lua/llex.c

528 lines
15 KiB
C
Raw Normal View History

1997-09-16 23:25:59 +04:00
/*
2012-12-05 23:57:00 +04:00
** $Id: llex.c,v 2.61 2012/01/23 23:05:51 roberto Exp roberto $
1999-02-26 00:07:26 +03:00
** Lexical Analyzer
1997-09-16 23:25:59 +04:00
** See Copyright Notice in lua.h
*/
1993-12-23 00:15:16 +03:00
#include <locale.h>
#include <string.h>
1993-12-23 00:15:16 +03:00
2002-12-04 20:38:31 +03:00
#define llex_c
#define LUA_CORE
2002-12-04 20:38:31 +03:00
#include "lua.h"
#include "lctype.h"
#include "ldo.h"
1997-09-16 23:25:59 +04:00
#include "llex.h"
#include "lobject.h"
#include "lparser.h"
#include "lstate.h"
1997-09-16 23:25:59 +04:00
#include "lstring.h"
#include "ltable.h"
1997-09-16 23:25:59 +04:00
#include "lzio.h"
1993-12-23 00:15:16 +03:00
#define next(ls) (ls->current = zgetc(ls->z))
1997-09-16 23:25:59 +04:00
1997-09-16 23:25:59 +04:00
#define currIsNewline(ls) (ls->current == '\n' || ls->current == '\r')
1998-05-27 17:08:34 +04:00
1999-07-22 23:29:42 +04:00
/* ORDER RESERVED */
static const char *const luaX_tokens [] = {
"and", "break", "do", "else", "elseif",
2011-02-02 17:55:17 +03:00
"end", "false", "for", "function", "goto", "if",
"in", "local", "nil", "not", "or", "repeat",
2004-12-03 23:54:12 +03:00
"return", "then", "true", "until", "while",
2011-06-23 18:54:48 +04:00
"..", "...", "==", ">=", "<=", "~=", "::", "<eof>",
"<number>", "<name>", "<string>"
2001-02-23 20:17:25 +03:00
};
1998-05-27 17:08:34 +04:00
1997-09-16 23:25:59 +04:00
#define save_and_next(ls) (save(ls, ls->current), next(ls))
2011-11-30 16:43:51 +04:00
static l_noret lexerror (LexState *ls, const char *msg, int token);
2007-10-25 20:45:47 +04:00
static void save (LexState *ls, int c) {
Mbuffer *b = ls->buff;
if (luaZ_bufflen(b) + 1 > luaZ_sizebuffer(b)) {
size_t newsize;
if (luaZ_sizebuffer(b) >= MAX_SIZET/2)
2007-10-25 20:45:47 +04:00
lexerror(ls, "lexical element too long", 0);
newsize = luaZ_sizebuffer(b) * 2;
luaZ_resizebuffer(ls->L, b, newsize);
}
b->buffer[luaZ_bufflen(b)++] = cast(char, c);
}
void luaX_init (lua_State *L) {
2000-03-03 17:58:26 +03:00
int i;
2000-02-08 19:39:42 +03:00
for (i=0; i<NUM_RESERVED; i++) {
2004-12-02 15:59:10 +03:00
TString *ts = luaS_new(L, luaX_tokens[i]);
luaS_fix(ts); /* reserved words are never collected */
ts->tsv.extra = cast_byte(i+1); /* reserved word */
1998-05-27 17:08:34 +04:00
}
}
2003-09-05 00:00:28 +04:00
const char *luaX_token2str (LexState *ls, int token) {
2012-12-05 23:57:00 +04:00
if (token < FIRST_RESERVED) { /* single-byte symbols? */
2004-11-24 22:16:03 +03:00
lua_assert(token == cast(unsigned char, token));
return (lisprint(token)) ? luaO_pushfstring(ls->L, LUA_QL("%c"), token) :
2007-08-10 00:29:15 +04:00
luaO_pushfstring(ls->L, "char(%d)", token);
}
else {
const char *s = luaX_tokens[token - FIRST_RESERVED];
2012-12-05 23:57:00 +04:00
if (token < TK_EOS) /* fixed format (symbols and reserved words)? */
return luaO_pushfstring(ls->L, LUA_QS, s);
2012-12-05 23:57:00 +04:00
else /* names, strings, and numerals */
return s;
2003-09-05 00:00:28 +04:00
}
}
static const char *txtToken (LexState *ls, int token) {
switch (token) {
case TK_NAME:
case TK_STRING:
case TK_NUMBER:
save(ls, '\0');
return luaO_pushfstring(ls->L, LUA_QS, luaZ_buffer(ls->buff));
default:
2003-09-05 00:00:28 +04:00
return luaX_token2str(ls, token);
1997-09-16 23:25:59 +04:00
}
1998-05-27 17:08:34 +04:00
}
2011-11-30 16:43:51 +04:00
static l_noret lexerror (LexState *ls, const char *msg, int token) {
2008-12-26 14:55:57 +03:00
char buff[LUA_IDSIZE];
luaO_chunkid(buff, getstr(ls->source), LUA_IDSIZE);
2004-03-12 22:53:56 +03:00
msg = luaO_pushfstring(ls->L, "%s:%d: %s", buff, ls->linenumber, msg);
if (token)
luaO_pushfstring(ls->L, "%s near %s", msg, txtToken(ls, token));
luaD_throw(ls->L, LUA_ERRSYNTAX);
}
2011-11-30 16:43:51 +04:00
l_noret luaX_syntaxerror (LexState *ls, const char *msg) {
2007-10-25 20:45:47 +04:00
lexerror(ls, msg, ls->t.token);
1997-09-16 23:25:59 +04:00
}
/*
** creates a new string and anchors it in function's table so that
** it will not be collected until the end of the function's compilation
** (by that time it should be anchored in function's prototype)
*/
TString *luaX_newstring (LexState *ls, const char *str, size_t l) {
lua_State *L = ls->L;
TValue *o; /* entry for `str' */
TString *ts = luaS_newlstr(L, str, l); /* create new string */
setsvalue2s(L, L->top++, ts); /* temporarily anchor it in stack */
o = luaH_set(L, ls->fs->h, L->top - 1);
2011-08-11 21:06:12 +04:00
if (ttisnil(o)) { /* not in use yet? (see 'addK') */
2011-08-15 23:41:58 +04:00
/* boolean value does not need GC barrier;
table has no metatable, so it does not need to invalidate cache */
setbvalue(o, 1); /* t[string] = true */
2009-11-17 19:33:38 +03:00
luaC_checkGC(L);
}
L->top--; /* remove string from stack */
2003-08-28 01:01:44 +04:00
return ts;
}
/*
** increment line number and skips newline sequence (any of
** \n, \r, \n\r, or \r\n)
*/
static void inclinenumber (LexState *ls) {
int old = ls->current;
lua_assert(currIsNewline(ls));
next(ls); /* skip `\n' or `\r' */
if (currIsNewline(ls) && ls->current != old)
next(ls); /* skip `\n\r' or `\r\n' */
2004-03-12 22:53:56 +03:00
if (++ls->linenumber >= MAX_INT)
luaX_syntaxerror(ls, "chunk has too many lines");
2000-05-26 18:04:04 +04:00
}
2011-02-23 16:13:10 +03:00
void luaX_setinput (lua_State *L, LexState *ls, ZIO *z, TString *source,
int firstchar) {
ls->decpoint = '.';
ls->L = L;
2011-02-23 16:13:10 +03:00
ls->current = firstchar;
ls->lookahead.token = TK_EOS; /* no look-ahead token */
ls->z = z;
ls->fs = NULL;
ls->linenumber = 1;
ls->lastline = 1;
ls->source = source;
ls->envn = luaS_new(L, LUA_ENV); /* create env name */
luaS_fix(ls->envn); /* never collect this name */
luaZ_resizebuffer(ls->L, ls->buff, LUA_MINBUFFER); /* initialize buffer */
2000-05-26 18:04:04 +04:00
}
1997-09-16 23:25:59 +04:00
/*
** =======================================================
1999-12-14 21:33:29 +03:00
** LEXICAL ANALYZER
1997-09-16 23:25:59 +04:00
** =======================================================
*/
2000-09-11 21:38:42 +04:00
static int check_next (LexState *ls, const char *set) {
if (ls->current == '\0' || !strchr(set, ls->current))
return 0;
save_and_next(ls);
return 1;
}
2011-01-26 19:30:02 +03:00
/*
** change all characters 'from' in buffer to 'to'
*/
static void buffreplace (LexState *ls, char from, char to) {
2006-02-06 21:28:16 +03:00
size_t n = luaZ_bufflen(ls->buff);
char *p = luaZ_buffer(ls->buff);
while (n--)
if (p[n] == from) p[n] = to;
}
#if !defined(getlocaledecpoint)
#define getlocaledecpoint() (localeconv()->decimal_point[0])
#endif
#define buff2d(b,e) luaO_str2d(luaZ_buffer(b), luaZ_bufflen(b) - 1, e)
/*
** in case of format error, try to change decimal point separator to
** the one defined in the current locale and check again
*/
static void trydecpoint (LexState *ls, SemInfo *seminfo) {
char old = ls->decpoint;
ls->decpoint = getlocaledecpoint();
buffreplace(ls, old, ls->decpoint); /* try new decimal separator */
if (!buff2d(ls->buff, &seminfo->r)) {
/* format error with correct decimal point: no more options */
buffreplace(ls, ls->decpoint, '.'); /* undo change (for error message) */
2007-10-25 20:45:47 +04:00
lexerror(ls, "malformed number", TK_NUMBER);
}
}
2000-09-11 21:38:42 +04:00
/* LUA_NUMBER */
/*
** this function is quite liberal in what it accepts, as 'luaO_str2d'
** will reject ill-formed numerals.
*/
static void read_numeral (LexState *ls, SemInfo *seminfo) {
const char *expo = "Ee";
int first = ls->current;
lua_assert(lisdigit(ls->current));
save_and_next(ls);
if (first == '0' && check_next(ls, "Xx")) /* hexadecimal? */
expo = "Pp";
for (;;) {
if (check_next(ls, expo)) /* exponent part? */
check_next(ls, "+-"); /* optional exponent sign */
if (lisxdigit(ls->current) || ls->current == '.')
save_and_next(ls);
else break;
}
save(ls, '\0');
buffreplace(ls, '.', ls->decpoint); /* follow locale for decimal point */
if (!buff2d(ls->buff, &seminfo->r)) /* format error? */
trydecpoint(ls, seminfo); /* try to update decimal point separator */
2000-09-11 21:38:42 +04:00
}
/*
2010-10-25 16:24:36 +04:00
** skip a sequence '[=*[' or ']=*]' and return its number of '='s or
** -1 if sequence is malformed
*/
static int skip_sep (LexState *ls) {
int count = 0;
int s = ls->current;
lua_assert(s == '[' || s == ']');
save_and_next(ls);
while (ls->current == '=') {
save_and_next(ls);
count++;
}
return (ls->current == s) ? count : (-count) - 1;
}
static void read_long_string (LexState *ls, SemInfo *seminfo, int sep) {
save_and_next(ls); /* skip 2nd `[' */
if (currIsNewline(ls)) /* string starts with a newline? */
inclinenumber(ls); /* skip it */
1998-12-03 18:45:15 +03:00
for (;;) {
switch (ls->current) {
1997-06-16 20:50:22 +04:00
case EOZ:
2007-10-25 20:45:47 +04:00
lexerror(ls, (seminfo) ? "unfinished long string" :
"unfinished long comment", TK_EOS);
2000-01-25 21:44:21 +03:00
break; /* to avoid warnings */
case ']': {
if (skip_sep(ls) == sep) {
save_and_next(ls); /* skip 2nd `]' */
goto endloop;
}
break;
}
case '\n': case '\r': {
save(ls, '\n');
inclinenumber(ls);
if (!seminfo) luaZ_resetbuffer(ls->buff); /* avoid wasting space */
break;
}
default: {
if (seminfo) save_and_next(ls);
else next(ls);
}
}
} endloop:
2002-03-08 22:07:01 +03:00
if (seminfo)
seminfo->ts = luaX_newstring(ls, luaZ_buffer(ls->buff) + (2 + sep),
luaZ_bufflen(ls->buff) - 2*(2 + sep));
2000-01-25 21:44:21 +03:00
}
2011-07-08 23:17:30 +04:00
static void escerror (LexState *ls, int *c, int n, const char *msg) {
int i;
luaZ_resetbuffer(ls->buff); /* prepare error message */
save(ls, '\\');
for (i = 0; i < n && c[i] != EOZ; i++)
save(ls, c[i]);
lexerror(ls, msg, TK_STRING);
}
static int readhexaesc (LexState *ls) {
2011-07-15 16:30:41 +04:00
int c[3], i; /* keep input for error message */
int r = 0; /* result accumulator */
c[0] = 'x'; /* for error message */
for (i = 1; i < 3; i++) { /* read two hexa digits */
c[i] = next(ls);
if (!lisxdigit(c[i]))
escerror(ls, c, i + 1, "hexadecimal digit expected");
r = (r << 4) + luaO_hexavalue(c[i]);
}
2011-07-15 16:30:41 +04:00
return r;
}
static int readdecesc (LexState *ls) {
2011-07-15 16:30:41 +04:00
int c[3], i;
int r = 0; /* result accumulator */
for (i = 0; i < 3 && lisdigit(ls->current); i++) { /* read up to 3 digits */
c[i] = ls->current;
r = 10*r + c[i] - '0';
next(ls);
}
2011-07-15 16:30:41 +04:00
if (r > UCHAR_MAX)
escerror(ls, c, i, "decimal escape too large");
2011-07-08 23:17:30 +04:00
return r;
}
static void read_string (LexState *ls, int del, SemInfo *seminfo) {
save_and_next(ls); /* keep delimiter (for error messages) */
while (ls->current != del) {
switch (ls->current) {
case EOZ:
2007-10-25 20:45:47 +04:00
lexerror(ls, "unfinished string", TK_EOS);
2010-02-28 00:16:24 +03:00
break; /* to avoid warnings */
case '\n':
case '\r':
2007-10-25 20:45:47 +04:00
lexerror(ls, "unfinished string", TK_STRING);
2010-02-28 00:16:24 +03:00
break; /* to avoid warnings */
case '\\': { /* escape sequences */
int c; /* final character to be saved */
next(ls); /* do not save the `\' */
switch (ls->current) {
case 'a': c = '\a'; goto read_save;
case 'b': c = '\b'; goto read_save;
case 'f': c = '\f'; goto read_save;
case 'n': c = '\n'; goto read_save;
case 'r': c = '\r'; goto read_save;
case 't': c = '\t'; goto read_save;
case 'v': c = '\v'; goto read_save;
case 'x': c = readhexaesc(ls); goto read_save;
case '\n': case '\r':
inclinenumber(ls); c = '\n'; goto only_save;
case '\\': case '\"': case '\'':
c = ls->current; goto read_save;
case EOZ: goto no_save; /* will raise an error next loop */
case 'z': { /* zap following span of spaces */
2011-07-06 20:45:14 +04:00
next(ls); /* skip the 'z' */
while (lisspace(ls->current)) {
if (currIsNewline(ls)) inclinenumber(ls);
else next(ls);
}
goto no_save;
}
2001-02-23 20:17:25 +03:00
default: {
if (!lisdigit(ls->current))
2011-07-08 23:17:30 +04:00
escerror(ls, &ls->current, 1, "invalid escape sequence");
/* digital escape \ddd */
c = readdecesc(ls);
goto only_save;
2000-01-25 21:44:21 +03:00
}
}
read_save: next(ls); /* read next character */
only_save: save(ls, c); /* save 'c' */
no_save: break;
}
2000-01-25 21:44:21 +03:00
default:
save_and_next(ls);
2000-01-25 21:44:21 +03:00
}
}
save_and_next(ls); /* skip delimiter */
seminfo->ts = luaX_newstring(ls, luaZ_buffer(ls->buff) + 1,
luaZ_bufflen(ls->buff) - 2);
}
1997-09-16 23:25:59 +04:00
2005-12-07 18:43:05 +03:00
static int llex (LexState *ls, SemInfo *seminfo) {
luaZ_resetbuffer(ls->buff);
1998-12-03 18:45:15 +03:00
for (;;) {
switch (ls->current) {
case '\n': case '\r': { /* line breaks */
inclinenumber(ls);
2010-02-28 00:16:24 +03:00
break;
}
case ' ': case '\f': case '\t': case '\v': { /* spaces */
next(ls);
break;
}
case '-': { /* '-' or '--' (comment) */
next(ls);
if (ls->current != '-') return '-';
2002-03-08 22:07:01 +03:00
/* else is a comment */
next(ls);
if (ls->current == '[') { /* long comment? */
int sep = skip_sep(ls);
luaZ_resetbuffer(ls->buff); /* `skip_sep' may dirty the buffer */
if (sep >= 0) {
read_long_string(ls, NULL, sep); /* skip long comment */
luaZ_resetbuffer(ls->buff); /* previous call may dirty the buff. */
2010-02-28 00:16:24 +03:00
break;
}
}
/* else short comment */
while (!currIsNewline(ls) && ls->current != EOZ)
next(ls); /* skip until end of line (or end of file) */
2010-02-28 00:16:24 +03:00
break;
}
case '[': { /* long string or simply '[' */
int sep = skip_sep(ls);
if (sep >= 0) {
read_long_string(ls, seminfo, sep);
2000-03-10 21:37:44 +03:00
return TK_STRING;
}
else if (sep == -1) return '[';
2007-10-25 20:45:47 +04:00
else lexerror(ls, "invalid long string delimiter", TK_STRING);
}
case '=': {
next(ls);
if (ls->current != '=') return '=';
else { next(ls); return TK_EQ; }
}
case '<': {
next(ls);
if (ls->current != '=') return '<';
else { next(ls); return TK_LE; }
}
case '>': {
next(ls);
if (ls->current != '=') return '>';
else { next(ls); return TK_GE; }
}
case '~': {
next(ls);
if (ls->current != '=') return '~';
else { next(ls); return TK_NE; }
}
2011-06-23 18:54:48 +04:00
case ':': {
next(ls);
if (ls->current != ':') return ':';
else { next(ls); return TK_DBCOLON; }
}
case '"': case '\'': { /* short literal strings */
read_string(ls, ls->current, seminfo);
2000-03-10 21:37:44 +03:00
return TK_STRING;
}
case '.': { /* '.', '..', '...', or number */
save_and_next(ls);
if (check_next(ls, ".")) {
if (check_next(ls, "."))
return TK_DOTS; /* '...' */
else return TK_CONCAT; /* '..' */
1993-12-23 00:15:16 +03:00
}
else if (!lisdigit(ls->current)) return '.';
/* else go through */
}
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9': {
read_numeral(ls, seminfo);
return TK_NUMBER;
}
case EOZ: {
2000-03-10 21:37:44 +03:00
return TK_EOS;
}
2001-02-23 20:17:25 +03:00
default: {
if (lislalpha(ls->current)) { /* identifier or reserved word? */
TString *ts;
do {
save_and_next(ls);
} while (lislalnum(ls->current));
ts = luaX_newstring(ls, luaZ_buffer(ls->buff),
luaZ_bufflen(ls->buff));
seminfo->ts = ts;
if (isreserved(ts)) /* reserved word? */
return ts->tsv.extra - 1 + FIRST_RESERVED;
2003-09-05 00:00:28 +04:00
else {
return TK_NAME;
}
1997-07-01 23:32:41 +04:00
}
else { /* single-char tokens (+ - / ...) */
int c = ls->current;
next(ls);
return c;
2001-02-23 20:17:25 +03:00
}
}
1993-12-23 00:15:16 +03:00
}
}
}
2005-12-07 18:43:05 +03:00
void luaX_next (LexState *ls) {
ls->lastline = ls->linenumber;
if (ls->lookahead.token != TK_EOS) { /* is there a look-ahead token? */
ls->t = ls->lookahead; /* use this one */
ls->lookahead.token = TK_EOS; /* and discharge it */
}
else
ls->t.token = llex(ls, &ls->t.seminfo); /* read next token */
}
2007-05-11 21:28:56 +04:00
int luaX_lookahead (LexState *ls) {
2005-12-07 18:43:05 +03:00
lua_assert(ls->lookahead.token == TK_EOS);
ls->lookahead.token = llex(ls, &ls->lookahead.seminfo);
2007-05-11 21:28:56 +04:00
return ls->lookahead.token;
2005-12-07 18:43:05 +03:00
}