qse/ase/awk/parse.c
2007-12-08 00:24:28 +00:00

5523 lines
118 KiB
C

/*
* $Id: parse.c,v 1.34 2007/11/10 15:21:40 bacon Exp $
*
* {License}
*/
#include <ase/awk/awk_i.h>
enum
{
TOKEN_EOF,
TOKEN_NEWLINE,
/* TOKEN_XXX_ASSIGNs should in sync
* with assop in assign_to_opcode */
TOKEN_ASSIGN,
TOKEN_PLUS_ASSIGN,
TOKEN_MINUS_ASSIGN,
TOKEN_MUL_ASSIGN,
TOKEN_DIV_ASSIGN,
TOKEN_IDIV_ASSIGN,
TOKEN_MOD_ASSIGN,
TOKEN_EXP_ASSIGN,
TOKEN_EQ,
TOKEN_NE,
TOKEN_LE,
TOKEN_LT,
TOKEN_GE,
TOKEN_GT,
TOKEN_NM, /* not match */
TOKEN_NOT,
TOKEN_PLUS,
TOKEN_PLUSPLUS,
TOKEN_MINUS,
TOKEN_MINUSMINUS,
TOKEN_MUL,
TOKEN_DIV,
TOKEN_IDIV,
TOKEN_MOD,
TOKEN_LOR,
TOKEN_LAND,
TOKEN_BOR,
TOKEN_BXOR,
TOKEN_BAND,
TOKEN_BORAND,
TOKEN_TILDE, /* used for unary bitwise-not and regex match */
TOKEN_RSHIFT,
TOKEN_LSHIFT,
TOKEN_IN,
TOKEN_EXP,
TOKEN_LPAREN,
TOKEN_RPAREN,
TOKEN_LBRACE,
TOKEN_RBRACE,
TOKEN_LBRACK,
TOKEN_RBRACK,
TOKEN_DOLLAR,
TOKEN_COMMA,
TOKEN_PERIOD,
TOKEN_SEMICOLON,
TOKEN_COLON,
TOKEN_QUEST,
TOKEN_BEGIN,
TOKEN_END,
TOKEN_FUNCTION,
TOKEN_LOCAL,
TOKEN_GLOBAL,
TOKEN_IF,
TOKEN_ELSE,
TOKEN_WHILE,
TOKEN_FOR,
TOKEN_DO,
TOKEN_BREAK,
TOKEN_CONTINUE,
TOKEN_RETURN,
TOKEN_EXIT,
TOKEN_NEXT,
TOKEN_NEXTFILE,
TOKEN_NEXTOFILE,
TOKEN_DELETE,
TOKEN_RESET,
TOKEN_PRINT,
TOKEN_PRINTF,
TOKEN_GETLINE,
TOKEN_IDENT,
TOKEN_INT,
TOKEN_REAL,
TOKEN_STR,
TOKEN_REX,
__TOKEN_COUNT__
};
enum
{
PARSE_GLOBAL,
PARSE_FUNCTION,
PARSE_BEGIN,
PARSE_END,
PARSE_BEGIN_BLOCK,
PARSE_END_BLOCK,
PARSE_PATTERN,
PARSE_ACTION_BLOCK
};
enum
{
PARSE_LOOP_NONE,
PARSE_LOOP_WHILE,
PARSE_LOOP_FOR,
PARSE_LOOP_DOWHILE
};
typedef struct binmap_t binmap_t;
struct binmap_t
{
int token;
int binop;
};
static int parse (ase_awk_t* awk);
static ase_awk_t* parse_progunit (ase_awk_t* awk);
static ase_awk_t* collect_globals (ase_awk_t* awk);
static void adjust_static_globals (ase_awk_t* awk);
static ase_size_t get_global (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len);
static ase_size_t find_global (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len);
static int add_global (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len,
ase_size_t line, int force);
static ase_awk_t* collect_locals (ase_awk_t* awk, ase_size_t nlocals);
static ase_awk_nde_t* parse_function (ase_awk_t* awk);
static ase_awk_nde_t* parse_begin (ase_awk_t* awk);
static ase_awk_nde_t* parse_end (ase_awk_t* awk);
static ase_awk_chain_t* parse_pattern_block (
ase_awk_t* awk, ase_awk_nde_t* ptn, ase_bool_t blockless);
static ase_awk_nde_t* parse_block (ase_awk_t* awk, ase_size_t line, ase_bool_t is_top);
static ase_awk_nde_t* parse_block_dc (ase_awk_t* awk, ase_size_t line, ase_bool_t is_top);
static ase_awk_nde_t* parse_statement (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_statement_nb (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_expression (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_expression0 (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_basic_expr (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_binary_expr (
ase_awk_t* awk, ase_size_t line, const binmap_t* binmap,
ase_awk_nde_t*(*next_level_func)(ase_awk_t*,ase_size_t));
static ase_awk_nde_t* parse_logical_or (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_logical_and (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_in (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_regex_match (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_bitwise_or (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_bitwise_or_with_extio (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_bitwise_xor (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_bitwise_and (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_equality (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_relational (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_shift (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_concat (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_additive (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_multiplicative (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_unary (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_exponent (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_unary_exp (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_increment (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_primary (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_primary_ident (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_hashidx (
ase_awk_t* awk, ase_char_t* name, ase_size_t name_len,
ase_size_t line);
static ase_awk_nde_t* parse_fncall (
ase_awk_t* awk, ase_char_t* name, ase_size_t name_len,
ase_awk_bfn_t* bfn, ase_size_t line);
static ase_awk_nde_t* parse_if (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_while (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_for (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_dowhile (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_break (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_continue (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_return (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_exit (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_next (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_nextfile (ase_awk_t* awk, ase_size_t line, int out);
static ase_awk_nde_t* parse_delete (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_reset (ase_awk_t* awk, ase_size_t line);
static ase_awk_nde_t* parse_print (ase_awk_t* awk, ase_size_t line, int type);
static int get_token (ase_awk_t* awk);
static int get_number (ase_awk_t* awk);
static int get_charstr (ase_awk_t* awk);
static int get_rexstr (ase_awk_t* awk);
static int get_string (
ase_awk_t* awk, ase_char_t end_char,
ase_char_t esc_char, ase_bool_t keep_esc_char);
static int get_char (ase_awk_t* awk);
static int unget_char (ase_awk_t* awk, ase_cint_t c);
static int skip_spaces (ase_awk_t* awk);
static int skip_comment (ase_awk_t* awk);
static int classify_ident (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len);
static int assign_to_opcode (ase_awk_t* awk);
static int is_plain_var (ase_awk_nde_t* nde);
static int is_var (ase_awk_nde_t* nde);
static int deparse (ase_awk_t* awk);
static int deparse_func (ase_awk_pair_t* pair, void* arg);
static int put_char (ase_awk_t* awk, ase_char_t c);
static int flush_out (ase_awk_t* awk);
typedef struct kwent_t kwent_t;
struct kwent_t
{
const ase_char_t* name;
ase_size_t name_len;
int type;
int valid; /* the entry is valid when this option is set */
};
static kwent_t kwtab[] =
{
/* operators */
{ ASE_T("in"), 2, TOKEN_IN, 0 },
/* top-level block starters */
{ ASE_T("BEGIN"), 5, TOKEN_BEGIN, ASE_AWK_PABLOCK },
{ ASE_T("END"), 3, TOKEN_END, ASE_AWK_PABLOCK },
{ ASE_T("function"), 8, TOKEN_FUNCTION, 0 },
/* keywords for variable declaration */
{ ASE_T("local"), 5, TOKEN_LOCAL, ASE_AWK_EXPLICIT },
{ ASE_T("global"), 6, TOKEN_GLOBAL, ASE_AWK_EXPLICIT },
/* keywords that start statements excluding expression statements */
{ ASE_T("if"), 2, TOKEN_IF, 0 },
{ ASE_T("else"), 4, TOKEN_ELSE, 0 },
{ ASE_T("while"), 5, TOKEN_WHILE, 0 },
{ ASE_T("for"), 3, TOKEN_FOR, 0 },
{ ASE_T("do"), 2, TOKEN_DO, 0 },
{ ASE_T("break"), 5, TOKEN_BREAK, 0 },
{ ASE_T("continue"), 8, TOKEN_CONTINUE, 0 },
{ ASE_T("return"), 6, TOKEN_RETURN, 0 },
{ ASE_T("exit"), 4, TOKEN_EXIT, 0 },
{ ASE_T("next"), 4, TOKEN_NEXT, ASE_AWK_PABLOCK },
{ ASE_T("nextfile"), 8, TOKEN_NEXTFILE, ASE_AWK_PABLOCK },
{ ASE_T("nextofile"), 9, TOKEN_NEXTOFILE, ASE_AWK_PABLOCK | ASE_AWK_NEXTOFILE },
{ ASE_T("delete"), 6, TOKEN_DELETE, 0 },
{ ASE_T("reset"), 5, TOKEN_RESET, ASE_AWK_RESET },
{ ASE_T("print"), 5, TOKEN_PRINT, ASE_AWK_EXTIO },
{ ASE_T("printf"), 6, TOKEN_PRINTF, ASE_AWK_EXTIO },
/* keywords that can start an expression */
{ ASE_T("getline"), 7, TOKEN_GETLINE, ASE_AWK_EXTIO },
{ ASE_NULL, 0, 0, 0 }
};
typedef struct global_t global_t;
struct global_t
{
const ase_char_t* name;
ase_size_t name_len;
int valid;
};
static global_t gtab[] =
{
{ ASE_T("ARGC"), 4, 0 },
{ ASE_T("ARGV"), 4, 0 },
/* output real-to-str conversion format for other cases than 'print' */
{ ASE_T("CONVFMT"), 7, 0 },
/* current input file name */
{ ASE_T("FILENAME"), 8, ASE_AWK_PABLOCK },
/* input record number in current file */
{ ASE_T("FNR"), 3, ASE_AWK_PABLOCK },
/* input field separator */
{ ASE_T("FS"), 2, 0 },
/* ignore case in string comparison */
{ ASE_T("IGNORECASE"), 10, 0 },
/* number of fields in current input record
* NF is also updated if you assign a value to $0. so it is not
* associated with ASE_AWK_PABLOCK */
{ ASE_T("NF"), 2, 0 },
/* input record number */
{ ASE_T("NR"), 2, ASE_AWK_PABLOCK },
/* current output file name */
{ ASE_T("OFILENAME"), 9, ASE_AWK_PABLOCK | ASE_AWK_NEXTOFILE },
/* output real-to-str conversion format for 'print' */
{ ASE_T("OFMT"), 4, ASE_AWK_EXTIO},
/* output field separator for 'print' */
{ ASE_T("OFS"), 3, ASE_AWK_EXTIO },
/* output record separator. used for 'print' and blockless output
* ASE_AWK_BLOCKLESS desn't have to be specified becuase
* it requires ASE_AWK_EXTIO to be ON. */
{ ASE_T("ORS"), 3, ASE_AWK_EXTIO },
{ ASE_T("RLENGTH"), 7, 0 },
{ ASE_T("RS"), 2, 0 },
{ ASE_T("RSTART"), 6, 0 },
{ ASE_T("SUBSEP"), 6, 0 }
};
#define GET_CHAR(awk) \
do { if (get_char (awk) == -1) return -1; } while(0)
#define GET_CHAR_TO(awk,c) \
do { \
if (get_char (awk) == -1) return -1; \
c = (awk)->src.lex.curc; \
} while(0)
#define SET_TOKEN_TYPE(awk,code) do { (awk)->token.type = (code); } while (0)
#define ADD_TOKEN_CHAR(awk,c) \
do { \
if (ase_str_ccat(&(awk)->token.name,(c)) == (ase_size_t)-1) \
{ \
ase_awk_seterror (awk, ASE_AWK_ENOMEM, (awk)->token.line, ASE_NULL, 0); \
return -1; \
} \
} while (0)
#define MATCH(awk,token_type) ((awk)->token.type == (token_type))
#define CLRERR(awk) ase_awk_seterrnum(awk,ASE_AWK_ENOERR)
#define ISNOERR(awk) ((awk)->errnum == ASE_AWK_ENOERR)
#define SETERR(awk,code) ase_awk_seterrnum(awk,code)
#define SETERRLIN(awk,code,line) ase_awk_seterror(awk,code,line,ASE_NULL,0)
#define SETERRTOK(awk,code) \
do { \
ase_cstr_t errarg; \
errarg.len = ASE_STR_LEN(&(awk)->token.name); \
errarg.ptr = ASE_STR_BUF(&(awk)->token.name); \
if (MATCH(awk,TOKEN_EOF)) \
ase_awk_seterror (awk, code, (awk)->token.prev.line, &errarg, 1); \
else \
ase_awk_seterror (awk, code, (awk)->token.line, &errarg, 1); \
} while (0)
#define SETERRARG(awk,code,line,arg,leng) \
do { \
ase_cstr_t errarg; \
errarg.len = (leng); \
errarg.ptr = (arg); \
ase_awk_seterror ((awk), (code), (line), &errarg, 1); \
} while (0)
ase_size_t ase_awk_getmaxdepth (ase_awk_t* awk, int type)
{
return (type == ASE_AWK_DEPTH_BLOCK_PARSE)? awk->parse.depth.max.block:
(type == ASE_AWK_DEPTH_BLOCK_RUN)? awk->run.depth.max.block:
(type == ASE_AWK_DEPTH_EXPR_PARSE)? awk->parse.depth.max.expr:
(type == ASE_AWK_DEPTH_EXPR_RUN)? awk->run.depth.max.expr:
(type == ASE_AWK_DEPTH_REX_BUILD)? awk->rex.depth.max.build:
(type == ASE_AWK_DEPTH_REX_MATCH)? awk->rex.depth.max.match: 0;
}
void ase_awk_setmaxdepth (ase_awk_t* awk, int types, ase_size_t depth)
{
if (types & ASE_AWK_DEPTH_BLOCK_PARSE)
{
awk->parse.depth.max.block = depth;
if (depth <= 0)
awk->parse.parse_block = parse_block;
else
awk->parse.parse_block = parse_block_dc;
}
if (types & ASE_AWK_DEPTH_EXPR_PARSE)
{
awk->parse.depth.max.expr = depth;
}
if (types & ASE_AWK_DEPTH_BLOCK_RUN)
{
awk->run.depth.max.block = depth;
}
if (types & ASE_AWK_DEPTH_EXPR_RUN)
{
awk->run.depth.max.expr = depth;
}
if (types & ASE_AWK_DEPTH_REX_BUILD)
{
awk->rex.depth.max.build = depth;
}
if (types & ASE_AWK_DEPTH_REX_MATCH)
{
awk->rex.depth.max.match = depth;
}
}
const ase_char_t* ase_awk_getglobalname (
ase_awk_t* awk, ase_size_t idx, ase_size_t* len)
{
/*
*len = gtab[idx].name_len;
return gtab[idx].name;
*/
ASE_ASSERT (idx < ase_awk_tab_getsize(&awk->parse.globals));
*len = awk->parse.globals.buf[idx].name.len;
return awk->parse.globals.buf[idx].name.ptr;
}
const ase_char_t* ase_awk_getkw (ase_awk_t* awk, const ase_char_t* kw)
{
ase_awk_pair_t* p;
ASE_ASSERT (kw != ASE_NULL);
p = ase_awk_map_get (awk->wtab, kw, ase_strlen(kw));
if (p != ASE_NULL) return ((ase_cstr_t*)p->val)->ptr;
return kw;
}
int ase_awk_parse (ase_awk_t* awk, ase_awk_srcios_t* srcios)
{
int n;
ASE_ASSERTX (
srcios != ASE_NULL && srcios->in != ASE_NULL,
"the source code input stream must be provided at least");
ASE_ASSERT (awk->parse.depth.cur.loop == 0);
ASE_ASSERT (awk->parse.depth.cur.expr == 0);
ase_awk_clear (awk);
ase_memcpy (&awk->src.ios, srcios, ASE_SIZEOF(awk->src.ios));
n = parse (awk);
ASE_ASSERT (awk->parse.depth.cur.loop == 0);
ASE_ASSERT (awk->parse.depth.cur.expr == 0);
return n;
}
static int parse (ase_awk_t* awk)
{
int n = 0;
ase_ssize_t op;
ASE_ASSERT (awk->src.ios.in != ASE_NULL);
CLRERR (awk);
op = awk->src.ios.in (
ASE_AWK_IO_OPEN, awk->src.ios.custom_data, ASE_NULL, 0);
if (op <= -1)
{
/* cannot open the source file.
* it doesn't even have to call CLOSE */
if (ISNOERR(awk)) SETERR (awk, ASE_AWK_ESINOP);
return -1;
}
adjust_static_globals (awk);
/* the user io handler for the source code input returns 0 when
* it doesn't have any files to open. this is the same condition
* as the source code file is empty. so it will perform the parsing
* when op is positive, which means there are something to parse */
if (op > 0)
{
/* get the first character */
if (get_char(awk) == -1)
{
n = -1;
goto exit_parse;
}
/* get the first token */
if (get_token(awk) == -1)
{
n = -1;
goto exit_parse;
}
while (1)
{
if (MATCH(awk,TOKEN_EOF)) break;
if (MATCH(awk,TOKEN_NEWLINE)) continue;
if (parse_progunit (awk) == ASE_NULL)
{
n = -1;
goto exit_parse;
}
}
}
ASE_ASSERT (awk->tree.nglobals == ase_awk_tab_getsize(&awk->parse.globals));
if (awk->src.ios.out != ASE_NULL)
{
if (deparse (awk) == -1)
{
n = -1;
goto exit_parse;
}
}
exit_parse:
if (n == 0) CLRERR (awk);
if (awk->src.ios.in (
ASE_AWK_IO_CLOSE, awk->src.ios.custom_data, ASE_NULL, 0) != 0)
{
if (n == 0)
{
/* this is to keep the earlier error above
* that might be more critical than this */
if (ISNOERR(awk)) SETERR (awk, ASE_AWK_ESINCL);
n = -1;
}
}
if (n == -1) ase_awk_clear (awk);
else awk->tree.ok = 1;
return n;
}
static ase_awk_t* parse_progunit (ase_awk_t* awk)
{
/*
pattern { action }
function name (parameter-list) { statement }
*/
ASE_ASSERT (awk->parse.depth.cur.loop == 0);
if ((awk->option & ASE_AWK_EXPLICIT) && MATCH(awk,TOKEN_GLOBAL))
{
ase_size_t nglobals;
awk->parse.id.block = PARSE_GLOBAL;
if (get_token(awk) == -1) return ASE_NULL;
ASE_ASSERT (awk->tree.nglobals == ase_awk_tab_getsize(&awk->parse.globals));
nglobals = awk->tree.nglobals;
if (collect_globals (awk) == ASE_NULL)
{
ase_awk_tab_remove (
&awk->parse.globals, nglobals,
ase_awk_tab_getsize(&awk->parse.globals) - nglobals);
awk->tree.nglobals = nglobals;
return ASE_NULL;
}
}
else if (MATCH(awk,TOKEN_FUNCTION))
{
awk->parse.id.block = PARSE_FUNCTION;
if (parse_function (awk) == ASE_NULL) return ASE_NULL;
}
else if (MATCH(awk,TOKEN_BEGIN))
{
if ((awk->option & ASE_AWK_PABLOCK) == 0)
{
SETERRTOK (awk, ASE_AWK_EFUNC);
return ASE_NULL;
}
/*
if (awk->tree.begin != ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_EDUPBEG, awk->token.prev.line);
return ASE_NULL;
}
*/
awk->parse.id.block = PARSE_BEGIN;
if (get_token(awk) == -1) return ASE_NULL;
if ((awk->option & ASE_AWK_BLOCKLESS) &&
(MATCH(awk,TOKEN_NEWLINE) || MATCH(awk,TOKEN_EOF)))
{
/* when the blockless pattern is supported
* BEGIN and { should be located on the same line */
SETERRLIN (awk, ASE_AWK_EBLKBEG, awk->token.prev.line);
return ASE_NULL;
}
if (!MATCH(awk,TOKEN_LBRACE))
{
SETERRTOK (awk, ASE_AWK_ELBRACE);
return ASE_NULL;
}
awk->parse.id.block = PARSE_BEGIN_BLOCK;
if (parse_begin (awk) == ASE_NULL) return ASE_NULL;
}
else if (MATCH(awk,TOKEN_END))
{
if ((awk->option & ASE_AWK_PABLOCK) == 0)
{
SETERRTOK (awk, ASE_AWK_EFUNC);
return ASE_NULL;
}
/*
if (awk->tree.end != ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_EDUPEND, awk->token.prev.line);
return ASE_NULL;
}
*/
awk->parse.id.block = PARSE_END;
if (get_token(awk) == -1) return ASE_NULL;
if ((awk->option & ASE_AWK_BLOCKLESS) &&
(MATCH(awk,TOKEN_NEWLINE) || MATCH(awk,TOKEN_EOF)))
{
/* when the blockless pattern is supported
* END and { should be located on the same line */
SETERRLIN (awk, ASE_AWK_EBLKEND, awk->token.prev.line);
return ASE_NULL;
}
if (!MATCH(awk,TOKEN_LBRACE))
{
SETERRTOK (awk, ASE_AWK_ELBRACE);
return ASE_NULL;
}
awk->parse.id.block = PARSE_END_BLOCK;
if (parse_end (awk) == ASE_NULL) return ASE_NULL;
}
else if (MATCH(awk,TOKEN_LBRACE))
{
/* patternless block */
if ((awk->option & ASE_AWK_PABLOCK) == 0)
{
SETERRTOK (awk, ASE_AWK_EFUNC);
return ASE_NULL;
}
awk->parse.id.block = PARSE_ACTION_BLOCK;
if (parse_pattern_block (
awk, ASE_NULL, ase_false) == ASE_NULL) return ASE_NULL;
}
else
{
/*
expressions
/regular expression/
pattern && pattern
pattern || pattern
!pattern
(pattern)
pattern, pattern
*/
ase_awk_nde_t* ptn;
if ((awk->option & ASE_AWK_PABLOCK) == 0)
{
SETERRTOK (awk, ASE_AWK_EFUNC);
return ASE_NULL;
}
awk->parse.id.block = PARSE_PATTERN;
ptn = parse_expression (awk, awk->token.line);
if (ptn == ASE_NULL) return ASE_NULL;
ASE_ASSERT (ptn->next == ASE_NULL);
if (MATCH(awk,TOKEN_COMMA))
{
if (get_token (awk) == -1)
{
ase_awk_clrpt (awk, ptn);
return ASE_NULL;
}
ptn->next = parse_expression (awk, awk->token.line);
if (ptn->next == ASE_NULL)
{
ase_awk_clrpt (awk, ptn);
return ASE_NULL;
}
}
if ((awk->option & ASE_AWK_BLOCKLESS) &&
(MATCH(awk,TOKEN_NEWLINE) || MATCH(awk,TOKEN_EOF)))
{
/* blockless pattern */
ase_bool_t newline = MATCH(awk,TOKEN_NEWLINE);
ase_size_t tline = awk->token.prev.line;
awk->parse.id.block = PARSE_ACTION_BLOCK;
if (parse_pattern_block (
awk, ptn, ase_true) == ASE_NULL)
{
ase_awk_clrpt (awk, ptn);
return ASE_NULL;
}
if (newline)
{
if (get_token (awk) == -1)
{
/* ptn has been added to the chain.
* it doesn't have to be cleared here
* as ase_awk_clear does it */
/*ase_awk_clrpt (awk, ptn);*/
return ASE_NULL;
}
}
if ((awk->option & ASE_AWK_EXTIO) != ASE_AWK_EXTIO)
{
/* blockless pattern requires ASE_AWK_EXTIO
* to be ON because the implicit block is
* "print $0" */
SETERRLIN (awk, ASE_AWK_ENOSUP, tline);
return ASE_NULL;
}
}
else
{
/* parse the action block */
if (!MATCH(awk,TOKEN_LBRACE))
{
ase_awk_clrpt (awk, ptn);
SETERRTOK (awk, ASE_AWK_ELBRACE);
return ASE_NULL;
}
awk->parse.id.block = PARSE_ACTION_BLOCK;
if (parse_pattern_block (
awk, ptn, ase_false) == ASE_NULL)
{
ase_awk_clrpt (awk, ptn);
return ASE_NULL;
}
}
}
return awk;
}
static ase_awk_nde_t* parse_function (ase_awk_t* awk)
{
ase_char_t* name;
ase_char_t* name_dup;
ase_size_t name_len;
ase_awk_nde_t* body;
ase_awk_afn_t* afn;
ase_size_t nargs;
ase_awk_pair_t* pair;
int n;
/* eat up the keyword 'function' and get the next token */
ASE_ASSERT (MATCH(awk,TOKEN_FUNCTION));
if (get_token(awk) == -1) return ASE_NULL;
/* match a function name */
if (!MATCH(awk,TOKEN_IDENT))
{
/* cannot find a valid identifier for a function name */
SETERRTOK (awk, ASE_AWK_EFNNAME);
return ASE_NULL;
}
name = ASE_STR_BUF(&awk->token.name);
name_len = ASE_STR_LEN(&awk->token.name);
/* check if it is a builtin function */
if (ase_awk_getbfn (awk, name, name_len) != ASE_NULL)
{
SETERRARG (awk, ASE_AWK_EBFNRED, awk->token.line, name, name_len);
return ASE_NULL;
}
if (ase_awk_map_get(awk->tree.afns, name, name_len) != ASE_NULL)
{
/* the function is defined previously */
SETERRARG (
awk, ASE_AWK_EAFNRED, awk->token.line,
name, name_len);
return ASE_NULL;
}
if (awk->option & ASE_AWK_UNIQUEFN)
{
/* check if it coincides to be a global variable name */
ase_size_t g = find_global (awk, name, name_len);
if (g != (ase_size_t)-1)
{
SETERRARG (
awk, ASE_AWK_EGBLRED, awk->token.line,
name, name_len);
return ASE_NULL;
}
}
/* clone the function name before it is overwritten */
name_dup = ase_awk_strxdup (awk, name, name_len);
if (name_dup == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, awk->token.line);
return ASE_NULL;
}
/* get the next token */
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, name_dup);
return ASE_NULL;
}
/* match a left parenthesis */
if (!MATCH(awk,TOKEN_LPAREN))
{
/* a function name is not followed by a left parenthesis */
ASE_AWK_FREE (awk, name_dup);
SETERRTOK (awk, ASE_AWK_ELPAREN);
return ASE_NULL;
}
/* get the next token */
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, name_dup);
return ASE_NULL;
}
/* make sure that parameter table is empty */
ASE_ASSERT (ase_awk_tab_getsize(&awk->parse.params) == 0);
/* read parameter list */
if (MATCH(awk,TOKEN_RPAREN))
{
/* no function parameter found. get the next token */
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, name_dup);
return ASE_NULL;
}
}
else
{
while (1)
{
ase_char_t* param;
ase_size_t param_len;
if (!MATCH(awk,TOKEN_IDENT))
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
SETERRTOK (awk, ASE_AWK_EBADPAR);
return ASE_NULL;
}
param = ASE_STR_BUF(&awk->token.name);
param_len = ASE_STR_LEN(&awk->token.name);
if (awk->option & ASE_AWK_UNIQUEFN)
{
/* check if a parameter conflicts with a function */
if (ase_strxncmp (name_dup, name_len, param, param_len) == 0 ||
ase_awk_map_get (awk->tree.afns, param, param_len) != ASE_NULL)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
SETERRARG (
awk, ASE_AWK_EAFNRED, awk->token.line,
param, param_len);
return ASE_NULL;
}
/* NOTE: the following is not a conflict
* global x;
* function f (x) { print x; }
* x in print x is a parameter
*/
}
/* check if a parameter conflicts with other parameters */
if (ase_awk_tab_find (
&awk->parse.params,
0, param, param_len) != (ase_size_t)-1)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
SETERRARG (
awk, ASE_AWK_EDUPPAR, awk->token.line,
param, param_len);
return ASE_NULL;
}
/* push the parameter to the parameter list */
if (ase_awk_tab_getsize (
&awk->parse.params) >= ASE_AWK_MAX_PARAMS)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
SETERRTOK (awk, ASE_AWK_EPARTM);
return ASE_NULL;
}
if (ase_awk_tab_add (
&awk->parse.params,
param, param_len) == (ase_size_t)-1)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
SETERRLIN (awk, ASE_AWK_ENOMEM, awk->token.line);
return ASE_NULL;
}
if (get_token (awk) == -1)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
return ASE_NULL;
}
if (MATCH(awk,TOKEN_RPAREN)) break;
if (!MATCH(awk,TOKEN_COMMA))
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
SETERRTOK (awk, ASE_AWK_ECOMMA);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
return ASE_NULL;
}
}
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
return ASE_NULL;
}
}
/* check if the function body starts with a left brace */
if (!MATCH(awk,TOKEN_LBRACE))
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
SETERRTOK (awk, ASE_AWK_ELBRACE);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
return ASE_NULL;
}
/* remember the current function name so that the body parser
* can know the name of the current function being parsed */
awk->tree.cur_afn.ptr = name_dup;
awk->tree.cur_afn.len = name_len;
/* actual function body */
body = awk->parse.parse_block (awk, awk->token.prev.line, ase_true);
/* clear the current function name remembered */
awk->tree.cur_afn.ptr = ASE_NULL;
awk->tree.cur_afn.len = 0;
if (body == ASE_NULL)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_tab_clear (&awk->parse.params);
return ASE_NULL;
}
/* TODO: study furthur if the parameter names should be saved
* for some reasons - might be needed for deparsing output */
nargs = ase_awk_tab_getsize (&awk->parse.params);
/* parameter names are not required anymore. clear them */
ase_awk_tab_clear (&awk->parse.params);
afn = (ase_awk_afn_t*) ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_afn_t));
if (afn == ASE_NULL)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_clrpt (awk, body);
SETERRLIN (awk, ASE_AWK_ENOMEM, awk->token.line);
return ASE_NULL;
}
afn->name = ASE_NULL; /* function name set below */
afn->name_len = 0;
afn->nargs = nargs;
afn->body = body;
n = ase_awk_map_putx (awk->tree.afns, name_dup, name_len, afn, &pair);
if (n < 0)
{
ASE_AWK_FREE (awk, name_dup);
ase_awk_clrpt (awk, body);
ASE_AWK_FREE (awk, afn);
SETERRLIN (awk, ASE_AWK_ENOMEM, awk->token.line);
return ASE_NULL;
}
/* duplicate functions should have been detected previously */
ASE_ASSERT (n != 0);
afn->name = ASE_AWK_PAIR_KEYPTR(pair); /* do some trick to save a string. */
afn->name_len = ASE_AWK_PAIR_KEYLEN(pair);
ASE_AWK_FREE (awk, name_dup);
return body;
}
static ase_awk_nde_t* parse_begin (ase_awk_t* awk)
{
ase_awk_nde_t* nde;
ASE_ASSERT (MATCH(awk,TOKEN_LBRACE));
if (get_token(awk) == -1) return ASE_NULL;
nde = awk->parse.parse_block (awk, awk->token.prev.line, ase_true);
if (nde == ASE_NULL) return ASE_NULL;
if (awk->tree.begin == ASE_NULL)
{
awk->tree.begin = nde;
awk->tree.begin_tail = nde;
}
else
{
awk->tree.begin_tail->next = nde;
awk->tree.begin_tail = nde;
}
return nde;
}
static ase_awk_nde_t* parse_end (ase_awk_t* awk)
{
ase_awk_nde_t* nde;
ASE_ASSERT (MATCH(awk,TOKEN_LBRACE));
if (get_token(awk) == -1) return ASE_NULL;
nde = awk->parse.parse_block (awk, awk->token.prev.line, ase_true);
if (nde == ASE_NULL) return ASE_NULL;
if (awk->tree.end == ASE_NULL)
{
awk->tree.end = nde;
awk->tree.end_tail = nde;
}
else
{
awk->tree.end_tail->next = nde;
awk->tree.end_tail = nde;
}
return nde;
}
static ase_awk_chain_t* parse_pattern_block (
ase_awk_t* awk, ase_awk_nde_t* ptn, ase_bool_t blockless)
{
ase_awk_nde_t* nde;
ase_awk_chain_t* chain;
ase_size_t line = awk->token.line;
if (blockless) nde = ASE_NULL;
else
{
ASE_ASSERT (MATCH(awk,TOKEN_LBRACE));
if (get_token(awk) == -1) return ASE_NULL;
nde = awk->parse.parse_block (awk, line, ase_true);
if (nde == ASE_NULL) return ASE_NULL;
}
chain = (ase_awk_chain_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_chain_t));
if (chain == ASE_NULL)
{
ase_awk_clrpt (awk, nde);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
chain->pattern = ptn;
chain->action = nde;
chain->next = ASE_NULL;
if (awk->tree.chain == ASE_NULL)
{
awk->tree.chain = chain;
awk->tree.chain_tail = chain;
awk->tree.chain_size++;
}
else
{
awk->tree.chain_tail->next = chain;
awk->tree.chain_tail = chain;
awk->tree.chain_size++;
}
return chain;
}
static ase_awk_nde_t* parse_block (
ase_awk_t* awk, ase_size_t line, ase_bool_t is_top)
{
ase_awk_nde_t* head, * curr, * nde;
ase_awk_nde_blk_t* block;
ase_size_t nlocals, nlocals_max, tmp;
nlocals = ase_awk_tab_getsize(&awk->parse.locals);
nlocals_max = awk->parse.nlocals_max;
/* local variable declarations */
if (awk->option & ASE_AWK_EXPLICIT)
{
while (1)
{
if (!MATCH(awk,TOKEN_LOCAL)) break;
if (get_token(awk) == -1)
{
ase_awk_tab_remove (
&awk->parse.locals, nlocals,
ase_awk_tab_getsize(&awk->parse.locals) - nlocals);
return ASE_NULL;
}
if (collect_locals (awk, nlocals) == ASE_NULL)
{
ase_awk_tab_remove (
&awk->parse.locals, nlocals,
ase_awk_tab_getsize(&awk->parse.locals) - nlocals);
return ASE_NULL;
}
}
}
/* block body */
head = ASE_NULL; curr = ASE_NULL;
while (1)
{
if (MATCH(awk,TOKEN_EOF))
{
ase_awk_tab_remove (
&awk->parse.locals, nlocals,
ase_awk_tab_getsize(&awk->parse.locals) - nlocals);
if (head != ASE_NULL) ase_awk_clrpt (awk, head);
SETERRLIN (awk, ASE_AWK_EENDSRC, awk->token.prev.line);
return ASE_NULL;
}
if (MATCH(awk,TOKEN_RBRACE))
{
if (get_token(awk) == -1)
{
ase_awk_tab_remove (
&awk->parse.locals, nlocals,
ase_awk_tab_getsize(&awk->parse.locals) - nlocals);
if (head != ASE_NULL) ase_awk_clrpt (awk, head);
return ASE_NULL;
}
break;
}
nde = parse_statement (awk, awk->token.line);
if (nde == ASE_NULL)
{
ase_awk_tab_remove (
&awk->parse.locals, nlocals,
ase_awk_tab_getsize(&awk->parse.locals)-nlocals);
if (head != ASE_NULL) ase_awk_clrpt (awk, head);
return ASE_NULL;
}
/* remove unnecessary statements */
if (nde->type == ASE_AWK_NDE_NULL)
{
ase_awk_clrpt (awk, nde);
continue;
}
if (nde->type == ASE_AWK_NDE_BLK &&
((ase_awk_nde_blk_t*)nde)->body == ASE_NULL)
{
ase_awk_clrpt (awk, nde);
continue;
}
if (curr == ASE_NULL) head = nde;
else curr->next = nde;
curr = nde;
}
block = (ase_awk_nde_blk_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_blk_t));
if (block == ASE_NULL)
{
ase_awk_tab_remove (
&awk->parse.locals, nlocals,
ase_awk_tab_getsize(&awk->parse.locals)-nlocals);
ase_awk_clrpt (awk, head);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
tmp = ase_awk_tab_getsize(&awk->parse.locals);
if (tmp > awk->parse.nlocals_max) awk->parse.nlocals_max = tmp;
/* remove all locals to move it up to the top level */
ase_awk_tab_remove (&awk->parse.locals, nlocals, tmp - nlocals);
/* adjust the number of locals for a block without any statements */
/* if (head == ASE_NULL) tmp = 0; */
block->type = ASE_AWK_NDE_BLK;
block->line = line;
block->next = ASE_NULL;
block->body = head;
/* TODO: not only local variables but also nested blocks,
unless it is part of other constructs such as if, can be promoted
and merged to top-level block */
/* migrate all block-local variables to a top-level block */
if (is_top)
{
block->nlocals = awk->parse.nlocals_max - nlocals;
awk->parse.nlocals_max = nlocals_max;
}
else
{
/*block->nlocals = tmp - nlocals;*/
block->nlocals = 0;
}
return (ase_awk_nde_t*)block;
}
static ase_awk_nde_t* parse_block_dc (
ase_awk_t* awk, ase_size_t line, ase_bool_t is_top)
{
ase_awk_nde_t* nde;
ASE_ASSERT (awk->parse.depth.max.block > 0);
if (awk->parse.depth.cur.block >= awk->parse.depth.max.block)
{
SETERRLIN (awk, ASE_AWK_EBLKNST, awk->token.prev.line);
return ASE_NULL;
}
awk->parse.depth.cur.block++;
nde = parse_block (awk, line, is_top);
awk->parse.depth.cur.block--;
return nde;
}
int ase_awk_initglobals (ase_awk_t* awk)
{
int id;
/* ase_awk_initglobals is not generic-purpose. call this from
* ase_awk_open only. */
ASE_ASSERT (awk->tree.nbglobals == 0 && awk->tree.nglobals == 0);
awk->tree.nbglobals = 0;
awk->tree.nglobals = 0;
for (id = ASE_AWK_MIN_GLOBAL_ID; id <= ASE_AWK_MAX_GLOBAL_ID; id++)
{
ase_size_t g;
g = ase_awk_tab_add (&awk->parse.globals,
gtab[id].name, gtab[id].name_len);
if (g == (ase_size_t)-1) return -1;
ASE_ASSERT ((int)g == id);
awk->tree.nbglobals++;
awk->tree.nglobals++;
}
ASE_ASSERT (awk->tree.nbglobals ==
ASE_AWK_MAX_GLOBAL_ID-ASE_AWK_MIN_GLOBAL_ID+1);
return 0;
}
static void adjust_static_globals (ase_awk_t* awk)
{
int id;
ASE_ASSERT (awk->tree.nbglobals >=
ASE_AWK_MAX_GLOBAL_ID - ASE_AWK_MAX_GLOBAL_ID + 1);
for (id = ASE_AWK_MIN_GLOBAL_ID; id <= ASE_AWK_MAX_GLOBAL_ID; id++)
{
if (gtab[id].valid != 0 &&
(awk->option & gtab[id].valid) != gtab[id].valid)
{
awk->parse.globals.buf[id].name.len = 0;
}
else
{
awk->parse.globals.buf[id].name.len = gtab[id].name_len;
}
}
}
static void trans_global (
ase_size_t index, ase_cstr_t* word, void* arg)
{
ase_awk_tab_t* tab = (ase_awk_tab_t*)arg;
ase_awk_t* awk = tab->awk;
/*
if (index >= ASE_AWK_MIN_GLOBAL_ID &&
index <= ASE_AWK_MAX_GLOBAL_ID)
*/
if (index < awk->tree.nbglobals)
{
ase_awk_pair_t* pair;
pair = ase_awk_map_get (awk->wtab, word->ptr, word->len);
if (pair != ASE_NULL)
{
word->ptr = ((ase_cstr_t*)(pair->val))->ptr;
word->len = ((ase_cstr_t*)(pair->val))->len;
}
}
}
static ase_size_t get_global (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len)
{
return ase_awk_tab_rrfindx (
&awk->parse.globals, 0, name, len,
trans_global, &awk->parse.globals);
}
static ase_size_t find_global (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len)
{
return ase_awk_tab_findx (
&awk->parse.globals, 0, name, len,
trans_global, &awk->parse.globals);
}
static int add_global (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len,
ase_size_t line, int disabled)
{
ase_size_t nglobals;
if (awk->option & ASE_AWK_UNIQUEFN)
{
/* check if it conflict with a builtin function name */
if (ase_awk_getbfn (awk, name, len) != ASE_NULL)
{
SETERRARG (
awk, ASE_AWK_EBFNRED, awk->token.line,
name, len);
return -1;
}
/* check if it conflict with a function name */
if (ase_awk_map_get (
awk->tree.afns, name, len) != ASE_NULL)
{
SETERRARG (
awk, ASE_AWK_EAFNRED, line,
name, len);
return -1;
}
}
/* check if it conflicts with other global variable names */
if (find_global (awk, name, len) != (ase_size_t)-1)
{
SETERRARG (awk, ASE_AWK_EDUPGBL, line, name, len);
return -1;
}
nglobals = ase_awk_tab_getsize (&awk->parse.globals);
if (nglobals >= ASE_AWK_MAX_GLOBALS)
{
SETERRLIN (awk, ASE_AWK_EGBLTM, line);
return -1;
}
if (ase_awk_tab_add (&awk->parse.globals, name, len) == (ase_size_t)-1)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return -1;
}
/* the disabled item is inserted normally but
* the name length is reset to zero. */
if (disabled) awk->parse.globals.buf[nglobals].name.len = 0;
awk->tree.nglobals = ase_awk_tab_getsize (&awk->parse.globals);
ASE_ASSERT (nglobals == awk->tree.nglobals-1);
/* return the id which is the index to the global table. */
return (int)nglobals;
}
int ase_awk_addglobal (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len)
{
int n;
if (len <= 0)
{
SETERR (awk, ASE_AWK_EINVAL);
return -1;
}
if (awk->tree.nglobals > awk->tree.nbglobals)
{
/* this function is not allow after ase_awk_parse is called */
SETERR (awk, ASE_AWK_ENOPER);
return -1;
}
n = add_global (awk, name, len, 0, 0);
/* update the count of the static globals.
* the total global count has been updated inside add_global. */
if (n >= 0) awk->tree.nbglobals++;
return n;
}
int ase_awk_delglobal (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len)
{
ase_size_t n;
#define ASE_AWK_NUM_STATIC_GLOBALS \
(ASE_AWK_MAX_GLOBAL_ID-ASE_AWK_MIN_GLOBAL_ID+1)
if (awk->tree.nglobals > awk->tree.nbglobals)
{
/* this function is not allow after ase_awk_parse is called */
SETERR (awk, ASE_AWK_ENOPER);
return -1;
}
n = ase_awk_tab_find (&awk->parse.globals,
ASE_AWK_NUM_STATIC_GLOBALS, name, len);
if (n == (ase_size_t)-1)
{
SETERRARG (awk, ASE_AWK_ENOENT, 0, name, len);
return -1;
}
/* invalidate the name if deletion is requested.
* this approach does not delete the entry.
* if ase_awk_addglobal is called with the same name
* again, the entry will be appended again.
* never call this funciton unless it is really required. */
awk->parse.globals.buf[n].name.ptr[0] = ASE_T('\0');;
awk->parse.globals.buf[n].name.len = 0;
return 0;
}
static ase_awk_t* collect_globals (ase_awk_t* awk)
{
while (1)
{
if (!MATCH(awk,TOKEN_IDENT))
{
SETERRTOK (awk, ASE_AWK_EBADVAR);
return ASE_NULL;
}
if (add_global (
awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name),
awk->token.line, 0) == -1) return ASE_NULL;
if (get_token(awk) == -1) return ASE_NULL;
if (MATCH(awk,TOKEN_SEMICOLON)) break;
if (!MATCH(awk,TOKEN_COMMA))
{
SETERRTOK (awk, ASE_AWK_ECOMMA);
return ASE_NULL;
}
if (get_token(awk) == -1) return ASE_NULL;
}
/* skip a semicolon */
if (get_token(awk) == -1) return ASE_NULL;
return awk;
}
static ase_awk_t* collect_locals (ase_awk_t* awk, ase_size_t nlocals)
{
ase_char_t* local;
ase_size_t local_len;
ase_size_t n;
while (1)
{
if (!MATCH(awk,TOKEN_IDENT))
{
SETERRTOK (awk, ASE_AWK_EBADVAR);
return ASE_NULL;
}
local = ASE_STR_BUF(&awk->token.name);
local_len = ASE_STR_LEN(&awk->token.name);
if (awk->option & ASE_AWK_UNIQUEFN)
{
/* check if it conflict with a builtin function name */
if (ase_awk_getbfn (awk, local, local_len) != ASE_NULL)
{
SETERRARG (
awk, ASE_AWK_EBFNRED, awk->token.line,
local, local_len);
return ASE_NULL;
}
/* check if it conflict with a function name */
if (ase_strxncmp (
awk->tree.cur_afn.ptr, awk->tree.cur_afn.len,
local, local_len) == 0 ||
ase_awk_map_get (
awk->tree.afns, local, local_len) != ASE_NULL)
{
SETERRARG (
awk, ASE_AWK_EAFNRED, awk->token.line,
local, local_len);
return ASE_NULL;
}
}
/* check if it conflicts with a paremeter name */
n = ase_awk_tab_find (&awk->parse.params, 0, local, local_len);
if (n != (ase_size_t)-1)
{
SETERRARG (
awk, ASE_AWK_EPARRED, awk->token.line,
local, local_len);
return ASE_NULL;
}
/* check if it conflicts with other local variable names */
n = ase_awk_tab_find (
&awk->parse.locals,
((awk->option & ASE_AWK_SHADING)? nlocals: 0),
local, local_len);
if (n != (ase_size_t)-1)
{
SETERRARG (
awk, ASE_AWK_EDUPLCL, awk->token.line,
local, local_len);
return ASE_NULL;
}
/* check if it conflicts with global variable names */
n = find_global (awk, local, local_len);
if (n != (ase_size_t)-1)
{
if (n < awk->tree.nbglobals)
{
/* conflicting with a static global variable */
SETERRARG (
awk, ASE_AWK_EDUPLCL, awk->token.line,
local, local_len);
return ASE_NULL;
}
if ((awk->option & ASE_AWK_SHADING) == 0)
{
/* conflicting with a normal global variable */
SETERRARG (
awk, ASE_AWK_EDUPLCL, awk->token.line,
local, local_len);
return ASE_NULL;
}
}
if (ase_awk_tab_getsize(&awk->parse.locals) >= ASE_AWK_MAX_LOCALS)
{
SETERRLIN (awk, ASE_AWK_ELCLTM, awk->token.line);
return ASE_NULL;
}
if (ase_awk_tab_add (
&awk->parse.locals, local, local_len) == (ase_size_t)-1)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, awk->token.line);
return ASE_NULL;
}
if (get_token(awk) == -1) return ASE_NULL;
if (MATCH(awk,TOKEN_SEMICOLON)) break;
if (!MATCH(awk,TOKEN_COMMA))
{
SETERRTOK (awk, ASE_AWK_ECOMMA);
return ASE_NULL;
}
if (get_token(awk) == -1) return ASE_NULL;
}
/* skip a semicolon */
if (get_token(awk) == -1) return ASE_NULL;
return awk;
}
static ase_awk_nde_t* parse_statement (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* nde;
if (MATCH(awk,TOKEN_SEMICOLON))
{
/* null statement */
nde = (ase_awk_nde_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_NULL;
nde->line = line;
nde->next = ASE_NULL;
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
}
else if (MATCH(awk,TOKEN_LBRACE))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = awk->parse.parse_block (
awk, awk->token.prev.line, ase_false);
}
else
{
/* the statement id held in awk->parse.id.stmnt denotes
* the token id of the statement currently being parsed.
* the current statement id is saved here because the
* statement id can be changed in parse_statement_nb.
* it will, in turn, call parse_statement which will
* eventually change the statement id. */
int old_id = awk->parse.id.stmnt;
/* set the current statement id */
awk->parse.id.stmnt = awk->token.type;
/* proceed parsing the statement */
nde = parse_statement_nb (awk, line);
/* restore the statement id saved previously */
awk->parse.id.stmnt = old_id;
}
return nde;
}
static ase_awk_nde_t* parse_statement_nb (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* nde;
/* keywords that don't require any terminating semicolon */
if (MATCH(awk,TOKEN_IF))
{
if (get_token(awk) == -1) return ASE_NULL;
return parse_if (awk, line);
}
else if (MATCH(awk,TOKEN_WHILE))
{
if (get_token(awk) == -1) return ASE_NULL;
awk->parse.depth.cur.loop++;
nde = parse_while (awk, line);
awk->parse.depth.cur.loop--;
return nde;
}
else if (MATCH(awk,TOKEN_FOR))
{
if (get_token(awk) == -1) return ASE_NULL;
awk->parse.depth.cur.loop++;
nde = parse_for (awk, line);
awk->parse.depth.cur.loop--;
return nde;
}
/* keywords that require a terminating semicolon */
if (MATCH(awk,TOKEN_DO))
{
if (get_token(awk) == -1) return ASE_NULL;
awk->parse.depth.cur.loop++;
nde = parse_dowhile (awk, line);
awk->parse.depth.cur.loop--;
return nde;
}
else if (MATCH(awk,TOKEN_BREAK))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_break (awk, line);
}
else if (MATCH(awk,TOKEN_CONTINUE))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_continue (awk, line);
}
else if (MATCH(awk,TOKEN_RETURN))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_return (awk, line);
}
else if (MATCH(awk,TOKEN_EXIT))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_exit (awk, line);
}
else if (MATCH(awk,TOKEN_NEXT))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_next (awk, line);
}
else if (MATCH(awk,TOKEN_NEXTFILE))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_nextfile (awk, line, 0);
}
else if (MATCH(awk,TOKEN_NEXTOFILE))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_nextfile (awk, line, 1);
}
else if (MATCH(awk,TOKEN_DELETE))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_delete (awk, line);
}
else if (MATCH(awk,TOKEN_RESET))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_reset (awk, line);
}
else if (MATCH(awk,TOKEN_PRINT))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_print (awk, line, ASE_AWK_NDE_PRINT);
}
else if (MATCH(awk,TOKEN_PRINTF))
{
if (get_token(awk) == -1) return ASE_NULL;
nde = parse_print (awk, line, ASE_AWK_NDE_PRINTF);
}
else
{
nde = parse_expression (awk, line);
}
if (nde == ASE_NULL) return ASE_NULL;
/* check if a statement ends with a semicolon */
if (!MATCH(awk,TOKEN_SEMICOLON))
{
if (nde != ASE_NULL) ase_awk_clrpt (awk, nde);
SETERRLIN (awk, ASE_AWK_ESTMEND, awk->token.prev.line);
return ASE_NULL;
}
/* eat up the semicolon and read in the next token */
if (get_token(awk) == -1)
{
if (nde != ASE_NULL) ase_awk_clrpt (awk, nde);
return ASE_NULL;
}
return nde;
}
static ase_awk_nde_t* parse_expression (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* nde;
if (awk->parse.depth.max.expr > 0 &&
awk->parse.depth.cur.expr >= awk->parse.depth.max.expr)
{
SETERRLIN (awk, ASE_AWK_EEXPRNST, line);
return ASE_NULL;
}
awk->parse.depth.cur.expr++;
nde = parse_expression0 (awk, line);
awk->parse.depth.cur.expr--;
return nde;
}
static ase_awk_nde_t* parse_expression0 (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* x, * y;
ase_awk_nde_ass_t* nde;
int opcode;
x = parse_basic_expr (awk, line);
if (x == ASE_NULL) return ASE_NULL;
opcode = assign_to_opcode (awk);
if (opcode == -1)
{
/* no assignment operator found. */
return x;
}
ASE_ASSERT (x->next == ASE_NULL);
if (!is_var(x) && x->type != ASE_AWK_NDE_POS)
{
ase_awk_clrpt (awk, x);
SETERRLIN (awk, ASE_AWK_EASSIGN, line);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, x);
return ASE_NULL;
}
/*y = parse_basic_expr (awk);*/
y = parse_expression (awk, awk->token.line);
if (y == ASE_NULL)
{
ase_awk_clrpt (awk, x);
return ASE_NULL;
}
nde = (ase_awk_nde_ass_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_ass_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, x);
ase_awk_clrpt (awk, y);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_ASS;
nde->line = line;
nde->next = ASE_NULL;
nde->opcode = opcode;
nde->left = x;
nde->right = y;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_basic_expr (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* nde, * n1, * n2;
nde = parse_logical_or (awk, line);
if (nde == ASE_NULL) return ASE_NULL;
if (MATCH(awk,TOKEN_QUEST))
{
ase_awk_nde_cnd_t* tmp;
if (get_token(awk) == -1) return ASE_NULL;
/*n1 = parse_basic_expr (awk, awk->token.line);*/
n1 = parse_expression (awk, awk->token.line);
if (n1 == ASE_NULL)
{
ase_awk_clrpt (awk, nde);
return ASE_NULL;
}
if (!MATCH(awk,TOKEN_COLON))
{
SETERRTOK (awk, ASE_AWK_ECOLON);
return ASE_NULL;
}
if (get_token(awk) == -1) return ASE_NULL;
/*n2 = parse_basic_expr (awk, awk->token.line);*/
n2 = parse_expression (awk, awk->token.line);
if (n2 == ASE_NULL)
{
ase_awk_clrpt (awk, nde);
ase_awk_clrpt (awk, n1);
return ASE_NULL;
}
tmp = (ase_awk_nde_cnd_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_cnd_t));
if (tmp == ASE_NULL)
{
ase_awk_clrpt (awk, nde);
ase_awk_clrpt (awk, n1);
ase_awk_clrpt (awk, n2);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
tmp->type = ASE_AWK_NDE_CND;
tmp->line = line;
tmp->next = ASE_NULL;
tmp->test = nde;
tmp->left = n1;
tmp->right = n2;
nde = (ase_awk_nde_t*)tmp;
}
return nde;
}
static ase_awk_nde_t* parse_binary_expr (
ase_awk_t* awk, ase_size_t line, const binmap_t* binmap,
ase_awk_nde_t*(*next_level_func)(ase_awk_t*,ase_size_t))
{
ase_awk_nde_exp_t* nde;
ase_awk_nde_t* left, * right;
int opcode;
left = next_level_func (awk, line);
if (left == ASE_NULL) return ASE_NULL;
while (1)
{
const binmap_t* p = binmap;
ase_bool_t matched = ase_false;
while (p->token != TOKEN_EOF)
{
if (MATCH(awk,p->token))
{
opcode = p->binop;
matched = ase_true;
break;
}
p++;
}
if (!matched) break;
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
right = next_level_func (awk, awk->token.line);
if (right == ASE_NULL)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
nde = (ase_awk_nde_exp_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_exp_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, right);
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_EXP_BIN;
nde->line = line;
nde->next = ASE_NULL;
nde->opcode = opcode;
nde->left = left;
nde->right = right;
left = (ase_awk_nde_t*)nde;
}
return left;
}
static ase_awk_nde_t* parse_logical_or (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_LOR, ASE_AWK_BINOP_LOR },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_logical_and);
}
static ase_awk_nde_t* parse_logical_and (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_LAND, ASE_AWK_BINOP_LAND },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_in);
}
static ase_awk_nde_t* parse_in (ase_awk_t* awk, ase_size_t line)
{
/*
static binmap_t map[] =
{
{ TOKEN_IN, ASE_AWK_BINOP_IN },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_regex_match);
*/
ase_awk_nde_exp_t* nde;
ase_awk_nde_t* left, * right;
ase_size_t line2;
left = parse_regex_match (awk, line);
if (left == ASE_NULL) return ASE_NULL;
while (1)
{
if (!MATCH(awk,TOKEN_IN)) break;
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
line2 = awk->token.line;
right = parse_regex_match (awk, line2);
if (right == ASE_NULL)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
if (!is_plain_var(right))
{
ase_awk_clrpt (awk, right);
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_ENOTVAR, line2);
return ASE_NULL;
}
nde = (ase_awk_nde_exp_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_exp_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, right);
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_EXP_BIN;
nde->line = line;
nde->next = ASE_NULL;
nde->opcode = ASE_AWK_BINOP_IN;
nde->left = left;
nde->right = right;
left = (ase_awk_nde_t*)nde;
}
return left;
}
static ase_awk_nde_t* parse_regex_match (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_TILDE, ASE_AWK_BINOP_MA },
{ TOKEN_NM, ASE_AWK_BINOP_NM },
{ TOKEN_EOF, 0 },
};
return parse_binary_expr (awk, line, map, parse_bitwise_or);
}
static ase_awk_nde_t* parse_bitwise_or (ase_awk_t* awk, ase_size_t line)
{
if (awk->option & ASE_AWK_EXTIO)
{
return parse_bitwise_or_with_extio (awk, line);
}
else
{
static binmap_t map[] =
{
{ TOKEN_BOR, ASE_AWK_BINOP_BOR },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (
awk, line, map, parse_bitwise_xor);
}
}
static ase_awk_nde_t* parse_bitwise_or_with_extio (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* left, * right;
left = parse_bitwise_xor (awk, line);
if (left == ASE_NULL) return ASE_NULL;
while (1)
{
int in_type;
if (MATCH(awk,TOKEN_BOR))
in_type = ASE_AWK_IN_PIPE;
else if (MATCH(awk,TOKEN_BORAND))
in_type = ASE_AWK_IN_COPROC;
else break;
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
if (MATCH(awk,TOKEN_GETLINE))
{
ase_awk_nde_getline_t* nde;
ase_awk_nde_t* var = ASE_NULL;
/* piped getline */
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
/* TODO: is this correct? */
if (MATCH(awk,TOKEN_IDENT))
{
/* command | getline var */
var = parse_primary (awk, awk->token.line);
if (var == ASE_NULL)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
}
nde = (ase_awk_nde_getline_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_getline_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_GETLINE;
nde->line = line;
nde->next = ASE_NULL;
nde->var = var;
nde->in_type = in_type;
nde->in = left;
left = (ase_awk_nde_t*)nde;
}
else
{
ase_awk_nde_exp_t* nde;
if (in_type == ASE_AWK_IN_COPROC)
{
ase_awk_clrpt (awk, left);
/* TODO: support coprocess */
SETERRLIN (awk, ASE_AWK_EGLNCPS, line);
return ASE_NULL;
}
right = parse_bitwise_xor (awk, awk->token.line);
if (right == ASE_NULL)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
nde = (ase_awk_nde_exp_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_exp_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, right);
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_EXP_BIN;
nde->line = line;
nde->next = ASE_NULL;
nde->opcode = ASE_AWK_BINOP_BOR;
nde->left = left;
nde->right = right;
left = (ase_awk_nde_t*)nde;
}
}
return left;
}
static ase_awk_nde_t* parse_bitwise_xor (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_BXOR, ASE_AWK_BINOP_BXOR },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_bitwise_and);
}
static ase_awk_nde_t* parse_bitwise_and (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_BAND, ASE_AWK_BINOP_BAND },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_equality);
}
static ase_awk_nde_t* parse_equality (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_EQ, ASE_AWK_BINOP_EQ },
{ TOKEN_NE, ASE_AWK_BINOP_NE },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_relational);
}
static ase_awk_nde_t* parse_relational (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_GT, ASE_AWK_BINOP_GT },
{ TOKEN_GE, ASE_AWK_BINOP_GE },
{ TOKEN_LT, ASE_AWK_BINOP_LT },
{ TOKEN_LE, ASE_AWK_BINOP_LE },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_shift);
}
static ase_awk_nde_t* parse_shift (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_LSHIFT, ASE_AWK_BINOP_LSHIFT },
{ TOKEN_RSHIFT, ASE_AWK_BINOP_RSHIFT },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_concat);
}
static ase_awk_nde_t* parse_concat (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_exp_t* nde;
ase_awk_nde_t* left, * right;
left = parse_additive (awk, line);
if (left == ASE_NULL) return ASE_NULL;
while (1)
{
if (MATCH(awk,TOKEN_PERIOD))
{
if (!(awk->option & ASE_AWK_EXPLICIT)) break;
if (get_token(awk) == -1) return ASE_NULL;
}
else if (MATCH(awk,TOKEN_LPAREN) ||
MATCH(awk,TOKEN_DOLLAR) ||
MATCH(awk,TOKEN_PLUS) ||
MATCH(awk,TOKEN_MINUS) ||
MATCH(awk,TOKEN_PLUSPLUS) ||
MATCH(awk,TOKEN_MINUSMINUS) ||
awk->token.type >= TOKEN_GETLINE)
{
/* TODO: is the check above sufficient? */
if (!(awk->option & ASE_AWK_IMPLICIT)) break;
}
else break;
right = parse_additive (awk, awk->token.line);
if (right == ASE_NULL)
{
ase_awk_clrpt (awk, left);
return ASE_NULL;
}
nde = (ase_awk_nde_exp_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_exp_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, left);
ase_awk_clrpt (awk, right);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_EXP_BIN;
nde->line = line;
nde->next = ASE_NULL;
nde->opcode = ASE_AWK_BINOP_CONCAT;
nde->left = left;
nde->right = right;
left = (ase_awk_nde_t*)nde;
}
return left;
}
static ase_awk_nde_t* parse_additive (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_PLUS, ASE_AWK_BINOP_PLUS },
{ TOKEN_MINUS, ASE_AWK_BINOP_MINUS },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_multiplicative);
}
static ase_awk_nde_t* parse_multiplicative (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_MUL, ASE_AWK_BINOP_MUL },
{ TOKEN_DIV, ASE_AWK_BINOP_DIV },
{ TOKEN_IDIV, ASE_AWK_BINOP_IDIV },
{ TOKEN_MOD, ASE_AWK_BINOP_MOD },
/* { TOKEN_EXP, ASE_AWK_BINOP_EXP }, */
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_unary);
}
static ase_awk_nde_t* parse_unary (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_exp_t* nde;
ase_awk_nde_t* left;
int opcode;
opcode = (MATCH(awk,TOKEN_PLUS))? ASE_AWK_UNROP_PLUS:
(MATCH(awk,TOKEN_MINUS))? ASE_AWK_UNROP_MINUS:
(MATCH(awk,TOKEN_NOT))? ASE_AWK_UNROP_NOT:
(MATCH(awk,TOKEN_TILDE))? ASE_AWK_UNROP_BNOT: -1;
/*if (opcode == -1) return parse_increment (awk);*/
if (opcode == -1) return parse_exponent (awk, line);
if (get_token(awk) == -1) return ASE_NULL;
if (awk->parse.depth.max.expr > 0 &&
awk->parse.depth.cur.expr >= awk->parse.depth.max.expr)
{
SETERRLIN (awk, ASE_AWK_EEXPRNST, awk->token.line);
return ASE_NULL;
}
awk->parse.depth.cur.expr++;
left = parse_unary (awk, awk->token.line);
awk->parse.depth.cur.expr--;
if (left == ASE_NULL) return ASE_NULL;
nde = (ase_awk_nde_exp_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_exp_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_EXP_UNR;
nde->line = line;
nde->next = ASE_NULL;
nde->opcode = opcode;
nde->left = left;
nde->right = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_exponent (ase_awk_t* awk, ase_size_t line)
{
static binmap_t map[] =
{
{ TOKEN_EXP, ASE_AWK_BINOP_EXP },
{ TOKEN_EOF, 0 }
};
return parse_binary_expr (awk, line, map, parse_unary_exp);
}
static ase_awk_nde_t* parse_unary_exp (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_exp_t* nde;
ase_awk_nde_t* left;
int opcode;
opcode = (MATCH(awk,TOKEN_PLUS))? ASE_AWK_UNROP_PLUS:
(MATCH(awk,TOKEN_MINUS))? ASE_AWK_UNROP_MINUS:
(MATCH(awk,TOKEN_NOT))? ASE_AWK_UNROP_NOT:
(MATCH(awk,TOKEN_TILDE))? ASE_AWK_UNROP_BNOT: -1;
if (opcode == -1) return parse_increment (awk, line);
if (get_token(awk) == -1) return ASE_NULL;
if (awk->parse.depth.max.expr > 0 &&
awk->parse.depth.cur.expr >= awk->parse.depth.max.expr)
{
SETERRLIN (awk, ASE_AWK_EEXPRNST, awk->token.line);
return ASE_NULL;
}
awk->parse.depth.cur.expr++;
left = parse_unary (awk, awk->token.line);
awk->parse.depth.cur.expr--;
if (left == ASE_NULL) return ASE_NULL;
nde = (ase_awk_nde_exp_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_exp_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_EXP_UNR;
nde->line = line;
nde->next = ASE_NULL;
nde->opcode = opcode;
nde->left = left;
nde->right = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_increment (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_exp_t* nde;
ase_awk_nde_t* left;
int type, opcode, opcode1, opcode2;
/* check for prefix increment operator */
opcode1 = MATCH(awk,TOKEN_PLUSPLUS)? ASE_AWK_INCOP_PLUS:
MATCH(awk,TOKEN_MINUSMINUS)? ASE_AWK_INCOP_MINUS: -1;
if (opcode1 != -1)
{
if (get_token(awk) == -1) return ASE_NULL;
}
left = parse_primary (awk, line);
if (left == ASE_NULL) return ASE_NULL;
if (awk->option & ASE_AWK_IMPLICIT)
{
/* concatenation operation by whitepaces are allowed
* if this option is set. the ++/-- operator following
* the primary should be treated specially.
* for example, "abc" ++ 10 => "abc" . ++10
*/
/* TOOD: probably left->type == ASE_AWK_NDE_POS should
* be inclued to handle $i++, $i--. */
if (!is_var(left)) return left;
}
/* check for postfix increment operator */
opcode2 = MATCH(awk,TOKEN_PLUSPLUS)? ASE_AWK_INCOP_PLUS:
MATCH(awk,TOKEN_MINUSMINUS)? ASE_AWK_INCOP_MINUS: -1;
if (opcode1 != -1 && opcode2 != -1)
{
/* both prefix and postfix increment operator.
* not allowed */
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_EPREPST, line);
return ASE_NULL;
}
else if (opcode1 == -1 && opcode2 == -1)
{
/* no increment operators */
return left;
}
else if (opcode1 != -1)
{
/* prefix increment operator */
type = ASE_AWK_NDE_EXP_INCPRE;
opcode = opcode1;
}
else if (opcode2 != -1)
{
/* postfix increment operator */
type = ASE_AWK_NDE_EXP_INCPST;
opcode = opcode2;
if (get_token(awk) == -1) return ASE_NULL;
}
nde = (ase_awk_nde_exp_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_exp_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, left);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = type;
nde->line = line;
nde->next = ASE_NULL;
nde->opcode = opcode;
nde->left = left;
nde->right = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_primary (ase_awk_t* awk, ase_size_t line)
{
if (MATCH(awk,TOKEN_IDENT))
{
return parse_primary_ident (awk, line);
}
else if (MATCH(awk,TOKEN_INT))
{
ase_awk_nde_int_t* nde;
nde = (ase_awk_nde_int_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_int_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_INT;
nde->line = line;
nde->next = ASE_NULL;
nde->val = ase_awk_strxtolong (awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name), 0, ASE_NULL);
nde->str = ase_awk_strxdup (awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name));
if (nde->str == ASE_NULL)
{
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
nde->len = ASE_STR_LEN(&awk->token.name);
ASE_ASSERT (
ASE_STR_LEN(&awk->token.name) ==
ase_strlen(ASE_STR_BUF(&awk->token.name)));
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, nde->str);
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
return (ase_awk_nde_t*)nde;
}
else if (MATCH(awk,TOKEN_REAL))
{
ase_awk_nde_real_t* nde;
nde = (ase_awk_nde_real_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_real_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_REAL;
nde->line = line;
nde->next = ASE_NULL;
nde->val = ase_awk_strxtoreal (awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name), ASE_NULL);
nde->str = ase_awk_strxdup (awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name));
if (nde->str == ASE_NULL)
{
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
nde->len = ASE_STR_LEN(&awk->token.name);
ASE_ASSERT (
ASE_STR_LEN(&awk->token.name) ==
ase_strlen(ASE_STR_BUF(&awk->token.name)));
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, nde->str);
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
return (ase_awk_nde_t*)nde;
}
else if (MATCH(awk,TOKEN_STR))
{
ase_awk_nde_str_t* nde;
nde = (ase_awk_nde_str_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_str_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_STR;
nde->line = line;
nde->next = ASE_NULL;
nde->len = ASE_STR_LEN(&awk->token.name);
nde->buf = ase_awk_strxdup (awk,
ASE_STR_BUF(&awk->token.name), nde->len);
if (nde->buf == ASE_NULL)
{
ASE_AWK_FREE (awk, nde);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, nde->buf);
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
return (ase_awk_nde_t*)nde;
}
else if (MATCH(awk,TOKEN_DIV))
{
ase_awk_nde_rex_t* nde;
int errnum;
/* the regular expression is tokenized here because
* of the context-sensitivity of the slash symbol */
SET_TOKEN_TYPE (awk, TOKEN_REX);
ase_str_clear (&awk->token.name);
if (get_rexstr (awk) == -1) return ASE_NULL;
ASE_ASSERT (MATCH(awk,TOKEN_REX));
nde = (ase_awk_nde_rex_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_rex_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_REX;
nde->line = line;
nde->next = ASE_NULL;
nde->len = ASE_STR_LEN(&awk->token.name);
nde->buf = ase_awk_strxdup (awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name));
if (nde->buf == ASE_NULL)
{
ASE_AWK_FREE (awk, nde);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->code = ase_awk_buildrex (awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name),
&errnum);
if (nde->code == ASE_NULL)
{
ASE_AWK_FREE (awk, nde->buf);
ASE_AWK_FREE (awk, nde);
SETERRLIN (awk, errnum, line);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, nde->buf);
ASE_AWK_FREE (awk, nde->code);
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
return (ase_awk_nde_t*)nde;
}
else if (MATCH(awk,TOKEN_DOLLAR))
{
ase_awk_nde_pos_t* nde;
ase_awk_nde_t* prim;
if (get_token(awk)) return ASE_NULL;
prim = parse_primary (awk, awk->token.line);
if (prim == ASE_NULL) return ASE_NULL;
nde = (ase_awk_nde_pos_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_pos_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, prim);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_POS;
nde->line = line;
nde->next = ASE_NULL;
nde->val = prim;
return (ase_awk_nde_t*)nde;
}
else if (MATCH(awk,TOKEN_LPAREN))
{
ase_awk_nde_t* nde;
ase_awk_nde_t* last;
/* eat up the left parenthesis */
if (get_token(awk) == -1) return ASE_NULL;
/* parse the sub-expression inside the parentheses */
nde = parse_expression (awk, awk->token.line);
if (nde == ASE_NULL) return ASE_NULL;
/* parse subsequent expressions separated by a comma, if any */
last = nde;
ASE_ASSERT (last->next == ASE_NULL);
while (MATCH(awk,TOKEN_COMMA))
{
ase_awk_nde_t* tmp;
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, nde);
return ASE_NULL;
}
tmp = parse_expression (awk, awk->token.line);
if (tmp == ASE_NULL)
{
ase_awk_clrpt (awk, nde);
return ASE_NULL;
}
ASE_ASSERT (tmp->next == ASE_NULL);
last->next = tmp;
last = tmp;
}
/* ----------------- */
/* check for the closing parenthesis */
if (!MATCH(awk,TOKEN_RPAREN))
{
ase_awk_clrpt (awk, nde);
SETERRTOK (awk, ASE_AWK_ERPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, nde);
return ASE_NULL;
}
/* check if it is a chained node */
if (nde->next != ASE_NULL)
{
/* if so, it is a expression group */
/* (expr1, expr2, expr2) */
ase_awk_nde_grp_t* tmp;
if ((awk->parse.id.stmnt != TOKEN_PRINT &&
awk->parse.id.stmnt != TOKEN_PRINTF) ||
awk->parse.depth.cur.expr != 1)
{
if (!MATCH(awk,TOKEN_IN))
{
ase_awk_clrpt (awk, nde);
SETERRTOK (awk, ASE_AWK_EIN);
return ASE_NULL;
}
}
tmp = (ase_awk_nde_grp_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_grp_t));
if (tmp == ASE_NULL)
{
ase_awk_clrpt (awk, nde);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
tmp->type = ASE_AWK_NDE_GRP;
tmp->line = line;
tmp->next = ASE_NULL;
tmp->body = nde;
nde = (ase_awk_nde_t*)tmp;
}
/* ----------------- */
return nde;
}
else if (MATCH(awk,TOKEN_GETLINE))
{
ase_awk_nde_getline_t* nde;
ase_awk_nde_t* var = ASE_NULL;
ase_awk_nde_t* in = ASE_NULL;
if (get_token(awk) == -1) return ASE_NULL;
if (MATCH(awk,TOKEN_IDENT))
{
/* getline var */
var = parse_primary (awk, awk->token.line);
if (var == ASE_NULL) return ASE_NULL;
}
if (MATCH(awk, TOKEN_LT))
{
/* getline [var] < file */
if (get_token(awk) == -1)
{
if (var != ASE_NULL) ase_awk_clrpt (awk, var);
return ASE_NULL;
}
/* TODO: is this correct? */
/*in = parse_expression (awk);*/
in = parse_primary (awk, awk->token.line);
if (in == ASE_NULL)
{
if (var != ASE_NULL) ase_awk_clrpt (awk, var);
return ASE_NULL;
}
}
nde = (ase_awk_nde_getline_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_getline_t));
if (nde == ASE_NULL)
{
if (var != ASE_NULL) ase_awk_clrpt (awk, var);
if (in != ASE_NULL) ase_awk_clrpt (awk, in);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_GETLINE;
nde->line = line;
nde->next = ASE_NULL;
nde->var = var;
nde->in_type = (in == ASE_NULL)?
ASE_AWK_IN_CONSOLE: ASE_AWK_IN_FILE;
nde->in = in;
return (ase_awk_nde_t*)nde;
}
/* valid expression introducer is expected */
SETERRLIN (awk, ASE_AWK_EEXPRES,
(MATCH(awk,TOKEN_EOF)? awk->token.prev.line: line));
return ASE_NULL;
}
static ase_awk_nde_t* parse_primary_ident (ase_awk_t* awk, ase_size_t line)
{
ase_char_t* name_dup;
ase_size_t name_len;
ase_awk_bfn_t* bfn;
ASE_ASSERT (MATCH(awk,TOKEN_IDENT));
name_dup = ase_awk_strxdup (awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name));
if (name_dup == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
name_len = ASE_STR_LEN(&awk->token.name);
if (get_token(awk) == -1)
{
ASE_AWK_FREE (awk, name_dup);
return ASE_NULL;
}
/* check if name_dup is an intrinsic function name */
bfn = ase_awk_getbfn (awk, name_dup, name_len);
if (bfn != ASE_NULL)
{
ase_awk_nde_t* nde;
if (!MATCH(awk,TOKEN_LPAREN))
{
/* an intrinsic function should be in the form
* of the function call */
ASE_AWK_FREE (awk, name_dup);
SETERRTOK (awk, ASE_AWK_ELPAREN);
return ASE_NULL;
}
nde = parse_fncall (awk, name_dup, name_len, bfn, line);
if (nde == ASE_NULL) ASE_AWK_FREE (awk, name_dup);
return (ase_awk_nde_t*)nde;
}
/* now we know that name_dup is a normal identifier. */
if (MATCH(awk,TOKEN_LBRACK))
{
ase_awk_nde_t* nde;
nde = parse_hashidx (awk, name_dup, name_len, line);
if (nde == ASE_NULL) ASE_AWK_FREE (awk, name_dup);
return (ase_awk_nde_t*)nde;
}
else if (MATCH(awk,TOKEN_LPAREN))
{
/* function call */
ase_awk_nde_t* nde;
nde = parse_fncall (awk, name_dup, name_len, ASE_NULL, line);
if (nde == ASE_NULL) ASE_AWK_FREE (awk, name_dup);
return (ase_awk_nde_t*)nde;
}
else
{
/* normal variable */
ase_awk_nde_var_t* nde;
ase_size_t idxa;
nde = (ase_awk_nde_var_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_var_t));
if (nde == ASE_NULL)
{
ASE_AWK_FREE (awk, name_dup);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
/* search the parameter name list */
idxa = ase_awk_tab_find (
&awk->parse.params, 0, name_dup, name_len);
if (idxa != (ase_size_t)-1)
{
nde->type = ASE_AWK_NDE_ARG;
nde->line = line;
nde->next = ASE_NULL;
/*nde->id.name = ASE_NULL;*/
nde->id.name = name_dup;
nde->id.name_len = name_len;
nde->id.idxa = idxa;
nde->idx = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
/* search the local variable list */
idxa = ase_awk_tab_rrfind (
&awk->parse.locals, 0, name_dup, name_len);
if (idxa != (ase_size_t)-1)
{
nde->type = ASE_AWK_NDE_LOCAL;
nde->line = line;
nde->next = ASE_NULL;
/*nde->id.name = ASE_NULL;*/
nde->id.name = name_dup;
nde->id.name_len = name_len;
nde->id.idxa = idxa;
nde->idx = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
/* gets the global variable index */
idxa = get_global (awk, name_dup, name_len);
if (idxa != (ase_size_t)-1)
{
nde->type = ASE_AWK_NDE_GLOBAL;
nde->line = line;
nde->next = ASE_NULL;
/*nde->id.name = ASE_NULL;*/
nde->id.name = name_dup;
nde->id.name_len = name_len;
nde->id.idxa = idxa;
nde->idx = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
if (awk->option & ASE_AWK_IMPLICIT)
{
if (awk->option & ASE_AWK_UNIQUEFN)
{
/* the name should not conflict with a function name */
/* check if it is a builtin function */
if (ase_awk_getbfn (awk, name_dup, name_len) != ASE_NULL)
{
SETERRARG (awk, ASE_AWK_EBFNRED, line, name_dup, name_len);
goto exit_func;
}
/* check if it is an AWK function */
if (ase_awk_map_get(awk->tree.afns, name_dup, name_len) != ASE_NULL)
{
/* the function is defined previously */
SETERRARG (awk, ASE_AWK_EAFNRED, line, name_dup, name_len);
goto exit_func;
}
}
nde->type = ASE_AWK_NDE_NAMED;
nde->line = line;
nde->next = ASE_NULL;
nde->id.name = name_dup;
nde->id.name_len = name_len;
nde->id.idxa = (ase_size_t)-1;
nde->idx = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
/* undefined variable */
SETERRARG (awk, ASE_AWK_EUNDEF, line, name_dup, name_len);
exit_func:
ASE_AWK_FREE (awk, name_dup);
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
}
static ase_awk_nde_t* parse_hashidx (
ase_awk_t* awk, ase_char_t* name, ase_size_t name_len, ase_size_t line)
{
ase_awk_nde_t* idx, * tmp, * last;
ase_awk_nde_var_t* nde;
ase_size_t idxa;
idx = ASE_NULL;
last = ASE_NULL;
do
{
if (get_token(awk) == -1)
{
if (idx != ASE_NULL) ase_awk_clrpt (awk, idx);
return ASE_NULL;
}
tmp = parse_expression (awk, awk->token.line);
if (tmp == ASE_NULL)
{
if (idx != ASE_NULL) ase_awk_clrpt (awk, idx);
return ASE_NULL;
}
if (idx == ASE_NULL)
{
ASE_ASSERT (last == ASE_NULL);
idx = tmp; last = tmp;
}
else
{
last->next = tmp;
last = tmp;
}
}
while (MATCH(awk,TOKEN_COMMA));
ASE_ASSERT (idx != ASE_NULL);
if (!MATCH(awk,TOKEN_RBRACK))
{
ase_awk_clrpt (awk, idx);
SETERRTOK (awk, ASE_AWK_ERBRACK);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, idx);
return ASE_NULL;
}
nde = (ase_awk_nde_var_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_var_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, idx);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
/* search the parameter name list */
idxa = ase_awk_tab_find (&awk->parse.params, 0, name, name_len);
if (idxa != (ase_size_t)-1)
{
nde->type = ASE_AWK_NDE_ARGIDX;
nde->line = line;
nde->next = ASE_NULL;
/*nde->id.name = ASE_NULL; */
nde->id.name = name;
nde->id.name_len = name_len;
nde->id.idxa = idxa;
nde->idx = idx;
return (ase_awk_nde_t*)nde;
}
/* search the local variable list */
idxa = ase_awk_tab_rrfind(&awk->parse.locals, 0, name, name_len);
if (idxa != (ase_size_t)-1)
{
nde->type = ASE_AWK_NDE_LOCALIDX;
nde->line = line;
nde->next = ASE_NULL;
/*nde->id.name = ASE_NULL; */
nde->id.name = name;
nde->id.name_len = name_len;
nde->id.idxa = idxa;
nde->idx = idx;
return (ase_awk_nde_t*)nde;
}
/* gets the global variable index */
idxa = get_global (awk, name, name_len);
if (idxa != (ase_size_t)-1)
{
nde->type = ASE_AWK_NDE_GLOBALIDX;
nde->line = line;
nde->next = ASE_NULL;
/*nde->id.name = ASE_NULL;*/
nde->id.name = name;
nde->id.name_len = name_len;
nde->id.idxa = idxa;
nde->idx = idx;
return (ase_awk_nde_t*)nde;
}
if (awk->option & ASE_AWK_IMPLICIT)
{
if (awk->option & ASE_AWK_UNIQUEFN)
{
/* check if it is a builtin function */
if (ase_awk_getbfn (awk, name, name_len) != ASE_NULL)
{
SETERRARG (awk, ASE_AWK_EBFNRED, line, name, name_len);
goto exit_func;
}
/* check if it is an AWK function */
if (ase_awk_map_get(awk->tree.afns, name, name_len) != ASE_NULL)
{
/* the function is defined previously */
SETERRARG (awk, ASE_AWK_EAFNRED, line, name, name_len);
goto exit_func;
}
}
nde->type = ASE_AWK_NDE_NAMEDIDX;
nde->line = line;
nde->next = ASE_NULL;
nde->id.name = name;
nde->id.name_len = name_len;
nde->id.idxa = (ase_size_t)-1;
nde->idx = idx;
return (ase_awk_nde_t*)nde;
}
/* undefined variable */
SETERRARG (awk, ASE_AWK_EUNDEF, line, name, name_len);
exit_func:
ase_awk_clrpt (awk, idx);
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
static ase_awk_nde_t* parse_fncall (
ase_awk_t* awk, ase_char_t* name, ase_size_t name_len,
ase_awk_bfn_t* bfn, ase_size_t line)
{
ase_awk_nde_t* head, * curr, * nde;
ase_awk_nde_call_t* call;
ase_size_t nargs;
if (get_token(awk) == -1) return ASE_NULL;
head = curr = ASE_NULL;
nargs = 0;
if (MATCH(awk,TOKEN_RPAREN))
{
/* no parameters to the function call */
if (get_token(awk) == -1) return ASE_NULL;
}
else
{
/* parse function parameters */
while (1)
{
nde = parse_expression (awk, awk->token.line);
if (nde == ASE_NULL)
{
if (head != ASE_NULL) ase_awk_clrpt (awk, head);
return ASE_NULL;
}
if (head == ASE_NULL) head = nde;
else curr->next = nde;
curr = nde;
nargs++;
if (MATCH(awk,TOKEN_RPAREN))
{
if (get_token(awk) == -1)
{
if (head != ASE_NULL)
ase_awk_clrpt (awk, head);
return ASE_NULL;
}
break;
}
if (!MATCH(awk,TOKEN_COMMA))
{
if (head != ASE_NULL) ase_awk_clrpt (awk, head);
SETERRTOK (awk, ASE_AWK_ECOMMA);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
if (head != ASE_NULL) ase_awk_clrpt (awk, head);
return ASE_NULL;
}
}
}
call = (ase_awk_nde_call_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_call_t));
if (call == ASE_NULL)
{
if (head != ASE_NULL) ase_awk_clrpt (awk, head);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
if (bfn != ASE_NULL)
{
call->type = ASE_AWK_NDE_BFN;
call->line = line;
call->next = ASE_NULL;
/*call->what.bfn = bfn; */
call->what.bfn.name.ptr = name;
call->what.bfn.name.len = name_len;
/* NOTE: oname is the original as in the bfn table.
* it would not duplicated here and not freed in
* ase_awk_clrpt either. so ase_awk_delfunc between
* ase_awk_parse and ase_awk_run may cause the program
* to fail. */
call->what.bfn.oname.ptr = bfn->name.ptr;
call->what.bfn.oname.len = bfn->name.len;
call->what.bfn.arg.min = bfn->arg.min;
call->what.bfn.arg.max = bfn->arg.max;
call->what.bfn.arg.spec = bfn->arg.spec;
call->what.bfn.handler = bfn->handler;
call->args = head;
call->nargs = nargs;
}
else
{
call->type = ASE_AWK_NDE_AFN;
call->line = line;
call->next = ASE_NULL;
call->what.afn.name.ptr = name;
call->what.afn.name.len = name_len;
call->args = head;
call->nargs = nargs;
}
return (ase_awk_nde_t*)call;
}
static ase_awk_nde_t* parse_if (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* test;
ase_awk_nde_t* then_part;
ase_awk_nde_t* else_part;
ase_awk_nde_if_t* nde;
if (!MATCH(awk,TOKEN_LPAREN))
{
SETERRTOK (awk, ASE_AWK_ELPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1) return ASE_NULL;
test = parse_expression (awk, awk->token.line);
if (test == ASE_NULL) return ASE_NULL;
if (!MATCH(awk,TOKEN_RPAREN))
{
ase_awk_clrpt (awk, test);
SETERRTOK (awk, ASE_AWK_ERPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
then_part = parse_statement (awk, awk->token.line);
if (then_part == ASE_NULL)
{
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
if (MATCH(awk,TOKEN_ELSE))
{
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, then_part);
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
else_part = parse_statement (awk, awk->token.prev.line);
if (else_part == ASE_NULL)
{
ase_awk_clrpt (awk, then_part);
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
}
else else_part = ASE_NULL;
nde = (ase_awk_nde_if_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_if_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, else_part);
ase_awk_clrpt (awk, then_part);
ase_awk_clrpt (awk, test);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_IF;
nde->line = line;
nde->next = ASE_NULL;
nde->test = test;
nde->then_part = then_part;
nde->else_part = else_part;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_while (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* test, * body;
ase_awk_nde_while_t* nde;
if (!MATCH(awk,TOKEN_LPAREN))
{
SETERRTOK (awk, ASE_AWK_ELPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1) return ASE_NULL;
test = parse_expression (awk, awk->token.line);
if (test == ASE_NULL) return ASE_NULL;
if (!MATCH(awk,TOKEN_RPAREN))
{
ase_awk_clrpt (awk, test);
SETERRTOK (awk, ASE_AWK_ERPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
body = parse_statement (awk, awk->token.line);
if (body == ASE_NULL)
{
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
nde = (ase_awk_nde_while_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_while_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, body);
ase_awk_clrpt (awk, test);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_WHILE;
nde->line = line;
nde->next = ASE_NULL;
nde->test = test;
nde->body = body;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_for (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* init, * test, * incr, * body;
ase_awk_nde_for_t* nde;
ase_awk_nde_foreach_t* nde2;
if (!MATCH(awk,TOKEN_LPAREN))
{
SETERRTOK (awk, ASE_AWK_ELPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1) return ASE_NULL;
if (MATCH(awk,TOKEN_SEMICOLON)) init = ASE_NULL;
else
{
/* this line is very ugly. it checks the entire next
* expression or the first element in the expression
* is wrapped by a parenthesis */
int no_foreach = MATCH(awk,TOKEN_LPAREN);
init = parse_expression (awk, awk->token.line);
if (init == ASE_NULL) return ASE_NULL;
if (!no_foreach && init->type == ASE_AWK_NDE_EXP_BIN &&
((ase_awk_nde_exp_t*)init)->opcode == ASE_AWK_BINOP_IN &&
is_plain_var(((ase_awk_nde_exp_t*)init)->left))
{
/* switch to foreach */
if (!MATCH(awk,TOKEN_RPAREN))
{
ase_awk_clrpt (awk, init);
SETERRTOK (awk, ASE_AWK_ERPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, init);
return ASE_NULL;
}
body = parse_statement (awk, awk->token.line);
if (body == ASE_NULL)
{
ase_awk_clrpt (awk, init);
return ASE_NULL;
}
nde2 = (ase_awk_nde_foreach_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_foreach_t));
if (nde2 == ASE_NULL)
{
ase_awk_clrpt (awk, init);
ase_awk_clrpt (awk, body);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde2->type = ASE_AWK_NDE_FOREACH;
nde2->line = line;
nde2->next = ASE_NULL;
nde2->test = init;
nde2->body = body;
return (ase_awk_nde_t*)nde2;
}
if (!MATCH(awk,TOKEN_SEMICOLON))
{
ase_awk_clrpt (awk, init);
SETERRTOK (awk, ASE_AWK_ESCOLON);
return ASE_NULL;
}
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, init);
return ASE_NULL;
}
if (MATCH(awk,TOKEN_SEMICOLON)) test = ASE_NULL;
else
{
test = parse_expression (awk, awk->token.line);
if (test == ASE_NULL)
{
ase_awk_clrpt (awk, init);
return ASE_NULL;
}
if (!MATCH(awk,TOKEN_SEMICOLON))
{
ase_awk_clrpt (awk, init);
ase_awk_clrpt (awk, test);
SETERRTOK (awk, ASE_AWK_ESCOLON);
return ASE_NULL;
}
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, init);
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
if (MATCH(awk,TOKEN_RPAREN)) incr = ASE_NULL;
else
{
incr = parse_expression (awk, awk->token.line);
if (incr == ASE_NULL)
{
ase_awk_clrpt (awk, init);
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
if (!MATCH(awk,TOKEN_RPAREN))
{
ase_awk_clrpt (awk, init);
ase_awk_clrpt (awk, test);
ase_awk_clrpt (awk, incr);
SETERRTOK (awk, ASE_AWK_ERPAREN);
return ASE_NULL;
}
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, init);
ase_awk_clrpt (awk, test);
ase_awk_clrpt (awk, incr);
return ASE_NULL;
}
body = parse_statement (awk, awk->token.line);
if (body == ASE_NULL)
{
ase_awk_clrpt (awk, init);
ase_awk_clrpt (awk, test);
ase_awk_clrpt (awk, incr);
return ASE_NULL;
}
nde = (ase_awk_nde_for_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_for_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, init);
ase_awk_clrpt (awk, test);
ase_awk_clrpt (awk, incr);
ase_awk_clrpt (awk, body);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_FOR;
nde->line = line;
nde->next = ASE_NULL;
nde->init = init;
nde->test = test;
nde->incr = incr;
nde->body = body;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_dowhile (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_t* test, * body;
ase_awk_nde_while_t* nde;
ASE_ASSERT (awk->token.prev.type == TOKEN_DO);
body = parse_statement (awk, awk->token.line);
if (body == ASE_NULL) return ASE_NULL;
if (!MATCH(awk,TOKEN_WHILE))
{
ase_awk_clrpt (awk, body);
SETERRTOK (awk, ASE_AWK_EWHILE);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, body);
return ASE_NULL;
}
if (!MATCH(awk,TOKEN_LPAREN))
{
ase_awk_clrpt (awk, body);
SETERRTOK (awk, ASE_AWK_ELPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, body);
return ASE_NULL;
}
test = parse_expression (awk, awk->token.line);
if (test == ASE_NULL)
{
ase_awk_clrpt (awk, body);
return ASE_NULL;
}
if (!MATCH(awk,TOKEN_RPAREN))
{
ase_awk_clrpt (awk, body);
ase_awk_clrpt (awk, test);
SETERRTOK (awk, ASE_AWK_ERPAREN);
return ASE_NULL;
}
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, body);
ase_awk_clrpt (awk, test);
return ASE_NULL;
}
nde = (ase_awk_nde_while_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_while_t));
if (nde == ASE_NULL)
{
ase_awk_clrpt (awk, body);
ase_awk_clrpt (awk, test);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_DOWHILE;
nde->line = line;
nde->next = ASE_NULL;
nde->test = test;
nde->body = body;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_break (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_break_t* nde;
ASE_ASSERT (awk->token.prev.type == TOKEN_BREAK);
if (awk->parse.depth.cur.loop <= 0)
{
SETERRLIN (awk, ASE_AWK_EBREAK, line);
return ASE_NULL;
}
nde = (ase_awk_nde_break_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_break_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_BREAK;
nde->line = line;
nde->next = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_continue (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_continue_t* nde;
ASE_ASSERT (awk->token.prev.type == TOKEN_CONTINUE);
if (awk->parse.depth.cur.loop <= 0)
{
SETERRLIN (awk, ASE_AWK_ECONTINUE, line);
return ASE_NULL;
}
nde = (ase_awk_nde_continue_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_continue_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_CONTINUE;
nde->line = line;
nde->next = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_return (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_return_t* nde;
ase_awk_nde_t* val;
ASE_ASSERT (awk->token.prev.type == TOKEN_RETURN);
nde = (ase_awk_nde_return_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_return_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_RETURN;
nde->line = line;
nde->next = ASE_NULL;
if (MATCH(awk,TOKEN_SEMICOLON))
{
/* no return value */
val = ASE_NULL;
}
else
{
val = parse_expression (awk, awk->token.line);
if (val == ASE_NULL)
{
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
}
nde->val = val;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_exit (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_exit_t* nde;
ase_awk_nde_t* val;
ASE_ASSERT (awk->token.prev.type == TOKEN_EXIT);
nde = (ase_awk_nde_exit_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_exit_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_EXIT;
nde->line = line;
nde->next = ASE_NULL;
if (MATCH(awk,TOKEN_SEMICOLON))
{
/* no exit code */
val = ASE_NULL;
}
else
{
val = parse_expression (awk, awk->token.line);
if (val == ASE_NULL)
{
ASE_AWK_FREE (awk, nde);
return ASE_NULL;
}
}
nde->val = val;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_next (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_next_t* nde;
ASE_ASSERT (awk->token.prev.type == TOKEN_NEXT);
if (awk->parse.id.block == PARSE_BEGIN_BLOCK)
{
SETERRLIN (awk, ASE_AWK_ENEXTBEG, line);
return ASE_NULL;
}
if (awk->parse.id.block == PARSE_END_BLOCK)
{
SETERRLIN (awk, ASE_AWK_ENEXTEND, line);
return ASE_NULL;
}
nde = (ase_awk_nde_next_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_next_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_NEXT;
nde->line = line;
nde->next = ASE_NULL;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_nextfile (ase_awk_t* awk, ase_size_t line, int out)
{
ase_awk_nde_nextfile_t* nde;
if (!out && awk->parse.id.block == PARSE_BEGIN_BLOCK)
{
SETERRLIN (awk, ASE_AWK_ENEXTFBEG, line);
return ASE_NULL;
}
if (!out && awk->parse.id.block == PARSE_END_BLOCK)
{
SETERRLIN (awk, ASE_AWK_ENEXTFEND, line);
return ASE_NULL;
}
nde = (ase_awk_nde_nextfile_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_nextfile_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_NEXTFILE;
nde->line = line;
nde->next = ASE_NULL;
nde->out = out;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_delete (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_delete_t* nde;
ase_awk_nde_t* var;
ASE_ASSERT (awk->token.prev.type == TOKEN_DELETE);
if (!MATCH(awk,TOKEN_IDENT))
{
SETERRTOK (awk, ASE_AWK_EIDENT);
return ASE_NULL;
}
var = parse_primary_ident (awk, awk->token.line);
if (var == ASE_NULL) return ASE_NULL;
if (!is_var (var))
{
/* a normal identifier is expected */
ase_awk_clrpt (awk, var);
SETERRLIN (awk, ASE_AWK_EDELETE, line);
return ASE_NULL;
}
nde = (ase_awk_nde_delete_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_delete_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_DELETE;
nde->line = line;
nde->next = ASE_NULL;
nde->var = var;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_reset (ase_awk_t* awk, ase_size_t line)
{
ase_awk_nde_reset_t* nde;
ase_awk_nde_t* var;
ASE_ASSERT (awk->token.prev.type == TOKEN_RESET);
if (!MATCH(awk,TOKEN_IDENT))
{
SETERRTOK (awk, ASE_AWK_EIDENT);
return ASE_NULL;
}
var = parse_primary_ident (awk, awk->token.line);
if (var == ASE_NULL) return ASE_NULL;
/* unlike delete, it must be followed by a plain variable only */
if (!is_plain_var (var))
{
/* a normal identifier is expected */
ase_awk_clrpt (awk, var);
SETERRLIN (awk, ASE_AWK_ERESET, line);
return ASE_NULL;
}
nde = (ase_awk_nde_reset_t*) ASE_AWK_MALLOC (
awk, ASE_SIZEOF(ase_awk_nde_reset_t));
if (nde == ASE_NULL)
{
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
nde->type = ASE_AWK_NDE_RESET;
nde->line = line;
nde->next = ASE_NULL;
nde->var = var;
return (ase_awk_nde_t*)nde;
}
static ase_awk_nde_t* parse_print (ase_awk_t* awk, ase_size_t line, int type)
{
ase_awk_nde_print_t* nde;
ase_awk_nde_t* args = ASE_NULL;
ase_awk_nde_t* out = ASE_NULL;
int out_type;
if (!MATCH(awk,TOKEN_SEMICOLON) &&
!MATCH(awk,TOKEN_GT) &&
!MATCH(awk,TOKEN_RSHIFT) &&
!MATCH(awk,TOKEN_BOR) &&
!MATCH(awk,TOKEN_BORAND))
{
ase_awk_nde_t* args_tail;
ase_awk_nde_t* tail_prev;
args = parse_expression (awk, awk->token.line);
if (args == ASE_NULL) return ASE_NULL;
args_tail = args;
tail_prev = ASE_NULL;
if (args->type != ASE_AWK_NDE_GRP)
{
/* args->type == ASE_AWK_NDE_GRP when print (a, b, c)
* args->type != ASE_AWK_NDE_GRP when print a, b, c */
while (MATCH(awk,TOKEN_COMMA))
{
if (get_token(awk) == -1)
{
ase_awk_clrpt (awk, args);
return ASE_NULL;
}
args_tail->next = parse_expression (awk, awk->token.line);
if (args_tail->next == ASE_NULL)
{
ase_awk_clrpt (awk, args);
return ASE_NULL;
}
tail_prev = args_tail;
args_tail = args_tail->next;
}
}
/* print 1 > 2 would print 1 to the file named 2.
* print (1 > 2) would print (1 > 2) on the console */
if (awk->token.prev.type != TOKEN_RPAREN &&
args_tail->type == ASE_AWK_NDE_EXP_BIN)
{
ase_awk_nde_exp_t* ep = (ase_awk_nde_exp_t*)args_tail;
if (ep->opcode == ASE_AWK_BINOP_GT)
{
ase_awk_nde_t* tmp = args_tail;
if (tail_prev != ASE_NULL)
tail_prev->next = ep->left;
else args = ep->left;
out = ep->right;
out_type = ASE_AWK_OUT_FILE;
ASE_AWK_FREE (awk, tmp);
}
else if (ep->opcode == ASE_AWK_BINOP_RSHIFT)
{
ase_awk_nde_t* tmp = args_tail;
if (tail_prev != ASE_NULL)
tail_prev->next = ep->left;
else args = ep->left;
out = ep->right;
out_type = ASE_AWK_OUT_FILE_APPEND;
ASE_AWK_FREE (awk, tmp);
}
else if (ep->opcode == ASE_AWK_BINOP_BOR)
{
ase_awk_nde_t* tmp = args_tail;
if (tail_prev != ASE_NULL)
tail_prev->next = ep->left;
else args = ep->left;
out = ep->right;
out_type = ASE_AWK_OUT_PIPE;
ASE_AWK_FREE (awk, tmp);
}
}
}
if (out == ASE_NULL)
{
out_type = MATCH(awk,TOKEN_GT)? ASE_AWK_OUT_FILE:
MATCH(awk,TOKEN_RSHIFT)? ASE_AWK_OUT_FILE_APPEND:
MATCH(awk,TOKEN_BOR)? ASE_AWK_OUT_PIPE:
MATCH(awk,TOKEN_BORAND)? ASE_AWK_OUT_COPROC:
ASE_AWK_OUT_CONSOLE;
if (out_type != ASE_AWK_OUT_CONSOLE)
{
if (get_token(awk) == -1)
{
if (args != ASE_NULL) ase_awk_clrpt (awk, args);
return ASE_NULL;
}
out = parse_expression (awk, awk->token.line);
if (out == ASE_NULL)
{
if (args != ASE_NULL) ase_awk_clrpt (awk, args);
return ASE_NULL;
}
}
}
nde = (ase_awk_nde_print_t*)
ASE_AWK_MALLOC (awk, ASE_SIZEOF(ase_awk_nde_print_t));
if (nde == ASE_NULL)
{
if (args != ASE_NULL) ase_awk_clrpt (awk, args);
if (out != ASE_NULL) ase_awk_clrpt (awk, out);
SETERRLIN (awk, ASE_AWK_ENOMEM, line);
return ASE_NULL;
}
ASE_ASSERTX (
type == ASE_AWK_NDE_PRINT || type == ASE_AWK_NDE_PRINTF,
"the node type should be either ASE_AWK_NDE_PRINT or ASE_AWK_NDE_PRINTF");
if (type == ASE_AWK_NDE_PRINTF && args == ASE_NULL)
{
if (out != ASE_NULL) ase_awk_clrpt (awk, out);
SETERRLIN (awk, ASE_AWK_EPRINTFARG, line);
return ASE_NULL;
}
nde->type = type;
nde->line = line;
nde->next = ASE_NULL;
nde->args = args;
nde->out_type = out_type;
nde->out = out;
return (ase_awk_nde_t*)nde;
}
static int get_token (ase_awk_t* awk)
{
ase_cint_t c;
ase_size_t line;
int n;
line = awk->token.line;
awk->token.prev.type = awk->token.type;
awk->token.prev.line = awk->token.line;
awk->token.prev.column = awk->token.column;
do
{
if (skip_spaces(awk) == -1) return -1;
if ((n = skip_comment(awk)) == -1) return -1;
}
while (n == 1);
ase_str_clear (&awk->token.name);
awk->token.line = awk->src.lex.line;
awk->token.column = awk->src.lex.column;
if (line != 0 && (awk->option & ASE_AWK_BLOCKLESS) &&
(awk->parse.id.block == PARSE_PATTERN ||
awk->parse.id.block == PARSE_BEGIN ||
awk->parse.id.block == PARSE_END))
{
if (awk->token.line != line)
{
SET_TOKEN_TYPE (awk, TOKEN_NEWLINE);
return 0;
}
}
c = awk->src.lex.curc;
if (c == ASE_CHAR_EOF)
{
ADD_TOKEN_CHAR (awk, ASE_T('<'));
ADD_TOKEN_CHAR (awk, ASE_T('E'));
ADD_TOKEN_CHAR (awk, ASE_T('O'));
ADD_TOKEN_CHAR (awk, ASE_T('F'));
ADD_TOKEN_CHAR (awk, ASE_T('>'));
SET_TOKEN_TYPE (awk, TOKEN_EOF);
}
else if (ASE_AWK_ISDIGIT (awk, c)/*|| c == ASE_T('.')*/)
{
if (get_number (awk) == -1) return -1;
}
else if (c == ASE_T('.'))
{
if (get_char (awk) == -1) return -1;
c = awk->src.lex.curc;
if ((awk->option & ASE_AWK_EXPLICIT) == 0 &&
ASE_AWK_ISDIGIT (awk, c))
{
awk->src.lex.curc = ASE_T('.');
if (unget_char (awk, c) == -1) return -1;
if (get_number (awk) == -1) return -1;
}
else /*if (ASE_AWK_ISSPACE (awk, c) || c == ASE_CHAR_EOF)*/
{
SET_TOKEN_TYPE (awk, TOKEN_PERIOD);
ADD_TOKEN_CHAR (awk, ASE_T('.'));
}
/*
else
{
ase_awk_seterror (
awk, ASE_AWK_ELXCHR, awk->token.line,
ASE_T("floating point not followed by any valid digits"));
return -1;
}
*/
}
else if (ASE_AWK_ISALPHA (awk, c) || c == ASE_T('_'))
{
int type;
/* identifier */
do
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
}
while (ASE_AWK_ISALPHA (awk, c) ||
c == ASE_T('_') || ASE_AWK_ISDIGIT(awk,c));
type = classify_ident (awk,
ASE_STR_BUF(&awk->token.name),
ASE_STR_LEN(&awk->token.name));
SET_TOKEN_TYPE (awk, type);
}
else if (c == ASE_T('\"'))
{
SET_TOKEN_TYPE (awk, TOKEN_STR);
if (get_charstr(awk) == -1) return -1;
while (awk->option & ASE_AWK_STRCONCAT)
{
do
{
if (skip_spaces(awk) == -1) return -1;
if ((n = skip_comment(awk)) == -1) return -1;
}
while (n == 1);
c = awk->src.lex.curc;
if (c != ASE_T('\"')) break;
if (get_charstr(awk) == -1) return -1;
}
}
else if (c == ASE_T('='))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_EQ);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_ASSIGN);
}
}
else if (c == ASE_T('!'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_NE);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('~'))
{
SET_TOKEN_TYPE (awk, TOKEN_NM);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_NOT);
}
}
else if (c == ASE_T('>'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if ((awk->option & ASE_AWK_SHIFT) && c == ASE_T('>'))
{
SET_TOKEN_TYPE (awk, TOKEN_RSHIFT);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_GE);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_GT);
}
}
else if (c == ASE_T('<'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if ((awk->option & ASE_AWK_SHIFT) && c == ASE_T('<'))
{
SET_TOKEN_TYPE (awk, TOKEN_LSHIFT);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_LE);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_LT);
}
}
else if (c == ASE_T('|'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('|'))
{
SET_TOKEN_TYPE (awk, TOKEN_LOR);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if ((awk->option & ASE_AWK_COPROC) && c == ASE_T('&'))
{
SET_TOKEN_TYPE (awk, TOKEN_BORAND);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_BOR);
}
}
else if (c == ASE_T('&'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('&'))
{
SET_TOKEN_TYPE (awk, TOKEN_LAND);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_BAND);
}
}
else if (c == ASE_T('~'))
{
SET_TOKEN_TYPE (awk, TOKEN_TILDE);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('^'))
{
SET_TOKEN_TYPE (awk, TOKEN_BXOR);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('+'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('+'))
{
SET_TOKEN_TYPE (awk, TOKEN_PLUSPLUS);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_PLUS_ASSIGN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_PLUS);
}
}
else if (c == ASE_T('-'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('-'))
{
SET_TOKEN_TYPE (awk, TOKEN_MINUSMINUS);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_MINUS_ASSIGN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_MINUS);
}
}
else if (c == ASE_T('*'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_MUL_ASSIGN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('*'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_EXP_ASSIGN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_EXP);
}
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_MUL);
}
}
else if (c == ASE_T('/'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_DIV_ASSIGN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if ((awk->option & ASE_AWK_IDIV) && c == ASE_T('/'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_IDIV_ASSIGN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_IDIV);
}
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_DIV);
}
}
#if 0
/* TODO: is it a good idea to use a back-slash for
* the idiv operator like BASIC? */
else if ((awk->option & ASE_AWK_IDIV) && c == ASE_T('\\'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_IDIV_ASSIGN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_IDIV);
}
}
#endif
else if (c == ASE_T('%'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('='))
{
SET_TOKEN_TYPE (awk, TOKEN_MOD_ASSIGN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
SET_TOKEN_TYPE (awk, TOKEN_MOD);
}
}
else if (c == ASE_T('('))
{
SET_TOKEN_TYPE (awk, TOKEN_LPAREN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T(')'))
{
SET_TOKEN_TYPE (awk, TOKEN_RPAREN);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('{'))
{
SET_TOKEN_TYPE (awk, TOKEN_LBRACE);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('}'))
{
SET_TOKEN_TYPE (awk, TOKEN_RBRACE);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('['))
{
SET_TOKEN_TYPE (awk, TOKEN_LBRACK);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T(']'))
{
SET_TOKEN_TYPE (awk, TOKEN_RBRACK);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('$'))
{
SET_TOKEN_TYPE (awk, TOKEN_DOLLAR);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T(','))
{
SET_TOKEN_TYPE (awk, TOKEN_COMMA);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T(';'))
{
SET_TOKEN_TYPE (awk, TOKEN_SEMICOLON);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T(':'))
{
SET_TOKEN_TYPE (awk, TOKEN_COLON);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else if (c == ASE_T('?'))
{
SET_TOKEN_TYPE (awk, TOKEN_QUEST);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR (awk);
}
else
{
ase_char_t cc = (ase_char_t)c;
SETERRARG (awk, ASE_AWK_ELXCHR, awk->token.line, &cc, 1);
return -1;
}
return 0;
}
static int get_number (ase_awk_t* awk)
{
ase_cint_t c;
ASE_ASSERT (ASE_STR_LEN(&awk->token.name) == 0);
SET_TOKEN_TYPE (awk, TOKEN_INT);
c = awk->src.lex.curc;
if (c == ASE_T('0'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('x') || c == ASE_T('X'))
{
/* hexadecimal number */
do
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
}
while (ASE_AWK_ISXDIGIT (awk, c));
return 0;
}
#if 0
else if (c == ASE_T('b') || c == ASE_T('B'))
{
/* binary number */
do
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
}
while (c == ASE_T('0') || c == ASE_T('1'));
return 0;
}
#endif
else if (c != '.')
{
/* octal number */
while (c >= ASE_T('0') && c <= ASE_T('7'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
}
if (c == ASE_T('8') || c == ASE_T('9'))
{
ase_char_t cc = (ase_char_t)c;
SETERRARG (awk, ASE_AWK_ELXDIG, awk->token.line, &cc, 1);
return -1;
}
return 0;
}
}
while (ASE_AWK_ISDIGIT (awk, c))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
}
if (c == ASE_T('.'))
{
/* floating-point number */
SET_TOKEN_TYPE (awk, TOKEN_REAL);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
while (ASE_AWK_ISDIGIT (awk, c))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
}
}
if (c == ASE_T('E') || c == ASE_T('e'))
{
SET_TOKEN_TYPE (awk, TOKEN_REAL);
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
if (c == ASE_T('+') || c == ASE_T('-'))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
}
while (ASE_AWK_ISDIGIT (awk, c))
{
ADD_TOKEN_CHAR (awk, c);
GET_CHAR_TO (awk, c);
}
}
return 0;
}
static int get_charstr (ase_awk_t* awk)
{
if (awk->src.lex.curc != ASE_T('\"'))
{
/* the starting quote has been consumed before this function
* has been called */
ADD_TOKEN_CHAR (awk, awk->src.lex.curc);
}
return get_string (awk, ASE_T('\"'), ASE_T('\\'), ase_false);
}
static int get_rexstr (ase_awk_t* awk)
{
if (awk->src.lex.curc == ASE_T('/'))
{
/* this part of the function is different from get_charstr
* because of the way this function is called */
GET_CHAR (awk);
return 0;
}
else
{
ADD_TOKEN_CHAR (awk, awk->src.lex.curc);
return get_string (awk, ASE_T('/'), ASE_T('\\'), ase_true);
}
}
static int get_string (
ase_awk_t* awk, ase_char_t end_char,
ase_char_t esc_char, ase_bool_t keep_esc_char)
{
ase_cint_t c;
int escaped = 0;
int digit_count = 0;
ase_cint_t c_acc = 0;
while (1)
{
GET_CHAR_TO (awk, c);
if (c == ASE_CHAR_EOF)
{
SETERRTOK (awk, ASE_AWK_EENDSTR);
return -1;
}
if (escaped == 3)
{
if (c >= ASE_T('0') && c <= ASE_T('7'))
{
c_acc = c_acc * 8 + c - ASE_T('0');
digit_count++;
if (digit_count >= escaped)
{
ADD_TOKEN_CHAR (awk, c_acc);
escaped = 0;
}
continue;
}
else
{
ADD_TOKEN_CHAR (awk, c_acc);
escaped = 0;
}
}
else if (escaped == 2 || escaped == 4 || escaped == 8)
{
if (c >= ASE_T('0') && c <= ASE_T('9'))
{
c_acc = c_acc * 16 + c - ASE_T('0');
digit_count++;
if (digit_count >= escaped)
{
ADD_TOKEN_CHAR (awk, c_acc);
escaped = 0;
}
continue;
}
else if (c >= ASE_T('A') && c <= ASE_T('F'))
{
c_acc = c_acc * 16 + c - ASE_T('A') + 10;
digit_count++;
if (digit_count >= escaped)
{
ADD_TOKEN_CHAR (awk, c_acc);
escaped = 0;
}
continue;
}
else if (c >= ASE_T('a') && c <= ASE_T('f'))
{
c_acc = c_acc * 16 + c - ASE_T('a') + 10;
digit_count++;
if (digit_count >= escaped)
{
ADD_TOKEN_CHAR (awk, c_acc);
escaped = 0;
}
continue;
}
else
{
ase_char_t rc;
rc = (escaped == 2)? ASE_T('x'):
(escaped == 4)? ASE_T('u'): ASE_T('U');
if (digit_count == 0) ADD_TOKEN_CHAR (awk, rc);
else ADD_TOKEN_CHAR (awk, c_acc);
escaped = 0;
}
}
if (escaped == 0 && c == end_char)
{
/* terminating quote */
/*GET_CHAR_TO (awk, c);*/
GET_CHAR (awk);
break;
}
if (escaped == 0 && c == esc_char)
{
escaped = 1;
continue;
}
if (escaped == 1)
{
if (c == ASE_T('n')) c = ASE_T('\n');
else if (c == ASE_T('r')) c = ASE_T('\r');
else if (c == ASE_T('t')) c = ASE_T('\t');
else if (c == ASE_T('f')) c = ASE_T('\f');
else if (c == ASE_T('b')) c = ASE_T('\b');
else if (c == ASE_T('v')) c = ASE_T('\v');
else if (c == ASE_T('a')) c = ASE_T('\a');
else if (c >= ASE_T('0') && c <= ASE_T('7'))
{
escaped = 3;
digit_count = 1;
c_acc = c - ASE_T('0');
continue;
}
else if (c == ASE_T('x'))
{
escaped = 2;
digit_count = 0;
c_acc = 0;
continue;
}
#ifdef ASE_CHAR_IS_WCHAR
else if (c == ASE_T('u') && ASE_SIZEOF(ase_char_t) >= 2)
{
escaped = 4;
digit_count = 0;
c_acc = 0;
continue;
}
else if (c == ASE_T('U') && ASE_SIZEOF(ase_char_t) >= 4)
{
escaped = 8;
digit_count = 0;
c_acc = 0;
continue;
}
#endif
else if (keep_esc_char)
{
ADD_TOKEN_CHAR (awk, esc_char);
}
escaped = 0;
}
ADD_TOKEN_CHAR (awk, c);
}
return 0;
}
static int get_char (ase_awk_t* awk)
{
ase_ssize_t n;
/*ase_char_t c;*/
if (awk->src.lex.ungotc_count > 0)
{
awk->src.lex.curc = awk->src.lex.ungotc[--awk->src.lex.ungotc_count];
awk->src.lex.line = awk->src.lex.ungotc_line[awk->src.lex.ungotc_count];
awk->src.lex.column = awk->src.lex.ungotc_column[awk->src.lex.ungotc_count];
return 0;
}
if (awk->src.shared.buf_pos >= awk->src.shared.buf_len)
{
CLRERR (awk);
n = awk->src.ios.in (
ASE_AWK_IO_READ, awk->src.ios.custom_data,
awk->src.shared.buf, ASE_COUNTOF(awk->src.shared.buf));
if (n <= -1)
{
if (ISNOERR(awk)) SETERR (awk, ASE_AWK_ESINRD);
return -1;
}
if (n == 0)
{
awk->src.lex.curc = ASE_CHAR_EOF;
return 0;
}
awk->src.shared.buf_pos = 0;
awk->src.shared.buf_len = n;
}
awk->src.lex.curc = awk->src.shared.buf[awk->src.shared.buf_pos++];
if (awk->src.lex.curc == ASE_T('\n'))
{
awk->src.lex.line++;
awk->src.lex.column = 1;
}
else awk->src.lex.column++;
return 0;
}
static int unget_char (ase_awk_t* awk, ase_cint_t c)
{
if (awk->src.lex.ungotc_count >= ASE_COUNTOF(awk->src.lex.ungotc))
{
SETERRLIN (awk, ASE_AWK_ELXUNG, awk->src.lex.line);
return -1;
}
awk->src.lex.ungotc_line[awk->src.lex.ungotc_count] = awk->src.lex.line;
awk->src.lex.ungotc_column[awk->src.lex.ungotc_count] = awk->src.lex.column;
awk->src.lex.ungotc[awk->src.lex.ungotc_count++] = c;
return 0;
}
static int skip_spaces (ase_awk_t* awk)
{
ase_cint_t c = awk->src.lex.curc;
while (ASE_AWK_ISSPACE (awk, c)) GET_CHAR_TO (awk, c);
return 0;
}
static int skip_comment (ase_awk_t* awk)
{
ase_cint_t c = awk->src.lex.curc;
ase_size_t line, column;
if (c == ASE_T('#'))
{
do
{
GET_CHAR_TO (awk, c);
}
while (c != ASE_T('\n') && c != ASE_CHAR_EOF);
GET_CHAR (awk);
return 1; /* comment by # */
}
if (c != ASE_T('/')) return 0; /* not a comment */
line = awk->src.lex.line;
column = awk->src.lex.column;
GET_CHAR_TO (awk, c);
if (c == ASE_T('*'))
{
do
{
GET_CHAR_TO (awk, c);
if (c == ASE_CHAR_EOF)
{
SETERRLIN (awk, ASE_AWK_EENDCMT, awk->src.lex.line);
return -1;
}
if (c == ASE_T('*'))
{
GET_CHAR_TO (awk, c);
if (c == ASE_CHAR_EOF)
{
SETERRLIN (awk, ASE_AWK_EENDCMT, awk->src.lex.line);
return -1;
}
if (c == ASE_T('/'))
{
/*GET_CHAR_TO (awk, c);*/
GET_CHAR (awk);
break;
}
}
}
while (1);
return 1; /* c-style comment */
}
if (unget_char (awk, c) == -1) return -1; /* error */
awk->src.lex.curc = ASE_T('/');
awk->src.lex.line = line;
awk->src.lex.column = column;
return 0;
}
static int classify_ident (
ase_awk_t* awk, const ase_char_t* name, ase_size_t len)
{
kwent_t* kwp;
ase_awk_pair_t* pair;
for (kwp = kwtab; kwp->name != ASE_NULL; kwp++)
{
const ase_char_t* k;
ase_size_t l;
if (kwp->valid != 0 &&
(awk->option & kwp->valid) != kwp->valid) continue;
pair = ase_awk_map_get (awk->wtab, kwp->name, kwp->name_len);
if (pair != ASE_NULL)
{
k = ((ase_cstr_t*)(pair->val))->ptr;
l = ((ase_cstr_t*)(pair->val))->len;
}
else
{
k = kwp->name;
l = kwp->name_len;
}
if (ase_strxncmp (k, l, name, len) == 0)
{
return kwp->type;
}
}
return TOKEN_IDENT;
}
static int assign_to_opcode (ase_awk_t* awk)
{
static int assop[] =
{
ASE_AWK_ASSOP_NONE,
ASE_AWK_ASSOP_PLUS,
ASE_AWK_ASSOP_MINUS,
ASE_AWK_ASSOP_MUL,
ASE_AWK_ASSOP_DIV,
ASE_AWK_ASSOP_IDIV,
ASE_AWK_ASSOP_MOD,
ASE_AWK_ASSOP_EXP
};
if (awk->token.type >= TOKEN_ASSIGN &&
awk->token.type <= TOKEN_EXP_ASSIGN)
{
return assop[awk->token.type - TOKEN_ASSIGN];
}
return -1;
}
static int is_plain_var (ase_awk_nde_t* nde)
{
return nde->type == ASE_AWK_NDE_GLOBAL ||
nde->type == ASE_AWK_NDE_LOCAL ||
nde->type == ASE_AWK_NDE_ARG ||
nde->type == ASE_AWK_NDE_NAMED;
}
static int is_var (ase_awk_nde_t* nde)
{
return nde->type == ASE_AWK_NDE_GLOBAL ||
nde->type == ASE_AWK_NDE_LOCAL ||
nde->type == ASE_AWK_NDE_ARG ||
nde->type == ASE_AWK_NDE_NAMED ||
nde->type == ASE_AWK_NDE_GLOBALIDX ||
nde->type == ASE_AWK_NDE_LOCALIDX ||
nde->type == ASE_AWK_NDE_ARGIDX ||
nde->type == ASE_AWK_NDE_NAMEDIDX;
}
struct deparse_func_t
{
ase_awk_t* awk;
ase_char_t* tmp;
ase_size_t tmp_len;
};
static int deparse (ase_awk_t* awk)
{
ase_awk_nde_t* nde;
ase_awk_chain_t* chain;
ase_char_t tmp[ASE_SIZEOF(ase_size_t)*8 + 32];
struct deparse_func_t df;
int n = 0;
ase_ssize_t op;
ASE_ASSERT (awk->src.ios.out != ASE_NULL);
awk->src.shared.buf_len = 0;
awk->src.shared.buf_pos = 0;
CLRERR (awk);
op = awk->src.ios.out (
ASE_AWK_IO_OPEN, awk->src.ios.custom_data, ASE_NULL, 0);
if (op <= -1)
{
if (ISNOERR(awk)) SETERR (awk, ASE_AWK_ESOUTOP);
return -1;
}
if (op == 0)
{
/* the result of the open operation indicates that the
* file has been open but reached the end. so it has to
* skip the entire deparsing procedure as it can't write
* any single characters on such an io handler. but note
* that this is not really an error for the parse and deparser.
*
* in fact, there are two ways to skip deparsing.
* 1. set awk->src.ios.out to NULL.
* 2. set awk->src.ios.out to a normal handler but
* make it return 0 on the OPEN request.
*/
n = 0;
goto exit_deparse;
}
#define EXIT_DEPARSE() do { n = -1; goto exit_deparse; } while(0)
if (awk->tree.nglobals > awk->tree.nbglobals)
{
ase_size_t i/*, len*/;
ASE_ASSERT (awk->tree.nglobals > 0);
if (ase_awk_putsrcstr(awk,ase_awk_getkw(awk,ASE_T("global"))) == -1)
{
EXIT_DEPARSE ();
}
if (ase_awk_putsrcstr (awk, ASE_T(" ")) == -1)
{
EXIT_DEPARSE ();
}
for (i = awk->tree.nbglobals; i < awk->tree.nglobals - 1; i++)
{
/*
len = ase_awk_longtostr ((ase_long_t)i,
10, ASE_T("__global"), tmp, ASE_COUNTOF(tmp));
ASE_ASSERT (len != (ase_size_t)-1);
if (ase_awk_putsrcstrx (awk, tmp, len) == -1)
{
EXIT_DEPARSE ();
}
*/
if (ase_awk_putsrcstrx (awk,
awk->parse.globals.buf[i].name.ptr,
awk->parse.globals.buf[i].name.len) == -1)
{
EXIT_DEPARSE ();
}
if (ase_awk_putsrcstr (awk, ASE_T(", ")) == -1)
EXIT_DEPARSE ();
}
/*
len = ase_awk_longtostr ((ase_long_t)i,
10, ASE_T("__global"), tmp, ASE_COUNTOF(tmp));
ASE_ASSERT (len != (ase_size_t)-1);
if (ase_awk_putsrcstrx (awk, tmp, len) == -1)
{
EXIT_DEPARSE ();
}
*/
if (ase_awk_putsrcstrx (awk,
awk->parse.globals.buf[i].name.ptr,
awk->parse.globals.buf[i].name.len) == -1)
{
EXIT_DEPARSE ();
}
if (awk->option & ASE_AWK_CRLF)
{
if (ase_awk_putsrcstr (awk, ASE_T(";\r\n\r\n")) == -1)
{
EXIT_DEPARSE ();
}
}
else
{
if (ase_awk_putsrcstr (awk, ASE_T(";\n\n")) == -1)
{
EXIT_DEPARSE ();
}
}
}
df.awk = awk;
df.tmp = tmp;
df.tmp_len = ASE_COUNTOF(tmp);
if (ase_awk_map_walk (awk->tree.afns, deparse_func, &df) == -1)
{
EXIT_DEPARSE ();
}
for (nde = awk->tree.begin; nde != ASE_NULL; nde = nde->next)
{
const ase_char_t* kw = ase_awk_getkw(awk,ASE_T("BEGIN"));
if (ase_awk_putsrcstr(awk,kw) == -1) EXIT_DEPARSE ();
if (ase_awk_putsrcstr (awk, ASE_T(" ")) == -1) EXIT_DEPARSE ();
if (ase_awk_prnnde (awk, nde) == -1) EXIT_DEPARSE ();
if (awk->option & ASE_AWK_CRLF)
{
if (put_char (awk, ASE_T('\r')) == -1) EXIT_DEPARSE ();
}
if (put_char (awk, ASE_T('\n')) == -1) EXIT_DEPARSE ();
}
chain = awk->tree.chain;
while (chain != ASE_NULL)
{
if (chain->pattern != ASE_NULL)
{
if (ase_awk_prnptnpt (awk, chain->pattern) == -1)
EXIT_DEPARSE ();
}
if (chain->action == ASE_NULL)
{
/* blockless pattern */
if (awk->option & ASE_AWK_CRLF)
{
if (put_char (awk, ASE_T('\r')) == -1)
EXIT_DEPARSE ();
}
if (put_char (awk, ASE_T('\n')) == -1)
EXIT_DEPARSE ();
}
else
{
if (chain->pattern != ASE_NULL)
{
if (put_char (awk, ASE_T(' ')) == -1)
EXIT_DEPARSE ();
}
if (ase_awk_prnpt (awk, chain->action) == -1)
EXIT_DEPARSE ();
}
if (awk->option & ASE_AWK_CRLF)
{
if (put_char (awk, ASE_T('\r')) == -1)
EXIT_DEPARSE ();
}
if (put_char (awk, ASE_T('\n')) == -1)
EXIT_DEPARSE ();
chain = chain->next;
}
for (nde = awk->tree.end; nde != ASE_NULL; nde = nde->next)
{
const ase_char_t* kw = ase_awk_getkw(awk,ASE_T("END"));
if (ase_awk_putsrcstr(awk,kw) == -1) EXIT_DEPARSE ();
if (ase_awk_putsrcstr (awk, ASE_T(" ")) == -1) EXIT_DEPARSE ();
if (ase_awk_prnnde (awk, nde) == -1) EXIT_DEPARSE ();
/*
if (awk->option & ASE_AWK_CRLF)
{
if (put_char (awk, ASE_T('\r')) == -1) EXIT_DEPARSE ();
}
if (put_char (awk, ASE_T('\n')) == -1) EXIT_DEPARSE ();
*/
}
if (flush_out (awk) == -1) EXIT_DEPARSE ();
exit_deparse:
if (n == 0) CLRERR (awk);
if (awk->src.ios.out (
ASE_AWK_IO_CLOSE, awk->src.ios.custom_data, ASE_NULL, 0) != 0)
{
if (n == 0)
{
if (ISNOERR(awk)) SETERR (awk, ASE_AWK_ESOUTCL);
n = -1;
}
}
return n;
}
static int deparse_func (ase_awk_pair_t* pair, void* arg)
{
struct deparse_func_t* df = (struct deparse_func_t*)arg;
ase_awk_afn_t* afn = (ase_awk_afn_t*)pair->val;
ase_size_t i, n;
ASE_ASSERT (ase_strxncmp (ASE_AWK_PAIR_KEYPTR(pair), ASE_AWK_PAIR_KEYLEN(pair), afn->name, afn->name_len) == 0);
if (ase_awk_putsrcstr(df->awk,ase_awk_getkw(df->awk,ASE_T("function"))) == -1) return -1;
if (ase_awk_putsrcstr (df->awk, ASE_T(" ")) == -1) return -1;
if (ase_awk_putsrcstr (df->awk, afn->name) == -1) return -1;
if (ase_awk_putsrcstr (df->awk, ASE_T(" (")) == -1) return -1;
for (i = 0; i < afn->nargs; )
{
n = ase_awk_longtostr (i++, 10,
ASE_T("__param"), df->tmp, df->tmp_len);
ASE_ASSERT (n != (ase_size_t)-1);
if (ase_awk_putsrcstrx (df->awk, df->tmp, n) == -1) return -1;
if (i >= afn->nargs) break;
if (ase_awk_putsrcstr (df->awk, ASE_T(", ")) == -1) return -1;
}
if (ase_awk_putsrcstr (df->awk, ASE_T(")")) == -1) return -1;
if (df->awk->option & ASE_AWK_CRLF)
{
if (put_char (df->awk, ASE_T('\r')) == -1) return -1;
}
if (put_char (df->awk, ASE_T('\n')) == -1) return -1;
if (ase_awk_prnpt (df->awk, afn->body) == -1) return -1;
if (df->awk->option & ASE_AWK_CRLF)
{
if (put_char (df->awk, ASE_T('\r')) == -1) return -1;
}
if (put_char (df->awk, ASE_T('\n')) == -1) return -1;
return 0;
}
static int put_char (ase_awk_t* awk, ase_char_t c)
{
awk->src.shared.buf[awk->src.shared.buf_len++] = c;
if (awk->src.shared.buf_len >= ASE_COUNTOF(awk->src.shared.buf))
{
if (flush_out (awk) == -1) return -1;
}
return 0;
}
static int flush_out (ase_awk_t* awk)
{
ase_ssize_t n;
ASE_ASSERT (awk->src.ios.out != ASE_NULL);
while (awk->src.shared.buf_pos < awk->src.shared.buf_len)
{
CLRERR (awk);
n = awk->src.ios.out (
ASE_AWK_IO_WRITE, awk->src.ios.custom_data,
&awk->src.shared.buf[awk->src.shared.buf_pos],
awk->src.shared.buf_len - awk->src.shared.buf_pos);
if (n <= 0)
{
if (ISNOERR(awk)) SETERR (awk, ASE_AWK_ESOUTWR);
return -1;
}
awk->src.shared.buf_pos += n;
}
awk->src.shared.buf_pos = 0;
awk->src.shared.buf_len = 0;
return 0;
}
int ase_awk_putsrcstr (ase_awk_t* awk, const ase_char_t* str)
{
while (*str != ASE_T('\0'))
{
if (put_char (awk, *str) == -1) return -1;
str++;
}
return 0;
}
int ase_awk_putsrcstrx (
ase_awk_t* awk, const ase_char_t* str, ase_size_t len)
{
const ase_char_t* end = str + len;
while (str < end)
{
if (put_char (awk, *str) == -1) return -1;
str++;
}
return 0;
}