0e78042e07
[DEVELOPMENT ENVIRONMENT] * NEW: Update French translation. * NEW: Support of the new '.list' file format. * NEW: Support of the new EXPORT AS syntax. * NEW: Update error messages. [COMPILER] * NEW: EXPORT AS is a new syntax for exporting a class with a specific name that can include a namespace. * NEW: New header file to start sharing reader code with the 'gb.eval' component. * NEW: Class names can include a colon. The part before the colon is assumed to be a namespace. [INTERPRETER] * NEW: Classes exported in a namespace are now declared both in the global symbol table and in the local symbol table of their component. * NEW: Class names are now allocated inside the symbol tables, and not in the class objects anymore. [GB.EVAL] * NEW: Share some reader code with the compiler.
1117 lines
19 KiB
C
1117 lines
19 KiB
C
/***************************************************************************
|
|
|
|
eval_read.c
|
|
|
|
(c) 2000-2017 Benoît Minisini <g4mba5@gmail.com>
|
|
|
|
This program is free software; you can redistribute it and/or modify
|
|
it under the terms of the GNU General Public License as published by
|
|
the Free Software Foundation; either version 2, or (at your option)
|
|
any later version.
|
|
|
|
This program is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with this program; if not, write to the Free Software
|
|
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
|
|
MA 02110-1301, USA.
|
|
|
|
***************************************************************************/
|
|
|
|
#define __EVAL_READ_C
|
|
|
|
#include <ctype.h>
|
|
|
|
#include "gb_common.h"
|
|
#include "gb_error.h"
|
|
#include "gb_table.h"
|
|
|
|
#include "eval.h"
|
|
#include "eval_read.h"
|
|
|
|
//#define DEBUG 1
|
|
//#define ENABLE_BIG_COMMENT
|
|
|
|
PUBLIC const char *READ_source_ptr;
|
|
#define source_ptr READ_source_ptr
|
|
|
|
static bool is_init = FALSE;
|
|
static bool _begin_line = FALSE;
|
|
|
|
static char ident_car[256];
|
|
static char first_car[256];
|
|
static char digit_car[256];
|
|
static char noop_car[256];
|
|
static char canres_car[256];
|
|
|
|
#undef isspace
|
|
#define isspace(_c) (((uchar)_c) <= ' ')
|
|
#undef isdigit
|
|
#define isdigit(_c) (digit_car[_c])
|
|
|
|
enum
|
|
{
|
|
GOTO_BREAK,
|
|
GOTO_SPACE,
|
|
GOTO_NEWLINE,
|
|
GOTO_COMMENT,
|
|
GOTO_STRING,
|
|
GOTO_IDENT,
|
|
GOTO_QUOTED_IDENT,
|
|
GOTO_NUMBER,
|
|
GOTO_ERROR,
|
|
GOTO_SHARP,
|
|
GOTO_OTHER
|
|
};
|
|
|
|
static void READ_init(void)
|
|
{
|
|
unsigned char i;
|
|
|
|
if (!is_init)
|
|
{
|
|
for (i = 0; i < 255; i++)
|
|
{
|
|
ident_car[i] = (i != 0) && ((i >= 'A' && i <= 'Z') || (i >= 'a' && i <= 'z') || (i >= '0' && i <= '9') || strchr("$_?@", i));
|
|
digit_car[i] = (i >= '0' && i <= '9');
|
|
noop_car[i] = ident_car[i] || digit_car[i] || i <= ' ';
|
|
canres_car[i] = (i != ':') && (i != '.') && (i != '!') && (i != '(');
|
|
|
|
if (i == 0)
|
|
first_car[i] = GOTO_BREAK;
|
|
else if (i == '\n')
|
|
first_car[i] = GOTO_NEWLINE;
|
|
else if (i <= ' ')
|
|
first_car[i] = GOTO_SPACE;
|
|
else if (i == '\'')
|
|
first_car[i] = GOTO_COMMENT;
|
|
else if (i == '"')
|
|
first_car[i] = GOTO_STRING;
|
|
else if (i == '#')
|
|
first_car[i] = GOTO_SHARP;
|
|
else if ((i >= 'A' && i <= 'Z') || (i >= 'a' && i <= 'z') || i == '$' || i == '_')
|
|
first_car[i] = GOTO_IDENT;
|
|
else if (i == '{')
|
|
first_car[i] = GOTO_QUOTED_IDENT;
|
|
else if (i >= '0' && i <= '9')
|
|
first_car[i] = GOTO_NUMBER;
|
|
else if (i >= 127)
|
|
first_car[i] = GOTO_ERROR;
|
|
else
|
|
first_car[i] = GOTO_OTHER;
|
|
}
|
|
|
|
is_init = TRUE;
|
|
}
|
|
}
|
|
|
|
static bool _no_quote = FALSE;
|
|
#define BUF_MAX 255
|
|
static char _buffer[BUF_MAX + 1];
|
|
|
|
PUBLIC char *READ_get_pattern(PATTERN *pattern)
|
|
{
|
|
int type = PATTERN_type(*pattern);
|
|
int index = PATTERN_index(*pattern);
|
|
const char *str;
|
|
const char *before = _no_quote ? "" : "'";
|
|
const char *after = _no_quote ? "" : "'";
|
|
|
|
switch(type)
|
|
{
|
|
case RT_RESERVED:
|
|
//snprintf(COMMON_buffer, COMMON_BUF_MAX, "%s%s%s", before, TABLE_get_symbol_name(COMP_res_table, index), after);
|
|
str = COMP_res_info[index].name;
|
|
if (ispunct((unsigned char)*str))
|
|
snprintf(_buffer, BUF_MAX, "%s%s%s", before, str, after);
|
|
else
|
|
strcpy(_buffer, str);
|
|
break;
|
|
|
|
case RT_NUMBER:
|
|
case RT_IDENTIFIER:
|
|
case RT_CLASS:
|
|
snprintf(_buffer, BUF_MAX, "%s%s%s", before, TABLE_get_symbol_name(EVAL->table, index), after);
|
|
break;
|
|
|
|
case RT_STRING:
|
|
case RT_TSTRING:
|
|
if (_no_quote)
|
|
snprintf(_buffer, BUF_MAX, "\"%s\"", TABLE_get_symbol_name(EVAL->string, index));
|
|
else
|
|
strcpy(_buffer, "string");
|
|
break;
|
|
|
|
case RT_NEWLINE:
|
|
case RT_END:
|
|
strcpy(_buffer, "end of expression");
|
|
break;
|
|
|
|
case RT_SUBR:
|
|
//snprintf(COMMON_buffer, COMMON_BUF_MAX, "%s%s%s", bafore, COMP_subr_info[index].name, after);
|
|
strcpy(_buffer, COMP_subr_info[index].name);
|
|
break;
|
|
|
|
case RT_COMMENT:
|
|
strncpy(_buffer, TABLE_get_symbol_name(EVAL->string, index), BUF_MAX);
|
|
_buffer[BUF_MAX] = 0;
|
|
break;
|
|
|
|
case RT_SPACE:
|
|
snprintf(_buffer, BUF_MAX, "[%d]", index);
|
|
break;
|
|
|
|
default:
|
|
sprintf(_buffer, "%s?%08X?%s", before, *pattern, after);
|
|
}
|
|
|
|
return _buffer;
|
|
}
|
|
|
|
#if DEBUG
|
|
PUBLIC void READ_dump_pattern(PATTERN *pattern)
|
|
{
|
|
int type = PATTERN_type(*pattern);
|
|
int index = PATTERN_index(*pattern);
|
|
int pos;
|
|
|
|
pos = (int)(pattern - EVAL->pattern);
|
|
if (pos < 0 || pos >= EVAL->pattern_count)
|
|
return;
|
|
|
|
printf("%d ", pos);
|
|
|
|
if (PATTERN_flag(*pattern) & RT_FIRST)
|
|
printf("!");
|
|
else
|
|
printf(" ");
|
|
|
|
if (PATTERN_flag(*pattern) & RT_POINT)
|
|
printf(".");
|
|
else
|
|
printf(" ");
|
|
|
|
if (PATTERN_flag(*pattern) & RT_CLASS)
|
|
printf("C");
|
|
else
|
|
printf(" ");
|
|
|
|
printf(" ");
|
|
|
|
_no_quote = TRUE;
|
|
|
|
if (type == RT_RESERVED)
|
|
printf("RESERVED %s\n", READ_get_pattern(pattern));
|
|
else if (type == RT_NUMBER)
|
|
printf("NUMBER %s\n", READ_get_pattern(pattern));
|
|
else if (type == RT_IDENTIFIER)
|
|
printf("IDENTIFIER %s\n", READ_get_pattern(pattern));
|
|
else if (type == RT_CLASS)
|
|
printf("CLASS %s\n", READ_get_pattern(pattern));
|
|
else if (type == RT_STRING)
|
|
printf("STRING %s\n", READ_get_pattern(pattern));
|
|
else if (type == RT_TSTRING)
|
|
printf("TSTRING %s\n", READ_get_pattern(pattern));
|
|
else if (type == RT_NEWLINE)
|
|
printf("NEWLINE (%d)\n", index);
|
|
else if (type == RT_END)
|
|
printf("END\n");
|
|
else if (type == RT_PARAM)
|
|
printf("PARAM %d\n", index);
|
|
else if (type == RT_SUBR)
|
|
printf("SUBR %s\n", READ_get_pattern(pattern));
|
|
else if (type == RT_COMMENT)
|
|
printf("COMMENT %s\n", READ_get_pattern(pattern));
|
|
else if (type == RT_SPACE)
|
|
printf("SPACE %s\n", READ_get_pattern(pattern));
|
|
else
|
|
printf("? %d\n", index);
|
|
|
|
_no_quote = FALSE;
|
|
}
|
|
#endif
|
|
|
|
#define get_char_offset(_offset) ((unsigned char)source_ptr[(_offset)])
|
|
#define get_char() ((unsigned char)(*source_ptr))
|
|
|
|
static unsigned char next_char(void)
|
|
{
|
|
source_ptr++;
|
|
return get_char();
|
|
}
|
|
|
|
#ifdef DEBUG
|
|
|
|
static void add_pattern(int type, int index)
|
|
{
|
|
EVAL->pattern[EVAL->pattern_count++] = PATTERN_make(type, index);
|
|
READ_dump_pattern(&EVAL->pattern[EVAL->pattern_count - 1]);
|
|
}
|
|
|
|
#else
|
|
|
|
#define add_pattern(_type, _index) EVAL->pattern[EVAL->pattern_count++] = PATTERN_make((_type), (_index));
|
|
|
|
#endif
|
|
|
|
static PATTERN get_previous_pattern(int n)
|
|
{
|
|
int i;
|
|
PATTERN pattern;
|
|
|
|
for (i = EVAL->pattern_count - 1; i >= 0; i--)
|
|
{
|
|
pattern = EVAL->pattern[i];
|
|
if (!PATTERN_is_space(pattern))
|
|
{
|
|
n--;
|
|
if (n == 0)
|
|
return pattern;
|
|
}
|
|
}
|
|
|
|
return NULL_PATTERN;
|
|
}
|
|
|
|
#define get_last_pattern() get_previous_pattern(1)
|
|
#define get_last_last_pattern() get_previous_pattern(2)
|
|
|
|
static void add_newline()
|
|
{
|
|
add_pattern(RT_NEWLINE, 0);
|
|
}
|
|
|
|
|
|
static void add_end()
|
|
{
|
|
add_pattern(RT_END, 0);
|
|
}
|
|
|
|
|
|
#include "gbc_read_temp.h"
|
|
|
|
#if 0
|
|
static void add_identifier()
|
|
{
|
|
unsigned char car;
|
|
const char *start;
|
|
int len;
|
|
int index;
|
|
int type;
|
|
int flag;
|
|
PATTERN last_pattern, last_last_pattern;
|
|
bool not_first;
|
|
bool can_be_reserved;
|
|
bool last_identifier, last_type, last_class, last_pub;
|
|
bool exist = TRUE;
|
|
|
|
last_pattern = get_last_pattern();
|
|
|
|
if (PATTERN_is_reserved(last_pattern))
|
|
{
|
|
flag = RES_get_ident_flag(PATTERN_index(last_pattern));
|
|
if (flag & RSF_PREV)
|
|
{
|
|
last_last_pattern = get_last_last_pattern();
|
|
if (PATTERN_is_reserved(last_last_pattern))
|
|
flag = RES_get_ident_flag(PATTERN_index(last_last_pattern));
|
|
else
|
|
flag = 0;
|
|
}
|
|
}
|
|
else
|
|
flag = 0;
|
|
|
|
type = RT_IDENTIFIER;
|
|
|
|
last_class = (flag & RSF_CLASS) != 0;
|
|
last_type = (flag & RSF_AS) != 0;
|
|
|
|
start = source_ptr;
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
if (!ident_car[get_char()])
|
|
break;
|
|
}
|
|
|
|
if (last_type && get_char() == ':')
|
|
{
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
if (!ident_car[get_char()])
|
|
break;
|
|
}
|
|
}
|
|
|
|
len = source_ptr - start;
|
|
|
|
if (last_type)
|
|
{
|
|
source_ptr--;
|
|
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
len++;
|
|
car = get_char();
|
|
if (car == '[')
|
|
{
|
|
car = get_char_offset(1);
|
|
if (car == ']')
|
|
{
|
|
source_ptr++;
|
|
len++;
|
|
index = TABLE_add_symbol(EVAL->table, start, len - 2);
|
|
continue;
|
|
}
|
|
}
|
|
|
|
len--;
|
|
break;
|
|
}
|
|
}
|
|
|
|
not_first = (flag & RSF_POINT) != 0;
|
|
|
|
car = get_char();
|
|
|
|
//can_be_reserved = !not_first && TABLE_find_symbol(COMP_res_table, &comp->source[start], len, NULL, &index);
|
|
|
|
can_be_reserved = !not_first && !last_class;
|
|
|
|
if (can_be_reserved)
|
|
{
|
|
index = RESERVED_find_word(start, len);
|
|
can_be_reserved = (index >= 0);
|
|
}
|
|
|
|
if (can_be_reserved)
|
|
{
|
|
static void *jump[] = {
|
|
&&__OTHERS, &&__ME_NEW_LAST_SUPER, &&__CLASS, &&__STRUCT, &&__SUB_PROCEDURE_FUNCTION, &&__CONST_EXTERN, &&__ENUM, &&__READ, &&__DATATYPE
|
|
};
|
|
|
|
last_identifier = (flag & RSF_IDENT) != 0;
|
|
last_pub = (flag & RSF_PUB) != 0;
|
|
|
|
goto *jump[RES_get_read_switch(index)];
|
|
|
|
do
|
|
{
|
|
__ME_NEW_LAST_SUPER:
|
|
can_be_reserved = !last_identifier;
|
|
break;
|
|
|
|
__CLASS:
|
|
can_be_reserved = canres_car[car] && (_begin_line || PATTERN_is(last_pattern, RS_END));
|
|
break;
|
|
|
|
__STRUCT:
|
|
can_be_reserved = canres_car[car] && (_begin_line || last_pub || PATTERN_is(last_pattern, RS_AS) || PATTERN_is(last_pattern, RS_END) || PATTERN_is(last_pattern, RS_NEW));
|
|
break;
|
|
|
|
__SUB_PROCEDURE_FUNCTION:
|
|
can_be_reserved = canres_car[car] && (_begin_line || last_pub || PATTERN_is(last_pattern, RS_END));
|
|
break;
|
|
|
|
__CONST_EXTERN:
|
|
can_be_reserved = canres_car[car] && (_begin_line || last_pub);
|
|
break;
|
|
|
|
__ENUM:
|
|
can_be_reserved = canres_car[car] && (_begin_line || last_pub);
|
|
break;
|
|
|
|
__READ:
|
|
can_be_reserved = canres_car[car] && (!last_identifier || PATTERN_is(last_pattern, RS_PROPERTY));
|
|
break;
|
|
|
|
__DATATYPE:
|
|
if (car == '[' && get_char_offset(1) == ']')
|
|
{
|
|
len += 2;
|
|
source_ptr += 2;
|
|
can_be_reserved = FALSE;
|
|
}
|
|
else
|
|
{
|
|
if (last_type || PATTERN_is(last_pattern, RS_OPEN))
|
|
can_be_reserved = TRUE;
|
|
else
|
|
can_be_reserved = FALSE;
|
|
}
|
|
break;
|
|
|
|
__OTHERS:
|
|
if (last_type || last_identifier || (PATTERN_is(last_pattern, RS_LBRA) && car == ')' && PATTERN_is_reserved(get_last_last_pattern())))
|
|
can_be_reserved = FALSE;
|
|
else
|
|
can_be_reserved = canres_car[car];
|
|
break;
|
|
}
|
|
while (0);
|
|
}
|
|
|
|
if (can_be_reserved)
|
|
{
|
|
type = RT_RESERVED;
|
|
goto __ADD_PATTERN;
|
|
}
|
|
|
|
if ((flag == 0) && car != '.' && car != '!')
|
|
{
|
|
index = RESERVED_find_subr(start, len);
|
|
if (index >= 0)
|
|
{
|
|
if (COMP_subr_info[index].min_param == 0 || car == '(')
|
|
{
|
|
type = RT_SUBR;
|
|
|
|
if (EVAL->custom)
|
|
{
|
|
GB_FUNCTION func;
|
|
GB_VALUE *ret;
|
|
|
|
if (!GB.GetFunction(&func, (void *)GB.GetClass(EVAL->parent), "IsSubr", NULL, NULL))
|
|
{
|
|
GB.Push(1, GB_T_STRING, start, len);
|
|
ret = GB.Call(&func, 1, FALSE);
|
|
if (!ret->_boolean.value)
|
|
{
|
|
exist = TABLE_add_symbol_exist(EVAL->table, start, len, &index);
|
|
type = RT_IDENTIFIER;
|
|
}
|
|
}
|
|
}
|
|
|
|
goto __ADD_PATTERN;
|
|
}
|
|
}
|
|
}
|
|
|
|
if (last_type)
|
|
type = RT_CLASS;
|
|
|
|
/*if (flag & RSF_EVENT)
|
|
{
|
|
start--;
|
|
len++;
|
|
*((char *)start) = ':';
|
|
}*/
|
|
|
|
if (!EVAL->analyze && PATTERN_is(last_pattern, RS_EXCL))
|
|
{
|
|
index = TABLE_add_symbol(EVAL->string, start, len);
|
|
type = RT_STRING;
|
|
}
|
|
else
|
|
{
|
|
exist = TABLE_add_symbol_exist(EVAL->table, start, len, &index);
|
|
}
|
|
|
|
__ADD_PATTERN:
|
|
|
|
if (EVAL->custom && !exist)
|
|
{
|
|
GB_FUNCTION func;
|
|
GB_VALUE *ret;
|
|
|
|
if (!GB.GetFunction(&func, (void *)GB.GetClass(EVAL->parent), "IsIdentifier", NULL, NULL))
|
|
{
|
|
GB.Push(1, GB_T_STRING, start, len);
|
|
ret = GB.Call(&func, 1, FALSE);
|
|
if (!ret->_boolean.value)
|
|
THROW("Unknown symbol");
|
|
}
|
|
}
|
|
|
|
add_pattern(type, index);
|
|
}
|
|
#endif
|
|
|
|
static void add_quoted_identifier(void)
|
|
{
|
|
unsigned char car;
|
|
const char *start;
|
|
int len;
|
|
int index;
|
|
int type;
|
|
PATTERN last_pattern;
|
|
|
|
last_pattern = get_last_pattern();
|
|
|
|
type = RT_IDENTIFIER;
|
|
|
|
start = source_ptr;
|
|
len = 1;
|
|
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
car = get_char();
|
|
len++;
|
|
if (!ident_car[car])
|
|
break;
|
|
}
|
|
|
|
source_ptr++;
|
|
|
|
if (!EVAL->analyze)
|
|
{
|
|
if (car != '}')
|
|
THROW("Missing '}'");
|
|
|
|
if (len == 2)
|
|
THROW("Void identifier");
|
|
}
|
|
else
|
|
{
|
|
if (!car)
|
|
len--;
|
|
|
|
if (car != '}' || len <= 2)
|
|
type = RT_ERROR;
|
|
}
|
|
|
|
if (!EVAL->analyze && PATTERN_is(last_pattern, RS_EXCL))
|
|
{
|
|
index = TABLE_add_symbol(EVAL->string, start + 1, len - 2);
|
|
type = RT_STRING;
|
|
}
|
|
else
|
|
{
|
|
if (!EVAL->rewrite && type != RT_ERROR)
|
|
{
|
|
start++;
|
|
len -= 2;
|
|
}
|
|
index = TABLE_add_symbol(type == RT_ERROR ? EVAL->string : EVAL->table, start, len);
|
|
}
|
|
|
|
add_pattern(type, index);
|
|
}
|
|
|
|
|
|
static void add_operator()
|
|
{
|
|
unsigned char car;
|
|
const char *start;
|
|
const char *end;
|
|
int len;
|
|
int op = NO_SYMBOL;
|
|
int index;
|
|
|
|
start = source_ptr;
|
|
end = start;
|
|
len = 1;
|
|
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
|
|
index = RESERVED_find_word(start, len);
|
|
if (index >= 0)
|
|
{
|
|
op = index;
|
|
end = source_ptr;
|
|
}
|
|
|
|
car = get_char();
|
|
//if (!isascii(car) || !ispunct(car))
|
|
if (noop_car[car])
|
|
break;
|
|
len++;
|
|
}
|
|
|
|
source_ptr = end;
|
|
|
|
if (EVAL->analyze && op == RS_QUES)
|
|
op = RS_PRINT;
|
|
|
|
if (op < 0)
|
|
THROW("Unknown operator");
|
|
|
|
add_pattern(RT_RESERVED, op);
|
|
}
|
|
|
|
|
|
static int xdigit_val(unsigned char c)
|
|
{
|
|
if (c >= '0' && c <= '9')
|
|
return (c - '0');
|
|
else if (c >= 'a' && c <= 'f')
|
|
return (c - 'a' + 10);
|
|
else if (c >= 'A' && c <= 'F')
|
|
return (c - 'A' + 10);
|
|
else
|
|
return (-1);
|
|
}
|
|
|
|
|
|
static void add_string()
|
|
{
|
|
unsigned char car;
|
|
const char *start;
|
|
int len;
|
|
int index;
|
|
ushort newline;
|
|
bool jump;
|
|
char *p;
|
|
int i;
|
|
|
|
start = source_ptr;
|
|
len = 0;
|
|
newline = 0;
|
|
jump = FALSE;
|
|
p = (char *)source_ptr;
|
|
|
|
//fprintf(stderr, "EVAL_read: add_string: [%d] %s\n", source_ptr - EVAL->source, source_ptr);
|
|
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
car = get_char();
|
|
|
|
//fprintf(stderr, "EVAL_read: car = %d jump = %d\n", car, jump);
|
|
|
|
if (jump)
|
|
{
|
|
if (car == '\n')
|
|
newline++;
|
|
else if (car == '"')
|
|
jump = FALSE;
|
|
else if (!car || !isspace(car))
|
|
break;
|
|
}
|
|
else
|
|
{
|
|
p++;
|
|
len++;
|
|
|
|
if (!car || car == '\n')
|
|
THROW("Non terminated string");
|
|
|
|
if (car == '\\')
|
|
{
|
|
source_ptr++;
|
|
car = get_char();
|
|
|
|
if (car == 'n')
|
|
*p = '\n';
|
|
else if (car == 't')
|
|
*p = '\t';
|
|
else if (car == 'r')
|
|
*p = '\r';
|
|
else if (car == 'b')
|
|
*p = '\b';
|
|
else if (car == 'v')
|
|
*p = '\v';
|
|
else if (car == 'f')
|
|
*p = '\f';
|
|
else if (car == 'e')
|
|
*p = '\x1B';
|
|
else if (car == '0')
|
|
*p = 0;
|
|
else if (car == '\"' || car == '\'' || car == '\\')
|
|
*p = car;
|
|
else
|
|
{
|
|
if (car == 'x')
|
|
{
|
|
i = xdigit_val(get_char_offset(1));
|
|
if (i >= 0)
|
|
{
|
|
car = i;
|
|
i = xdigit_val(get_char_offset(2));
|
|
if (i >= 0)
|
|
{
|
|
car = (car << 4) | (uchar)i;
|
|
*p = car;
|
|
source_ptr += 2;
|
|
continue;
|
|
}
|
|
}
|
|
}
|
|
|
|
THROW("Bad character constant in string");
|
|
}
|
|
}
|
|
else if (car == '"')
|
|
{
|
|
p--;
|
|
len--;
|
|
jump = TRUE;
|
|
}
|
|
else
|
|
*p = car;
|
|
}
|
|
}
|
|
|
|
p[1] = 0;
|
|
|
|
//fprintf(stderr, "EVAL_read: add_string (end): [%d] %s\n", source_ptr - EVAL->source, source_ptr);
|
|
if (len > 0)
|
|
{
|
|
index = TABLE_add_symbol(EVAL->string, start + 1, len);
|
|
add_pattern(RT_STRING, index);
|
|
}
|
|
else
|
|
add_pattern(RT_STRING, VOID_STRING_INDEX);
|
|
|
|
for (i = 0; i < newline; i++)
|
|
add_newline();
|
|
|
|
source_ptr -= newline;
|
|
}
|
|
|
|
#if ENABLE_BIG_COMMENT
|
|
static void add_big_comment()
|
|
{
|
|
unsigned char car;
|
|
const char *start;
|
|
int len;
|
|
int index;
|
|
int type;
|
|
//bool space = FALSE;
|
|
|
|
start = source_ptr;
|
|
len = 0;
|
|
EVAL->comment = TRUE;
|
|
|
|
/*for(;;)
|
|
{
|
|
if (start == EVAL->source)
|
|
break;
|
|
|
|
start--;
|
|
car = *start;
|
|
if (car == '\n')
|
|
break;
|
|
|
|
if (car > ' ')
|
|
{
|
|
start++;
|
|
space = TRUE;
|
|
break;
|
|
}
|
|
len++;
|
|
}
|
|
|
|
if (!space)
|
|
{
|
|
start = source_ptr;
|
|
len = 1;
|
|
}*/
|
|
|
|
for(;;)
|
|
{
|
|
car = get_char();
|
|
if (car == 0)
|
|
break;
|
|
|
|
/*if (car == '\n')
|
|
{
|
|
TABLE_add_symbol(EVAL->string, start, len, NULL, &index);
|
|
type = RT_COMMENT;
|
|
add_pattern(type, index);
|
|
add_newline();
|
|
len = 1;
|
|
start = source_ptr + 1;
|
|
continue;
|
|
}*/
|
|
|
|
if (car == '*' && get_char_offset(1) == '/')
|
|
{
|
|
source_ptr += 2;
|
|
len += 2;
|
|
EVAL->comment = FALSE;
|
|
break;
|
|
}
|
|
len++;
|
|
source_ptr++;
|
|
}
|
|
|
|
index = TABLE_add_symbol(EVAL->string, start, len);
|
|
type = RT_COMMENT;
|
|
add_pattern(type, index);
|
|
}
|
|
#endif
|
|
|
|
static void add_comment()
|
|
{
|
|
unsigned char car;
|
|
const char *start;
|
|
int len;
|
|
int index;
|
|
int type;
|
|
bool space = FALSE;
|
|
|
|
start = source_ptr;
|
|
len = 1;
|
|
|
|
for(;;)
|
|
{
|
|
if (start == EVAL->source)
|
|
break;
|
|
|
|
start--;
|
|
car = *start;
|
|
if (car == '\n')
|
|
break;
|
|
|
|
if (car > ' ')
|
|
{
|
|
start++;
|
|
space = TRUE;
|
|
break;
|
|
}
|
|
len++;
|
|
}
|
|
|
|
if (!space)
|
|
{
|
|
start = source_ptr;
|
|
len = 1;
|
|
}
|
|
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
car = get_char();
|
|
if (car == 0 || car == '\n')
|
|
break;
|
|
len++;
|
|
}
|
|
|
|
index = TABLE_add_symbol(EVAL->string, start, len);
|
|
type = RT_COMMENT;
|
|
|
|
add_pattern(type, index);
|
|
}
|
|
|
|
|
|
static void add_string_for_analyze()
|
|
{
|
|
unsigned char car;
|
|
const char *start;
|
|
int len;
|
|
int index;
|
|
int type;
|
|
|
|
start = source_ptr;
|
|
len = 0;
|
|
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
car = get_char();
|
|
if (car == '\\')
|
|
{
|
|
source_ptr++;
|
|
car = get_char();
|
|
len++;
|
|
if (car == 0)
|
|
break;
|
|
}
|
|
else if (car == 0 || car == '\n' || car == '"')
|
|
break;
|
|
len++;
|
|
}
|
|
|
|
if (car == '"')
|
|
source_ptr++;
|
|
|
|
index = TABLE_add_symbol(EVAL->string, start + 1, len);
|
|
type = RT_STRING;
|
|
|
|
add_pattern(type, index);
|
|
//fprintf(stderr, "add_string_for_analyze: %s\n", TABLE_get_symbol_name(EVAL->string, index));
|
|
}
|
|
|
|
|
|
static void add_spaces()
|
|
{
|
|
unsigned char car;
|
|
int len;
|
|
|
|
len = 1;
|
|
|
|
for(;;)
|
|
{
|
|
source_ptr++;
|
|
car = get_char();
|
|
if (car > ' ' || car == '\n' || car == 0)
|
|
break;
|
|
len++;
|
|
}
|
|
|
|
add_pattern(RT_SPACE, len);
|
|
}
|
|
|
|
|
|
PUBLIC void EVAL_read(void)
|
|
{
|
|
static void *jump_char[11] =
|
|
{
|
|
&&__BREAK,
|
|
&&__SPACE,
|
|
&&__NEWLINE,
|
|
&&__COMMENT,
|
|
&&__STRING,
|
|
&&__IDENT,
|
|
&&__QUOTED_IDENT,
|
|
&&__NUMBER,
|
|
&&__ERROR,
|
|
&&__SHARP,
|
|
&&__OTHER
|
|
};
|
|
|
|
unsigned char car;
|
|
|
|
READ_init();
|
|
|
|
source_ptr = EVAL->source;
|
|
_begin_line = TRUE;
|
|
|
|
#if ENABLE_BIG_COMMENT
|
|
if (EVAL->comment)
|
|
goto __BIG_COMMENT;
|
|
#endif
|
|
|
|
for(;;)
|
|
{
|
|
car = get_char();
|
|
goto *jump_char[(int)first_car[car]];
|
|
|
|
__ERROR:
|
|
|
|
THROW(E_SYNTAX);
|
|
|
|
__SPACE:
|
|
|
|
if (EVAL->analyze)
|
|
add_spaces();
|
|
else
|
|
source_ptr++;
|
|
|
|
continue;
|
|
|
|
__NEWLINE:
|
|
|
|
source_ptr++;
|
|
add_newline();
|
|
_begin_line = TRUE;
|
|
continue;
|
|
|
|
__COMMENT:
|
|
|
|
if (EVAL->analyze)
|
|
{
|
|
add_comment();
|
|
}
|
|
else
|
|
{
|
|
do
|
|
{
|
|
source_ptr++;
|
|
car = get_char();
|
|
}
|
|
while (car != '\n' && car != 0);
|
|
}
|
|
|
|
_begin_line = FALSE;
|
|
continue;
|
|
|
|
__STRING:
|
|
|
|
if (EVAL->analyze)
|
|
add_string_for_analyze();
|
|
else
|
|
add_string();
|
|
_begin_line = FALSE;
|
|
continue;
|
|
|
|
__IDENT:
|
|
|
|
add_identifier();
|
|
_begin_line = FALSE;
|
|
continue;
|
|
|
|
__QUOTED_IDENT:
|
|
|
|
add_quoted_identifier();
|
|
_begin_line = FALSE;
|
|
continue;
|
|
|
|
__NUMBER:
|
|
|
|
add_number();
|
|
_begin_line = FALSE;
|
|
continue;
|
|
|
|
__SHARP:
|
|
|
|
if (_begin_line)
|
|
{
|
|
_begin_line = FALSE;
|
|
if (get_char_offset(1) == '!' && EVAL->analyze)
|
|
add_comment();
|
|
else
|
|
add_identifier();
|
|
continue;
|
|
}
|
|
|
|
__OTHER:
|
|
|
|
#if ENABLE_BIG_COMMENT
|
|
if (car == '/' && get_char_offset(1) == '*')
|
|
goto __BIG_COMMENT;
|
|
#endif
|
|
|
|
if (add_number())
|
|
add_operator();
|
|
|
|
_begin_line = FALSE;
|
|
continue;
|
|
|
|
#if ENABLE_BIG_COMMENT
|
|
__BIG_COMMENT:
|
|
|
|
if (EVAL->analyze)
|
|
add_big_comment();
|
|
else
|
|
{
|
|
for(;;)
|
|
{
|
|
car = get_char();
|
|
if (car == 0)
|
|
break;
|
|
if (car == '*' && get_char_offset(1) == '/')
|
|
{
|
|
source_ptr += 2;
|
|
break;
|
|
}
|
|
if (car == '\n')
|
|
add_newline();
|
|
source_ptr++;
|
|
}
|
|
}
|
|
|
|
_begin_line = FALSE;
|
|
continue;
|
|
#endif
|
|
}
|
|
|
|
__BREAK:
|
|
|
|
add_end();
|
|
add_end();
|
|
add_end();
|
|
add_end();
|
|
}
|
|
|