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

lex.c 6.1KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351
  1. #include "parse/lex.h"
  2. #include <stdlib.h>
  3. static int parse_number(const char *str, double *num) {
  4. // TODO: Floats
  5. size_t len = strlen(str);
  6. *num = 0;
  7. int power = 1;
  8. for (int i = (int)len - 1; i >= 0; --i) {
  9. char ch = str[i];
  10. if (ch >= '0' && ch <= '9') {
  11. *num += (ch - '0') * power;
  12. power *= 10;
  13. } else {
  14. return -1;
  15. }
  16. }
  17. return 0;
  18. }
  19. const char *l2_token_kind_name(enum l2_token_kind kind) {
  20. switch (kind) {
  21. case L2_TOK_OPEN_PAREN:
  22. return "open-paren";
  23. case L2_TOK_CLOSE_PAREN:
  24. return "close-paren";
  25. case L2_TOK_OPEN_BRACE:
  26. return "open-brace";
  27. case L2_TOK_CLOSE_BRACE:
  28. return "close-brace";
  29. case L2_TOK_OPEN_BRACKET:
  30. return "open-bracket";
  31. case L2_TOK_CLOSE_BRACKET:
  32. return "close-bracket";
  33. case L2_TOK_COMMA:
  34. return "comma";
  35. case L2_TOK_PERIOD:
  36. return "period";
  37. case L2_TOK_COLON_EQ:
  38. return "colon-equals";
  39. case L2_TOK_EOL:
  40. return "end-of-line";
  41. case L2_TOK_EOF:
  42. return "end-of-file";
  43. case L2_TOK_NUMBER:
  44. return "number";
  45. case L2_TOK_STRING:
  46. return "string";
  47. case L2_TOK_IDENT:
  48. return "ident";
  49. case L2_TOK_ERROR:
  50. return "error";
  51. }
  52. }
  53. void l2_token_free(struct l2_token *tok) {
  54. if (tok->kind == L2_TOK_STRING || tok->kind == L2_TOK_IDENT) {
  55. free(tok->v.str);
  56. }
  57. }
  58. char *l2_token_extract_str(struct l2_token *tok) {
  59. char *str = tok->v.str;
  60. tok->v.str = NULL;
  61. return str;
  62. }
  63. void l2_lexer_init(struct l2_lexer *lexer, struct l2_io_reader *r) {
  64. lexer->toks[0].kind = L2_TOK_EOF,
  65. lexer->tokidx = 0;
  66. lexer->line = 1;
  67. lexer->ch = 1;
  68. lexer->parens = 0;
  69. l2_bufio_reader_init(&lexer->reader, r);
  70. }
  71. static int peek_ch(struct l2_lexer *lexer) {
  72. int ch = l2_bufio_peek(&lexer->reader, 1);
  73. return ch;
  74. }
  75. static int read_ch(struct l2_lexer *lexer) {
  76. int ch = l2_bufio_get(&lexer->reader);
  77. lexer->ch += 1;
  78. if (ch == '\n') {
  79. lexer->ch = 1;
  80. lexer->line += 1;
  81. }
  82. return ch;
  83. }
  84. static int is_whitespace(int ch) {
  85. return ch == ' ' || ch == '\r' || ch == '\n' || ch == '\t';
  86. }
  87. static int skip_whitespace(struct l2_lexer *lexer) {
  88. int nl = 0;
  89. while (is_whitespace(l2_bufio_peek(&lexer->reader, 1))) {
  90. int ch = read_ch(lexer);
  91. if (ch == '\n') {
  92. nl = 1;
  93. }
  94. }
  95. return nl;
  96. }
  97. static void read_string(struct l2_lexer *lexer, struct l2_token *tok) {
  98. tok->kind = L2_TOK_STRING;
  99. tok->v.str = malloc(16);
  100. if (tok->v.str == NULL) {
  101. tok->kind = L2_TOK_ERROR;
  102. tok->v.str = "Allocaton failure";
  103. return;
  104. }
  105. size_t size = 16;
  106. size_t idx = 0;
  107. while (1) {
  108. int ch = read_ch(lexer);
  109. if (ch == '"') {
  110. return;
  111. } else if (ch == EOF) {
  112. tok->kind = L2_TOK_EOF;
  113. free(tok->v.str);
  114. tok->v.str = "Unexpected EOF";
  115. return;
  116. } else if (ch == '\\') {
  117. int ch2 = read_ch(lexer);
  118. switch (ch2) {
  119. case 'n':
  120. ch = '\n';
  121. break;
  122. case 'r':
  123. ch = '\r';
  124. break;
  125. case 't':
  126. ch = '\t';
  127. break;
  128. case EOF:
  129. tok->kind = L2_TOK_EOF;
  130. free(tok->v.str);
  131. tok->v.str = "Unexpected EOF";
  132. return;
  133. default:
  134. ch = ch2;
  135. break;
  136. }
  137. }
  138. tok->v.str[idx++] = (char)ch;
  139. if (idx >= size) {
  140. size *= 2;
  141. char *newbuf = realloc(tok->v.str, size);
  142. if (newbuf == NULL) {
  143. free(tok->v.str);
  144. tok->kind = L2_TOK_ERROR;
  145. tok->v.str = "Allocation failure";
  146. return;
  147. }
  148. tok->v.str = newbuf;
  149. }
  150. }
  151. }
  152. static void read_ident(struct l2_lexer *lexer, struct l2_token *tok) {
  153. tok->kind = L2_TOK_IDENT;
  154. tok->v.str = malloc(16);
  155. if (tok->v.str == NULL) {
  156. tok->kind = L2_TOK_ERROR;
  157. tok->v.str = "Allocaton failure";
  158. return;
  159. }
  160. size_t size = 16;
  161. size_t idx = 0;
  162. while (1) {
  163. int ch = peek_ch(lexer);
  164. if (is_whitespace(ch)) {
  165. tok->v.str[idx] = '\0';
  166. return;
  167. }
  168. switch (ch) {
  169. case '(':
  170. case ')':
  171. case '{':
  172. case '}':
  173. case '[':
  174. case ']':
  175. case ',':
  176. case '.':
  177. case ':':
  178. case ';':
  179. case EOF:
  180. tok->v.str[idx] = '\0';
  181. return;
  182. }
  183. tok->v.str[idx++] = (char)read_ch(lexer);
  184. if (idx + 1 >= size) {
  185. size *= 2;
  186. char *newbuf = realloc(tok->v.str, size);
  187. if (newbuf == NULL) {
  188. free(tok->v.str);
  189. tok->kind = L2_TOK_ERROR;
  190. tok->v.str = "Allocation failure";
  191. return;
  192. }
  193. tok->v.str = newbuf;
  194. }
  195. }
  196. }
  197. static void read_tok(struct l2_lexer *lexer, struct l2_token *tok) {
  198. tok->line = lexer->line;
  199. tok->ch = lexer->ch;
  200. int nl = skip_whitespace(lexer);
  201. if (nl && lexer->parens == 0) {
  202. tok->kind = L2_TOK_EOL;
  203. return;
  204. }
  205. int ch = peek_ch(lexer);
  206. switch (ch) {
  207. case '(':
  208. read_ch(lexer);
  209. tok->kind = L2_TOK_OPEN_PAREN;
  210. lexer->parens += 1;
  211. break;
  212. case ')':
  213. read_ch(lexer);
  214. tok->kind = L2_TOK_CLOSE_PAREN;
  215. lexer->parens -= 1;
  216. break;
  217. case '{':
  218. read_ch(lexer);
  219. tok->kind = L2_TOK_OPEN_BRACE;
  220. break;
  221. case '}':
  222. read_ch(lexer);
  223. tok->kind = L2_TOK_CLOSE_BRACE;
  224. break;
  225. case '[':
  226. read_ch(lexer);
  227. tok->kind = L2_TOK_OPEN_BRACKET;
  228. break;
  229. case ']':
  230. read_ch(lexer);
  231. tok->kind = L2_TOK_CLOSE_BRACKET;
  232. break;
  233. case ';':
  234. tok->kind = L2_TOK_EOL;
  235. do {
  236. read_ch(lexer);
  237. skip_whitespace(lexer);
  238. } while (peek_ch(lexer) == ';');
  239. break;
  240. case ',':
  241. read_ch(lexer);
  242. tok->kind = L2_TOK_COMMA;
  243. break;
  244. case '.':
  245. read_ch(lexer);
  246. tok->kind = L2_TOK_PERIOD;
  247. break;
  248. case ':':
  249. read_ch(lexer);
  250. {
  251. ch = read_ch(lexer);
  252. switch (ch) {
  253. case '=':
  254. tok->kind = L2_TOK_COLON_EQ;
  255. break;
  256. default:
  257. tok->kind = L2_TOK_ERROR;
  258. tok->v.str = "Unexpected character";
  259. break;
  260. }
  261. }
  262. break;
  263. case EOF:
  264. tok->kind = L2_TOK_EOF;
  265. break;
  266. case '"':
  267. read_ch(lexer);
  268. read_string(lexer, tok);
  269. break;
  270. default:
  271. read_ident(lexer, tok);
  272. if (tok->kind != L2_TOK_IDENT) {
  273. break;
  274. }
  275. double num;
  276. if (parse_number(tok->v.str, &num) >= 0) {
  277. free(tok->v.str);
  278. tok->kind = L2_TOK_NUMBER;
  279. tok->v.num = num;
  280. }
  281. break;
  282. }
  283. }
  284. struct l2_token *l2_lexer_peek(struct l2_lexer *lexer, int count) {
  285. int offset = count - 1;
  286. while (offset >= lexer->tokidx) {
  287. read_tok(lexer, &lexer->toks[lexer->tokidx++]);
  288. }
  289. return &lexer->toks[offset];
  290. }
  291. void l2_lexer_consume(struct l2_lexer *lexer) {
  292. l2_token_free(&lexer->toks[0]);
  293. lexer->tokidx -= 1;
  294. memmove(lexer->toks, lexer->toks + 1, lexer->tokidx * sizeof(*lexer->toks));
  295. }
  296. void l2_lexer_skip_opt(struct l2_lexer *lexer, enum l2_token_kind kind) {
  297. if (l2_lexer_peek(lexer, 1)->kind == kind) {
  298. l2_lexer_consume(lexer);
  299. }
  300. }