diff options
Diffstat (limited to 'src/llex.c')
| -rw-r--r-- | src/llex.c | 365 |
1 files changed, 201 insertions, 164 deletions
@@ -1,24 +1,23 @@ /* -** $Id: llex.c,v 1.72 2000/10/20 16:39:03 roberto Exp $ +** $Id: llex.c,v 1.119 2003/03/24 12:39:34 roberto Exp $ ** Lexical Analyzer ** See Copyright Notice in lua.h */ #include <ctype.h> -#include <stdio.h> #include <string.h> +#define llex_c + #include "lua.h" +#include "ldo.h" #include "llex.h" -#include "lmem.h" #include "lobject.h" #include "lparser.h" #include "lstate.h" #include "lstring.h" -#include "ltable.h" -#include "luadebug.h" #include "lzio.h" @@ -29,16 +28,22 @@ /* ORDER RESERVED */ static const char *const token2string [] = { - "and", "break", "do", "else", "elseif", "end", "for", - "function", "if", "local", "nil", "not", "or", "repeat", "return", "then", - "until", "while", "", "..", "...", "==", ">=", "<=", "~=", "", "", "<eof>"}; + "and", "break", "do", "else", "elseif", + "end", "false", "for", "function", "if", + "in", "local", "nil", "not", "or", "repeat", + "return", "then", "true", "until", "while", "*name", + "..", "...", "==", ">=", "<=", "~=", + "*number", "*string", "<eof>" +}; void luaX_init (lua_State *L) { int i; for (i=0; i<NUM_RESERVED; i++) { TString *ts = luaS_new(L, token2string[i]); - ts->marked = (unsigned char)(RESERVEDMARK+i); /* reserved word */ + luaS_fix(ts); /* reserved words are never collected */ + lua_assert(strlen(token2string[i])+1 <= TOKEN_LEN); + ts->tsv.reserved = cast(lu_byte, i+1); /* reserved word */ } } @@ -48,50 +53,64 @@ void luaX_init (lua_State *L) { void luaX_checklimit (LexState *ls, int val, int limit, const char *msg) { if (val > limit) { - char buff[100]; - sprintf(buff, "too many %.50s (limit=%d)", msg, limit); - luaX_error(ls, buff, ls->t.token); + msg = luaO_pushfstring(ls->L, "too many %s (limit=%d)", msg, limit); + luaX_syntaxerror(ls, msg); } } -void luaX_syntaxerror (LexState *ls, const char *s, const char *token) { +void luaX_errorline (LexState *ls, const char *s, const char *token, int line) { + lua_State *L = ls->L; char buff[MAXSRC]; - luaO_chunkid(buff, ls->source->str, sizeof(buff)); - luaO_verror(ls->L, "%.99s;\n last token read: `%.30s' at line %d in %.80s", - s, token, ls->linenumber, buff); + luaO_chunkid(buff, getstr(ls->source), MAXSRC); + luaO_pushfstring(L, "%s:%d: %s near `%s'", buff, line, s, token); + luaD_throw(L, LUA_ERRSYNTAX); } -void luaX_error (LexState *ls, const char *s, int token) { - char buff[TOKEN_LEN]; - luaX_token2str(token, buff); - if (buff[0] == '\0') - luaX_syntaxerror(ls, s, ls->L->Mbuffer); - else - luaX_syntaxerror(ls, s, buff); +static void luaX_error (LexState *ls, const char *s, const char *token) { + luaX_errorline(ls, s, token, ls->linenumber); +} + + +void luaX_syntaxerror (LexState *ls, const char *msg) { + const char *lasttoken; + switch (ls->t.token) { + case TK_NAME: + lasttoken = getstr(ls->t.seminfo.ts); + break; + case TK_STRING: + case TK_NUMBER: + lasttoken = luaZ_buffer(ls->buff); + break; + default: + lasttoken = luaX_token2str(ls, ls->t.token); + break; + } + luaX_error(ls, msg, lasttoken); } -void luaX_token2str (int token, char *s) { - if (token < 256) { - s[0] = (char)token; - s[1] = '\0'; +const char *luaX_token2str (LexState *ls, int token) { + if (token < FIRST_RESERVED) { + lua_assert(token == (unsigned char)token); + return luaO_pushfstring(ls->L, "%c", token); } else - strcpy(s, token2string[token-FIRST_RESERVED]); + return token2string[token-FIRST_RESERVED]; } -static void luaX_invalidchar (LexState *ls, int c) { - char buff[8]; - sprintf(buff, "0x%02X", c); - luaX_syntaxerror(ls, "invalid control char", buff); +static void luaX_lexerror (LexState *ls, const char *s, int token) { + if (token == TK_EOS) + luaX_error(ls, s, luaX_token2str(ls, token)); + else + luaX_error(ls, s, luaZ_buffer(ls->buff)); } static void inclinenumber (LexState *LS) { - next(LS); /* skip '\n' */ + next(LS); /* skip `\n' */ ++LS->linenumber; luaX_checklimit(LS, LS->linenumber, MAX_INT, "lines in a chunk"); } @@ -122,159 +141,173 @@ void luaX_setinput (lua_State *L, LexState *LS, ZIO *z, TString *source) { */ -/* use Mbuffer to store names, literal strings and numbers */ +/* use buffer to store names, literal strings and numbers */ + +/* extra space to allocate when growing buffer */ +#define EXTRABUFF 32 + +/* maximum number of chars that can be read without checking buffer size */ +#define MAXNOCHECK 5 -#define EXTRABUFF 128 -#define checkbuffer(L, n, len) if ((len)+(n) > L->Mbuffsize) \ - luaO_openspace(L, (len)+(n)+EXTRABUFF) +#define checkbuffer(LS, len) \ + if (((len)+MAXNOCHECK)*sizeof(char) > luaZ_sizebuffer((LS)->buff)) \ + luaZ_openspace((LS)->L, (LS)->buff, (len)+EXTRABUFF) -#define save(L, c, l) (L->Mbuffer[l++] = (char)c) -#define save_and_next(L, LS, l) (save(L, LS->current, l), next(LS)) +#define save(LS, c, l) \ + (luaZ_buffer((LS)->buff)[l++] = cast(char, c)) +#define save_and_next(LS, l) (save(LS, LS->current, l), next(LS)) -static const char *readname (LexState *LS) { - lua_State *L = LS->L; +static size_t readname (LexState *LS) { size_t l = 0; - checkbuffer(L, 10, l); + checkbuffer(LS, l); do { - checkbuffer(L, 10, l); - save_and_next(L, LS, l); + checkbuffer(LS, l); + save_and_next(LS, l); } while (isalnum(LS->current) || LS->current == '_'); - save(L, '\0', l); - return L->Mbuffer; + save(LS, '\0', l); + return l-1; } /* LUA_NUMBER */ -static void read_number (LexState *LS, int comma, SemInfo *seminfo) { - lua_State *L = LS->L; +static void read_numeral (LexState *LS, int comma, SemInfo *seminfo) { size_t l = 0; - checkbuffer(L, 10, l); - if (comma) save(L, '.', l); + checkbuffer(LS, l); + if (comma) save(LS, '.', l); while (isdigit(LS->current)) { - checkbuffer(L, 10, l); - save_and_next(L, LS, l); + checkbuffer(LS, l); + save_and_next(LS, l); } if (LS->current == '.') { - save_and_next(L, LS, l); + save_and_next(LS, l); if (LS->current == '.') { - save_and_next(L, LS, l); - save(L, '\0', l); - luaX_error(LS, "ambiguous syntax" - " (decimal point x string concatenation)", TK_NUMBER); + save_and_next(LS, l); + save(LS, '\0', l); + luaX_lexerror(LS, + "ambiguous syntax (decimal point x string concatenation)", + TK_NUMBER); } } while (isdigit(LS->current)) { - checkbuffer(L, 10, l); - save_and_next(L, LS, l); + checkbuffer(LS, l); + save_and_next(LS, l); } if (LS->current == 'e' || LS->current == 'E') { - save_and_next(L, LS, l); /* read 'E' */ + save_and_next(LS, l); /* read `E' */ if (LS->current == '+' || LS->current == '-') - save_and_next(L, LS, l); /* optional exponent sign */ + save_and_next(LS, l); /* optional exponent sign */ while (isdigit(LS->current)) { - checkbuffer(L, 10, l); - save_and_next(L, LS, l); + checkbuffer(LS, l); + save_and_next(LS, l); } } - save(L, '\0', l); - if (!luaO_str2d(L->Mbuffer, &seminfo->r)) - luaX_error(LS, "malformed number", TK_NUMBER); + save(LS, '\0', l); + if (!luaO_str2d(luaZ_buffer(LS->buff), &seminfo->r)) + luaX_lexerror(LS, "malformed number", TK_NUMBER); } static void read_long_string (LexState *LS, SemInfo *seminfo) { - lua_State *L = LS->L; int cont = 0; size_t l = 0; - checkbuffer(L, 10, l); - save(L, '[', l); /* save first '[' */ - save_and_next(L, LS, l); /* pass the second '[' */ + checkbuffer(LS, l); + save(LS, '[', l); /* save first `[' */ + save_and_next(LS, l); /* pass the second `[' */ + if (LS->current == '\n') /* string starts with a newline? */ + inclinenumber(LS); /* skip it */ for (;;) { - checkbuffer(L, 10, l); + checkbuffer(LS, l); switch (LS->current) { case EOZ: - save(L, '\0', l); - luaX_error(LS, "unfinished long string", TK_STRING); + save(LS, '\0', l); + luaX_lexerror(LS, (seminfo) ? "unfinished long string" : + "unfinished long comment", TK_EOS); break; /* to avoid warnings */ case '[': - save_and_next(L, LS, l); + save_and_next(LS, l); if (LS->current == '[') { cont++; - save_and_next(L, LS, l); + save_and_next(LS, l); } continue; case ']': - save_and_next(L, LS, l); + save_and_next(LS, l); if (LS->current == ']') { if (cont == 0) goto endloop; cont--; - save_and_next(L, LS, l); + save_and_next(LS, l); } continue; case '\n': - save(L, '\n', l); + save(LS, '\n', l); inclinenumber(LS); + if (!seminfo) l = 0; /* reset buffer to avoid wasting space */ continue; default: - save_and_next(L, LS, l); + save_and_next(LS, l); } } endloop: - save_and_next(L, LS, l); /* skip the second ']' */ - save(L, '\0', l); - seminfo->ts = luaS_newlstr(L, L->Mbuffer+2, l-5); + save_and_next(LS, l); /* skip the second `]' */ + save(LS, '\0', l); + if (seminfo) + seminfo->ts = luaS_newlstr(LS->L, luaZ_buffer(LS->buff) + 2, l - 5); } static void read_string (LexState *LS, int del, SemInfo *seminfo) { - lua_State *L = LS->L; size_t l = 0; - checkbuffer(L, 10, l); - save_and_next(L, LS, l); + checkbuffer(LS, l); + save_and_next(LS, l); while (LS->current != del) { - checkbuffer(L, 10, l); + checkbuffer(LS, l); switch (LS->current) { - case EOZ: case '\n': - save(L, '\0', l); - luaX_error(LS, "unfinished string", TK_STRING); + case EOZ: + save(LS, '\0', l); + luaX_lexerror(LS, "unfinished string", TK_EOS); + break; /* to avoid warnings */ + case '\n': + save(LS, '\0', l); + luaX_lexerror(LS, "unfinished string", TK_STRING); break; /* to avoid warnings */ case '\\': - next(LS); /* do not save the '\' */ + next(LS); /* do not save the `\' */ switch (LS->current) { - case 'a': save(L, '\a', l); next(LS); break; - case 'b': save(L, '\b', l); next(LS); break; - case 'f': save(L, '\f', l); next(LS); break; - case 'n': save(L, '\n', l); next(LS); break; - case 'r': save(L, '\r', l); next(LS); break; - case 't': save(L, '\t', l); next(LS); break; - case 'v': save(L, '\v', l); next(LS); break; - case '\n': save(L, '\n', l); inclinenumber(LS); break; - case '0': case '1': case '2': case '3': case '4': - case '5': case '6': case '7': case '8': case '9': { - int c = 0; - int i = 0; - do { - c = 10*c + (LS->current-'0'); - next(LS); - } while (++i<3 && isdigit(LS->current)); - if (c != (unsigned char)c) { - save(L, '\0', l); - luaX_error(LS, "escape sequence too large", TK_STRING); + case 'a': save(LS, '\a', l); next(LS); break; + case 'b': save(LS, '\b', l); next(LS); break; + case 'f': save(LS, '\f', l); next(LS); break; + case 'n': save(LS, '\n', l); next(LS); break; + case 'r': save(LS, '\r', l); next(LS); break; + case 't': save(LS, '\t', l); next(LS); break; + case 'v': save(LS, '\v', l); next(LS); break; + case '\n': save(LS, '\n', l); inclinenumber(LS); break; + case EOZ: break; /* will raise an error next loop */ + default: { + if (!isdigit(LS->current)) + save_and_next(LS, l); /* handles \\, \", \', and \? */ + else { /* \xxx */ + int c = 0; + int i = 0; + do { + c = 10*c + (LS->current-'0'); + next(LS); + } while (++i<3 && isdigit(LS->current)); + if (c > UCHAR_MAX) { + save(LS, '\0', l); + luaX_lexerror(LS, "escape sequence too large", TK_STRING); + } + save(LS, c, l); } - save(L, c, l); - break; } - default: /* handles \\, \", \', and \? */ - save_and_next(L, LS, l); } break; default: - save_and_next(L, LS, l); + save_and_next(LS, l); } } - save_and_next(L, LS, l); /* skip delimiter */ - save(L, '\0', l); - seminfo->ts = luaS_newlstr(L, L->Mbuffer+1, l-3); + save_and_next(LS, l); /* skip delimiter */ + save(LS, '\0', l); + seminfo->ts = luaS_newlstr(LS->L, luaZ_buffer(LS->buff) + 1, l - 3); } @@ -282,58 +315,56 @@ int luaX_lex (LexState *LS, SemInfo *seminfo) { for (;;) { switch (LS->current) { - case ' ': case '\t': case '\r': /* `\r' to avoid problems with DOS */ - next(LS); - continue; - - case '\n': + case '\n': { inclinenumber(LS); continue; - - case '$': - luaX_error(LS, "unexpected `$' (pragmas are no longer supported)", '$'); - break; - - case '-': + } + case '-': { next(LS); if (LS->current != '-') return '-'; - do { next(LS); } while (LS->current != '\n' && LS->current != EOZ); + /* else is a comment */ + next(LS); + if (LS->current == '[' && (next(LS), LS->current == '[')) + read_long_string(LS, NULL); /* long comment */ + else /* short comment */ + while (LS->current != '\n' && LS->current != EOZ) + next(LS); continue; - - case '[': + } + case '[': { next(LS); if (LS->current != '[') return '['; else { read_long_string(LS, seminfo); return TK_STRING; } - - case '=': + } + case '=': { next(LS); if (LS->current != '=') return '='; else { next(LS); return TK_EQ; } - - case '<': + } + case '<': { next(LS); if (LS->current != '=') return '<'; else { next(LS); return TK_LE; } - - case '>': + } + case '>': { next(LS); if (LS->current != '=') return '>'; else { next(LS); return TK_GE; } - - case '~': + } + case '~': { next(LS); if (LS->current != '=') return '~'; else { next(LS); return TK_NE; } - + } case '"': - case '\'': + case '\'': { read_string(LS, LS->current, seminfo); return TK_STRING; - - case '.': + } + case '.': { next(LS); if (LS->current == '.') { next(LS); @@ -345,36 +376,42 @@ int luaX_lex (LexState *LS, SemInfo *seminfo) { } else if (!isdigit(LS->current)) return '.'; else { - read_number(LS, 1, seminfo); + read_numeral(LS, 1, seminfo); return TK_NUMBER; } - - case '0': case '1': case '2': case '3': case '4': - case '5': case '6': case '7': case '8': case '9': - read_number(LS, 0, seminfo); - return TK_NUMBER; - - case EOZ: + } + case EOZ: { return TK_EOS; - - case '_': goto tname; - - default: - if (!isalpha(LS->current)) { - int c = LS->current; - if (iscntrl(c)) - luaX_invalidchar(LS, c); + } + default: { + if (isspace(LS->current)) { next(LS); - return c; + continue; + } + else if (isdigit(LS->current)) { + read_numeral(LS, 0, seminfo); + return TK_NUMBER; } - tname: { /* identifier or reserved word */ - TString *ts = luaS_new(LS->L, readname(LS)); - if (ts->marked >= RESERVEDMARK) /* reserved word? */ - return ts->marked-RESERVEDMARK+FIRST_RESERVED; + else if (isalpha(LS->current) || LS->current == '_') { + /* identifier or reserved word */ + size_t l = readname(LS); + TString *ts = luaS_newlstr(LS->L, luaZ_buffer(LS->buff), l); + if (ts->tsv.reserved > 0) /* reserved word? */ + return ts->tsv.reserved - 1 + FIRST_RESERVED; seminfo->ts = ts; return TK_NAME; } + else { + int c = LS->current; + if (iscntrl(c)) + luaX_error(LS, "invalid control char", + luaO_pushfstring(LS->L, "char(%d)", c)); + next(LS); + return c; /* single-char tokens (+ - / ...) */ + } + } } } } +#undef next |