obj));
  597. else if (jerr != step->expected_error)
  598. printf("ERROR: got wrong error: %s\n",
  599. json_tokener_error_desc(jerr));
  600. else if (json_tokener_get_parse_end(tok) != expected_char_offset)
  601. printf("ERROR: wrong char_offset %zu != expected %zu\n",
  602. json_tokener_get_parse_end(tok), expected_char_offset);
  603. else
  604. {
  605. printf("OK: got correct error: %s\n",
  606. json_tokener_error_desc(jerr));
  607. this_step_ok = 1;
  608. }
  609. }
  610. else
  611. {
  612. if (new_obj == NULL &&
  613. !(step->length >= 4 && strncmp(step->string_to_parse, "null", 4) == 0))
  614. printf("ERROR: expected valid object, instead: %s\n",
  615. json_tokener_error_desc(jerr));
  616. else if (json_tokener_get_parse_end(tok) != expected_char_offset)
  617. printf("ERROR: wrong char_offset %zu != expected %zu\n",
  618. json_tokener_get_parse_end(tok), expected_char_offset);
  619. else
  620. {
  621. printf("OK: got object of type [%s]: %s\n",
  622. json_type_to_name(json_object_get_type(new_obj)),
  623. json_object_to_json_string(new_obj));
  624. this_step_ok = 1;
  625. }
  626. }
  627. if (new_obj)
  628. json_object_put(new_obj);
  629. if (step->reset_tokener & 1)
  630. json_tokener_reset(tok);
  631. if (this_step_ok)
  632. num_ok++;
  633. else
  634. num_error++;
  635. }
  636. json_tokener_free(tok);
  637. printf("End Incremental Tests OK=%d ERROR=%d\n", num_ok, num_error);
  638. }