gecko-dev/js/ref/jsarray.c

1300 строки
31 KiB
C
Исходник Обычный вид История

/* -*- Mode: C; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 4 -*-
*
* The contents of this file are subject to the Netscape Public License
* Version 1.0 (the "NPL"); you may not use this file except in
* compliance with the NPL. You may obtain a copy of the NPL at
* http://www.mozilla.org/NPL/
*
* Software distributed under the NPL is distributed on an "AS IS" basis,
* WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
* for the specific language governing rights and limitations under the
* NPL.
*
* The Initial Developer of this code under the NPL is Netscape
* Communications Corporation. Portions created by Netscape are
* Copyright (C) 1998 Netscape Communications Corporation. All Rights
* Reserved.
*/
/*
* JS array class.
*/
#include "jsstddef.h"
#include <stdlib.h>
#include <string.h>
#include "prtypes.h"
#include "prassert.h"
#include "jsapi.h"
#include "jsarray.h"
#include "jsatom.h"
#include "jscntxt.h"
#include "jsconfig.h"
#include "jsfun.h"
#include "jsgc.h"
#include "jsinterp.h"
#include "jslock.h"
#include "jsnum.h"
#include "jsobj.h"
#include "jsstr.h"
/* 2^32 - 1 as a number and a string */
#define MAXINDEX 4294967295u
#define MAXSTR "4294967295"
/*
* Determine if the id represents an array index.
*
* An id is an array index according to ECMA by (15.4):
*
* "Array objects give special treatment to a certain class of property names.
* A property name P (in the form of a string value) is an array index if and
* only if ToString(ToUint32(P)) is equal to P and ToUint32(P) is not equal
* to 2^32-1."
*
* In our implementation, it would be sufficient to check for JSVAL_IS_INT(id)
* except that by using signed 32-bit integers we miss the top half of the
* valid range. This function checks the string representation itself; note
* that calling a standard conversion routine might allow strings such as
* "08" or "4.0" as array indices, which they are not.
*/
static JSBool
IdIsIndex(jsid id, jsuint *indexp)
{
JSString *str;
jschar *cp;
if (JSVAL_IS_INT(id)) {
jsint i;
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
i = JSVAL_TO_INT(id);
if (i < 0)
return JS_FALSE;
*indexp = (jsuint)i;
return JS_TRUE;
}
/* It must be a string. */
str = JSVAL_TO_STRING(id);
cp = str->chars;
if (JS7_ISDEC(*cp) && str->length < sizeof(MAXSTR)) {
jsuint index = JS7_UNDEC(*cp++);
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
jsuint oldIndex = 0;
jsint c;
if (index != 0) {
while (JS7_ISDEC(*cp)) {
oldIndex = index;
c = JS7_UNDEC(*cp);
index = 10*index + c;
cp++;
}
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
}
/* Make sure all characters were consumed and that it couldn't
* have overflowed.
*/
if (*cp == 0 &&
(oldIndex < (MAXINDEX / 10) ||
(oldIndex == (MAXINDEX / 10) && c < (MAXINDEX % 10))))
{
*indexp = index;
return JS_TRUE;
}
}
return JS_FALSE;
}
static JSBool
ValueIsLength(JSContext *cx, jsval v, jsuint *lengthp)
{
jsint i;
/*
* It's only an array length if it's an int or a double. Some relevant
* ECMA language is 15.4.2.2 - 'If the argument len is a number, then the
* length property of the newly constructed object is set to ToUint32(len)
* - I take 'is a number' to mean 'typeof len' returns 'number' in
* javascript.
*/
if (JSVAL_IS_INT(v)) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
i = JSVAL_TO_INT(v);
/* jsuint cast does ToUint32 */
if (lengthp)
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
*lengthp = (jsuint) i;
return JS_TRUE;
}
if (JSVAL_IS_DOUBLE(v)) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
/*
* XXXmccabe I'd love to add another check here, against
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
* js_DoubleToInteger(d) != d), so that ValueIsLength matches
* IdIsIndex, but it doesn't seem to follow from ECMA.
* (seems to be appropriate for IdIsIndex, though).
*/
return js_ValueToECMAUint32(cx, v, (uint32 *)lengthp);
}
return JS_FALSE;
}
static JSBool
ValueToIndex(JSContext *cx, jsval v, jsuint *lengthp)
{
jsint i;
if (JSVAL_IS_INT(v)) {
i = JSVAL_TO_INT(v);
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
/* jsuint cast does ToUint32. */
if (lengthp)
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
*lengthp = (jsuint)i;
return JS_TRUE;
}
return js_ValueToECMAUint32(cx, v, (uint32 *)lengthp);
}
JSBool
js_GetLengthProperty(JSContext *cx, JSObject *obj, jsuint *lengthp)
{
jsid id;
jsint i;
jsval v;
id = (jsid) cx->runtime->atomState.lengthAtom;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
/* Short-circuit, because js_ValueToECMAUint32 fails when
* called during init time.
*/
if (JSVAL_IS_INT(v)) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
i = JSVAL_TO_INT(v);
/* jsuint cast does ToUint32. */
*lengthp = (jsuint)i;
return JS_TRUE;
}
return js_ValueToECMAUint32(cx, v, (uint32 *)lengthp);
}
static JSBool
IndexToValue(JSContext *cx, jsuint length, jsval *vp)
{
if (length <= JSVAL_INT_MAX) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
*vp = INT_TO_JSVAL(length);
return JS_TRUE;
}
return js_NewDoubleValue(cx, (jsdouble)length, vp);
}
static JSBool
IndexToId(JSContext *cx, jsuint length, jsid *idp)
{
JSString *str;
JSAtom *atom;
if (length <= JSVAL_INT_MAX) {
*idp = (jsid) INT_TO_JSVAL(length);
} else {
str = js_NumberToString(cx, (jsdouble)length);
if (!str)
return JS_FALSE;
atom = js_AtomizeString(cx, str, 0);
if (!atom)
return JS_FALSE;
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
*idp = (jsid)atom;
}
return JS_TRUE;
}
JSBool
js_SetLengthProperty(JSContext *cx, JSObject *obj, jsuint length)
{
jsval v;
jsid id;
if (!IndexToValue(cx, length, &v))
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
return JS_FALSE;
id = (jsid) cx->runtime->atomState.lengthAtom;
return OBJ_SET_PROPERTY(cx, obj, id, &v);
}
JSBool
js_HasLengthProperty(JSContext *cx, JSObject *obj, jsuint *lengthp)
{
JSErrorReporter older;
jsid id;
JSBool ok;
jsval v;
older = JS_SetErrorReporter(cx, NULL);
id = (jsid) cx->runtime->atomState.lengthAtom;
ok = OBJ_GET_PROPERTY(cx, obj, id, &v);
JS_SetErrorReporter(cx, older);
if (!ok)
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
return JS_FALSE;
return ValueIsLength(cx, v, lengthp);
}
/*
* This get function is specific to Array.prototype.length and other array
* instance length properties. It calls back through the class get function
* in case some magic happens there (see call_getProperty in jsfun.c).
*/
static JSBool
array_length_getter(JSContext *cx, JSObject *obj, jsval id, jsval *vp)
{
return OBJ_GET_CLASS(cx, obj)->getProperty(cx, obj, id, vp);
}
static JSBool
array_length_setter(JSContext *cx, JSObject *obj, jsval id, jsval *vp)
{
jsuint newlen, oldlen, slot;
jsid id2;
jsval junk;
if (!ValueToIndex(cx, *vp, &newlen))
return JS_FALSE;
if (!js_GetLengthProperty(cx, obj, &oldlen))
return JS_FALSE;
for (slot = newlen; slot < oldlen; slot++) {
if (!IndexToId(cx, slot, &id2))
return JS_FALSE;
if (!OBJ_DELETE_PROPERTY(cx, obj, id2, &junk))
return JS_FALSE;
}
return IndexToValue(cx, newlen, vp);
}
static JSBool
array_addProperty(JSContext *cx, JSObject *obj, jsval id, jsval *vp)
{
jsuint index, length;
if (!(IdIsIndex(id, &index)))
return JS_TRUE;
if (!js_GetLengthProperty(cx, obj, &length))
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
return JS_FALSE;
if (index >= length) {
length = index + 1;
return js_SetLengthProperty(cx, obj, length);
}
return JS_TRUE;
}
static JSBool
array_convert(JSContext *cx, JSObject *obj, JSType type, jsval *vp)
{
jsuint length;
if (cx->version == JSVERSION_1_2) {
if (!js_GetLengthProperty(cx, obj, &length))
return JS_FALSE;
switch (type) {
case JSTYPE_NUMBER:
return IndexToValue(cx, length, vp);
case JSTYPE_BOOLEAN:
*vp = BOOLEAN_TO_JSVAL(length > 0);
return JS_TRUE;
default:
return JS_TRUE;
}
}
js_TryValueOf(cx, obj, type, vp);
return JS_TRUE;
}
JSClass js_ArrayClass = {
"Array",
0,
array_addProperty, JS_PropertyStub, JS_PropertyStub, JS_PropertyStub,
JS_EnumerateStub, JS_ResolveStub, array_convert, JS_FinalizeStub
};
static JSBool
array_join_sub(JSContext *cx, JSObject *obj, JSString *sep, JSBool literalize,
jsval *rval)
{
JSBool ok;
jsval v;
jsuint length, index;
jschar *chars, *ochars;
size_t nchars, growth, seplen;
const jschar *sepstr;
JSString *str;
PRHashEntry *he;
ok = js_GetLengthProperty(cx, obj, &length);
if (!ok)
return JS_FALSE;
ok = JS_TRUE;
if (literalize) {
he = js_EnterSharpObject(cx, obj, NULL, &chars);
if (!he)
return JS_FALSE;
if (IS_SHARP(he)) {
#if JS_HAS_SHARP_VARS
nchars = js_strlen(chars);
#else
chars[0] = '[';
chars[1] = ']';
chars[2] = 0;
nchars = 2;
#endif
goto make_string;
}
/*
* Allocate 1 + 3 + 1 for "[", the worst-case closing ", ]", and the
* terminating 0.
*/
growth = (1 + 3 + 1) * sizeof(jschar);
if (!chars) {
nchars = 0;
chars = malloc(growth);
if (!chars)
goto done;
} else {
MAKE_SHARP(he);
nchars = js_strlen(chars);
chars = realloc((ochars = chars), nchars * sizeof(jschar) + growth);
if (!chars) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
free(ochars);
goto done;
}
}
chars[nchars++] = '[';
} else {
if (length == 0) {
*rval = JS_GetEmptyStringValue(cx);
return ok;
}
chars = NULL;
nchars = 0;
}
sepstr = NULL;
seplen = sep->length;
v = JSVAL_NULL;
for (index = 0; index < length; index++) {
ok = JS_GetElement(cx, obj, index, &v);
if (!ok)
goto done;
if (JSVAL_IS_VOID(v) || JSVAL_IS_NULL(v)) {
str = cx->runtime->emptyString;
} else {
str = (literalize ? js_ValueToSource : js_ValueToString)(cx, v);
if (!str) {
ok = JS_FALSE;
goto done;
}
}
/* Allocate 3 + 1 at end for ", ", closing bracket, and zero. */
growth = (nchars + (sepstr ? seplen : 0) +
str->length +
3 + 1) * sizeof(jschar);
if (!chars) {
chars = malloc(growth);
if (!chars)
goto done;
} else {
chars = realloc((ochars = chars), growth);
if (!chars) {
free(ochars);
goto done;
}
}
if (sepstr) {
js_strncpy(&chars[nchars], sepstr, seplen);
nchars += seplen;
}
sepstr = sep->chars;
js_strncpy(&chars[nchars], str->chars, str->length);
nchars += str->length;
}
done:
if (literalize) {
if (chars) {
if (JSVAL_IS_VOID(v)) {
chars[nchars++] = ',';
chars[nchars++] = ' ';
}
chars[nchars++] = ']';
}
js_LeaveSharpObject(cx, NULL);
}
if (!ok) {
if (chars)
free(chars);
return ok;
}
make_string:
if (!chars) {
JS_ReportOutOfMemory(cx);
return JS_FALSE;
}
chars[nchars] = 0;
str = js_NewString(cx, chars, nchars, 0);
if (!str) {
free(chars);
return JS_FALSE;
}
*rval = STRING_TO_JSVAL(str);
return JS_TRUE;
}
static jschar comma_space_ucstr[] = {',', ' ', 0};
static jschar comma_ucstr[] = {',', 0};
static JSString comma_space = {2, comma_space_ucstr};
static JSString comma = {1, comma_ucstr};
#if JS_HAS_TOSOURCE
static JSBool
array_toSource(JSContext *cx, JSObject *obj, uintN argc, jsval *argv,
jsval *rval)
{
return array_join_sub(cx, obj, &comma_space, JS_TRUE, rval);
}
#endif
static JSBool
array_toString(JSContext *cx, JSObject *obj, uintN argc, jsval *argv,
jsval *rval)
{
JSBool literalize;
/*
* JS1.2 arrays convert to array literals, with a comma followed by a space
* between each element.
*/
literalize = (cx->version == JSVERSION_1_2);
return array_join_sub(cx, obj, literalize ? &comma_space : &comma,
literalize, rval);
}
static JSBool
array_join(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
JSString *str;
if (argc == 0)
return array_join_sub(cx, obj, &comma, JS_FALSE, rval);
str = js_ValueToString(cx, argv[0]);
if (!str)
return JS_FALSE;
argv[0] = STRING_TO_JSVAL(str);
return array_join_sub(cx, obj, str, JS_FALSE, rval);
}
#if !JS_HAS_MORE_PERL_FUN
static JSBool
array_nyi(JSContext *cx, const char *what)
{
1998-09-04 00:29:37 +04:00
JS_ReportErrorNumber(cx, js_GetErrorMessage, NULL, JSMSG_NO_PROTO, what);
return JS_FALSE;
}
#endif
static JSBool
InitArrayObject(JSContext *cx, JSObject *obj, jsuint length, jsval *vector)
{
jsval v;
jsid id;
jsuint index;
if (!IndexToValue(cx, length, &v))
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
return JS_FALSE;
id = (jsid) cx->runtime->atomState.lengthAtom;
if (!js_DefineProperty(cx, obj, id, v,
array_length_getter, array_length_setter,
JSPROP_PERMANENT,
NULL)) {
return JS_FALSE;
}
if (!vector)
return JS_TRUE;
for (index = 0; index < length; index++) {
if (!IndexToId(cx, index, &id))
return JS_FALSE;
if (!js_DefineProperty(cx, obj, id, vector[index],
JS_PropertyStub, JS_PropertyStub,
JSPROP_ENUMERATE,
NULL)) {
return JS_FALSE;
}
}
return JS_TRUE;
}
static JSBool
array_reverse(JSContext *cx, JSObject *obj, uintN argc, jsval *argv,
jsval *rval)
{
jsuint len, half, i;
jsid id, id2;
jsval v, v2;
if (!js_GetLengthProperty(cx, obj, &len))
return JS_FALSE;
half = len / 2;
for (i = 0; i < half; i++) {
if (!IndexToId(cx, i, &id))
return JS_FALSE;
if (!IndexToId(cx, len - i - 1, &id2))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id2, &v2))
return JS_FALSE;
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
if (!OBJ_SET_PROPERTY(cx, obj, id, &v2))
return JS_FALSE;
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
if (!OBJ_SET_PROPERTY(cx, obj, id2, &v))
return JS_FALSE;
}
*rval = OBJECT_TO_JSVAL(obj);
return JS_TRUE;
}
typedef struct QSortArgs {
void *vec;
size_t elsize;
void *pivot;
JSComparator cmp;
void *arg;
} QSortArgs;
static void
js_qsort_r(QSortArgs *qa, int lo, int hi)
{
void *pivot, *a, *b;
int i, j;
pivot = qa->pivot;
while (lo < hi) {
i = lo;
j = hi;
a = (char *)qa->vec + i * qa->elsize;
memmove(pivot, a, qa->elsize);
while (i < j) {
for (;;) {
b = (char *)qa->vec + j * qa->elsize;
if ((*qa->cmp)(b, pivot, qa->arg) <= 0)
break;
j--;
}
memmove(a, b, qa->elsize);
while (i < j && (*qa->cmp)(a, pivot, qa->arg) <= 0) {
i++;
a = (char *)qa->vec + i * qa->elsize;
}
memmove(b, a, qa->elsize);
}
memmove(a, pivot, qa->elsize);
if (i - lo < hi - i) {
js_qsort_r(qa, lo, i - 1);
lo = i + 1;
} else {
js_qsort_r(qa, i + 1, hi);
hi = i - 1;
}
}
}
PRBool
js_qsort(void *vec, size_t nel, size_t elsize, JSComparator cmp, void *arg)
{
void *pivot;
QSortArgs qa;
pivot = malloc(elsize);
if (!pivot)
return PR_FALSE;
qa.vec = vec;
qa.elsize = elsize;
qa.pivot = pivot;
qa.cmp = cmp;
qa.arg = arg;
js_qsort_r(&qa, 0, (int)(nel - 1));
free(pivot);
return PR_TRUE;
}
typedef struct CompareArgs {
JSContext *context;
jsval fval;
JSBool status;
} CompareArgs;
static int
sort_compare(const void *a, const void *b, void *arg)
{
jsval av = *(const jsval *)a, bv = *(const jsval *)b;
CompareArgs *ca = arg;
JSContext *cx = ca->context;
jsdouble cmp = -1;
jsval fval, argv[2], rval;
JSBool ok;
fval = ca->fval;
if (fval == JSVAL_NULL) {
JSString *astr, *bstr;
if (av == bv) {
cmp = 0;
} else if (av == JSVAL_VOID || bv == JSVAL_VOID) {
/* Put undefined properties at the end. */
cmp = (av == JSVAL_VOID) ? 1 : -1;
} else if ((astr = js_ValueToString(cx, av)) != NULL &&
(bstr = js_ValueToString(cx, bv)) != NULL) {
cmp = js_CompareStrings(astr, bstr);
} else {
ca->status = JS_FALSE;
}
} else {
argv[0] = av;
argv[1] = bv;
ok = js_CallFunctionValue(cx,
OBJ_GET_PARENT(cx, JSVAL_TO_OBJECT(fval)),
fval, 2, argv, &rval);
if (ok) {
ok = js_ValueToNumber(cx, rval, &cmp);
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
/* Clamp cmp to -1, 0, 1. */
if (JSDOUBLE_IS_NaN(cmp)) {
/* XXX report some kind of error here? ECMA talks about
* 'consistent compare functions' that don't return NaN, but is
* silent about what the result should be. So we currently
* ignore it.
*/
cmp = 0;
} else if (cmp != 0) {
cmp = cmp > 0 ? 1 : -1;
}
} else {
ca->status = ok;
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
}
}
return (int)cmp;
}
/* XXXmccabe do the sort helper functions need to take int? (Or can we claim
* that 2^32 * 32 is too large to worry about?) Something dumps when I change
* to unsigned int; is qsort using -1 as a fencepost?
*/
static JSBool
array_sort(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
jsval fval;
CompareArgs ca;
jsuint len, i;
jsval *vec;
jsid id;
if (argc > 0) {
if (JSVAL_IS_PRIMITIVE(argv[0])) {
1998-09-04 00:29:37 +04:00
JS_ReportErrorNumber(cx, js_GetErrorMessage, NULL,
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
JSMSG_BAD_SORT_ARG);
return JS_FALSE;
}
fval = argv[0];
} else {
fval = JSVAL_NULL;
}
if (!js_GetLengthProperty(cx, obj, &len))
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
return JS_FALSE;
vec = JS_malloc(cx, (size_t) len * sizeof(jsval));
if (!vec)
return JS_FALSE;
for (i = 0; i < len; i++) {
ca.status = IndexToId(cx, i, &id);
if (!ca.status)
goto out;
ca.status = OBJ_GET_PROPERTY(cx, obj, id, &vec[i]);
if (!ca.status)
goto out;
}
ca.context = cx;
ca.fval = fval;
ca.status = JS_TRUE;
if (!js_qsort(vec, (size_t) len, sizeof(jsval), sort_compare, &ca)) {
JS_ReportOutOfMemory(cx);
ca.status = JS_FALSE;
}
if (ca.status) {
ca.status = InitArrayObject(cx, obj, len, vec);
if (ca.status)
*rval = OBJECT_TO_JSVAL(obj);
}
out:
if (vec)
JS_free(cx, vec);
return ca.status;
}
#ifdef NOTYET
/*
* From "Programming perl", Larry Wall and Randall L. Schwartz, Copyright XXX
* O'Reilly & Associates, Inc., but with Java primitive type sizes for i, l,
* and so on:
*
* a An ASCII string, will be null padded.
* A An ASCII string, will be space padded.
* b A bit string, low-to-high order.
* B A bit string, high-to-low order.
* h A hexadecimal string, low nybble first.
* H A hexadecimal string, high nybble first.
* c A signed char value.
* C An unsigned char value.
* s A signed short (16-bit) value.
* S An unsigned short (16-bit) value.
* i A signed integer (32-bit) value.
* I An unsigned integer (32-bit) value.
* l A signed long (64-bit) value.
* L An unsigned long (64-bit) value.
* n A short in "network" byte order.
* N An integer in "network" byte order.
* f A single-precision float in IEEE format.
* d A double-precision float in IEEE format.
* p A pointer to a string.
* x A null byte.
* X Back up one byte.
* @ Null-fill to absolute position.
* u A uuencoded string.
*
* Each letter may be followed by a number giving the repeat count. Together
* the letter and repeat count make a field specifier. Field specifiers may
* be separated by whitespace, which will be ignored.
*/
static JSBool
array_pack(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
#if JS_HAS_MORE_PERL_FUN
#else
return array_nyi(cx, "pack");
#endif
}
#endif /* NOTYET */
static JSBool
array_push(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
#if JS_HAS_MORE_PERL_FUN
jsuint length;
uintN i;
jsid id;
if (!js_GetLengthProperty(cx, obj, &length))
return JS_FALSE;
for (i = 0; i < argc; i++) {
if (!IndexToId(cx, length + i, &id))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, obj, id, &argv[i]))
return JS_FALSE;
}
/*
* If JS1.2, follow Perl4 by returning the last thing pushed. Otherwise,
* return the new array length.
*/
length += argc;
if (cx->version == JSVERSION_1_2) {
*rval = argc ? argv[argc-1] : JSVAL_VOID;
} else {
if (!IndexToValue(cx, length, rval))
return JS_FALSE;
}
return js_SetLengthProperty(cx, obj, length);
#else
return array_nyi(cx, "push");
#endif
}
static JSBool
array_pop(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
#if JS_HAS_MORE_PERL_FUN
jsuint index;
jsid id;
jsval junk;
if (!js_GetLengthProperty(cx, obj, &index))
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
return JS_FALSE;
if (index > 0) {
index--;
if (!IndexToId(cx, index, &id))
return JS_FALSE;
/* Get the to-be-deleted property's value into rval. */
if (!OBJ_GET_PROPERTY(cx, obj, id, rval))
return JS_FALSE;
if (!OBJ_DELETE_PROPERTY(cx, obj, id, &junk))
return JS_FALSE;
}
return js_SetLengthProperty(cx, obj, index);
#else
return array_nyi(cx, "pop");
#endif
}
static JSBool
array_shift(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
#if JS_HAS_MORE_PERL_FUN
jsuint length, i;
jsid id, id2;
jsval v, junk;
if (!js_GetLengthProperty(cx, obj, &length))
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
return JS_FALSE;
if (length > 0) {
length--;
id = JSVAL_ZERO;
/* Get the to-be-deleted property's value into rval ASAP. */
if (!OBJ_GET_PROPERTY(cx, obj, id, rval))
return JS_FALSE;
/*
* Slide down the array above the first element.
*/
if (length > 0) {
for (i = 1; i <= length; i++) {
if (!IndexToId(cx, i, &id))
return JS_FALSE;
if (!IndexToId(cx, i - 1, &id2))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, obj, id2, &v))
return JS_FALSE;
}
}
/* Delete the only or last element. */
if (!OBJ_DELETE_PROPERTY(cx, obj, id, &junk))
return JS_FALSE;
}
return js_SetLengthProperty(cx, obj, length);
#else
return array_nyi(cx, "shift");
#endif
}
static JSBool
array_unshift(JSContext *cx, JSObject *obj, uintN argc, jsval *argv,
jsval *rval)
{
#if JS_HAS_MORE_PERL_FUN
jsuint length, last;
uintN i;
jsid id, id2;
jsval v;
if (!js_GetLengthProperty(cx, obj, &length))
return JS_FALSE;
if (argc > 0) {
/* Slide up the array to make room for argc at the bottom. */
if (length > 0) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
last = length;
while (last--) {
if (!IndexToId(cx, last, &id))
return JS_FALSE;
if (!IndexToId(cx, last + argc, &id2))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, obj, id2, &v))
return JS_FALSE;
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
}
}
/* Copy from argv to the bottom of the array. */
for (i = 0; i < argc; i++) {
if (!IndexToId(cx, i, &id))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, obj, id, &argv[i]))
return JS_FALSE;
}
/* Follow Perl by returning the new array length. */
length += argc;
if (!js_SetLengthProperty(cx, obj, length))
return JS_FALSE;
}
return IndexToValue(cx, length, rval);
#else
return array_nyi(cx, "unshift");
#endif
}
static JSBool
array_splice(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
#if JS_HAS_MORE_PERL_FUN
jsuint length, begin, end, count, delta, last;
uintN i;
jsdouble d;
jsid id, id2;
jsval v;
JSObject *obj2;
/* Nothing to do if no args. Otherwise lock and load length. */
if (argc == 0)
return JS_TRUE;
if (!js_GetLengthProperty(cx, obj, &length))
return JS_FALSE;
/* Convert the first argument into a starting index. */
if (!js_ValueToNumber(cx, *argv, &d))
return JS_FALSE;
d = js_DoubleToInteger(d);
if (d < 0) {
d += length;
if (d < 0)
d = 0;
} else if (d > length) {
d = length;
}
begin = (jsuint)d; /* d has been clamped to uint32 */
argc--;
argv++;
/* Convert the second argument from a count into a fencepost index. */
delta = length - begin;
if (argc == 0) {
count = delta;
end = length;
} else {
if (!js_ValueToNumber(cx, *argv, &d))
return JS_FALSE;
d = js_DoubleToInteger(d);
if (d < 0)
d = 0;
else if (d > delta)
d = delta;
count = (jsuint)d;
end = begin + count;
argc--;
argv++;
}
if (count == 1 && cx->version == JSVERSION_1_2) {
/*
* JS lacks "list context", whereby in Perl one turns the single
* scalar that's spliced out into an array just by assigning it to
* @single instead of $single, or by using it as Perl push's first
* argument, for instance.
*
* JS1.2 emulated Perl too closely and returned a non-Array for
* the single-splice-out case, requiring callers to test and wrap
* in [] if necessary. So JS1.3, default, and other versions all
* return an array of length 1 for uniformity.
*/
if (!IndexToId(cx, begin, &id))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, rval))
return JS_FALSE;
} else {
if (cx->version != JSVERSION_1_2 || count > 0) {
/*
* Create a new array value to return. Our ECMA v2 proposal specs
* that splice always returns an array value, even when given no
* arguments. We think this is best because it eliminates the need
* for callers to do an extra test to handle the empty splice case.
*/
obj2 = js_NewArrayObject(cx, 0, NULL);
if (!obj2)
return JS_FALSE;
*rval = OBJECT_TO_JSVAL(obj2);
}
/* If there are elements to remove, put them into the return value. */
if (count > 0) {
for (last = begin; last < end; last++) {
if (!IndexToId(cx, last, &id))
return JS_FALSE;
if (!IndexToId(cx, last - begin, &id2))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, obj2, id2, &v))
return JS_FALSE;
}
}
}
/* Find the direction (up or down) to copy and make way for argv. */
if (argc > count) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
delta = (jsuint)argc - count;
last = length;
/* (uint) end could be 0, so can't use vanilla >= test */
while (last-- > end) {
if (!IndexToId(cx, last, &id))
return JS_FALSE;
if (!IndexToId(cx, last + delta, &id2))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, obj, id2, &v))
return JS_FALSE;
}
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
length += delta;
} else if (argc < count) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
delta = count - (jsuint)argc;
for (last = end; last < length; last++) {
if (!IndexToId(cx, last, &id))
return JS_FALSE;
if (!IndexToId(cx, last - delta, &id2))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, obj, id2, &v))
return JS_FALSE;
}
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
length -= delta;
}
/* Copy from argv into the hole to complete the splice. */
for (i = 0; i < argc; i++) {
if (!IndexToId(cx, begin + i, &id))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, obj, id, &argv[i]))
return JS_FALSE;
}
/* Update length in case we deleted elements from the end. */
return js_SetLengthProperty(cx, obj, length);
#else
return array_nyi(cx, "splice");
#endif
}
#if JS_HAS_SEQUENCE_OPS
/*
* Python-esque sequence operations.
*/
static JSBool
array_concat(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
JSObject *nobj, *aobj;
jsuint slot, length, alength;
jsid id, id2;
jsval v;
uintN i;
nobj = js_NewArrayObject(cx, 0, NULL);
if (!nobj)
return JS_FALSE;
/* Only add the first element as an array if it looks like one. Treat the
* target the same way as the arguments.
*/
/* XXXmccabe Might make sense to recast all of this as a do-while. */
if (js_HasLengthProperty(cx, obj, &length)) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
for (slot = 0; slot < length; slot++) {
if (!IndexToId(cx, slot, &id))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, nobj, id, &v))
return JS_FALSE;
}
} else {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
length = 1;
v = OBJECT_TO_JSVAL(obj);
if (!OBJ_SET_PROPERTY(cx, nobj, JSVAL_ZERO, &v))
return JS_FALSE;
}
for (i = 0; i < argc; i++) {
v = argv[i];
if (JSVAL_IS_OBJECT(v)) {
aobj = JSVAL_TO_OBJECT(v);
if (aobj && js_HasLengthProperty(cx, aobj, &alength)) {
for (slot = 0; slot < alength; slot++) {
if (!IndexToId(cx, slot, &id))
return JS_FALSE;
if (!IndexToId(cx, length + slot, &id2))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, aobj, id, &v))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, nobj, id2, &v))
return JS_FALSE;
}
length += alength;
continue;
}
}
if (!IndexToId(cx, length, &id))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, nobj, id, &v))
return JS_FALSE;
length++;
}
*rval = OBJECT_TO_JSVAL(nobj);
return JS_TRUE;
}
static JSBool
array_slice(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
JSObject *nobj;
jsuint length, begin, end, slot;
jsdouble d;
jsid id, id2;
jsval v;
nobj = js_NewArrayObject(cx, 0, NULL);
if (!nobj)
return JS_FALSE;
if (!js_GetLengthProperty(cx, obj, &length))
return JS_FALSE;
begin = 0;
end = length;
if (argc > 0) {
if (!js_ValueToNumber(cx, argv[0], &d))
return JS_FALSE;
d = js_DoubleToInteger(d);
if (d < 0) {
d += length;
if (d < 0)
d = 0;
} else if (d > length) {
d = length;
}
begin = (jsuint)d;
if (argc > 1) {
if (!js_ValueToNumber(cx, argv[1], &d))
return JS_FALSE;
d = js_DoubleToInteger(d);
if (d < 0) {
d += length;
if (d < 0)
d = 0;
} else if (d > length) {
d = length;
}
end = (jsuint)d;
}
}
for (slot = begin; slot < end; slot++) {
if (!IndexToId(cx, slot, &id))
return JS_FALSE;
if (!IndexToId(cx, slot - begin, &id2))
return JS_FALSE;
if (!OBJ_GET_PROPERTY(cx, obj, id, &v))
return JS_FALSE;
if (!OBJ_SET_PROPERTY(cx, nobj, id2, &v))
return JS_FALSE;
}
*rval = OBJECT_TO_JSVAL(nobj);
return JS_TRUE;
}
#endif /* JS_HAS_SEQUENCE_OPS */
static JSFunctionSpec array_methods[] = {
#if JS_HAS_TOSOURCE
{js_toSource_str, array_toSource, 0},
#endif
{js_toString_str, array_toString, 0},
/* Perl-ish methods. */
{"join", array_join, 1},
{"reverse", array_reverse, 0},
{"sort", array_sort, 1},
#ifdef NOTYET
{"pack", array_pack, 1},
#endif
{"push", array_push, 1},
{"pop", array_pop, 0},
{"shift", array_shift, 0},
{"unshift", array_unshift, 1},
{"splice", array_splice, 1},
/* Python-esque sequence methods. */
#if JS_HAS_SEQUENCE_OPS
{"concat", array_concat, 0},
{"slice", array_slice, 0},
#endif
{0}
};
static JSBool
Array(JSContext *cx, JSObject *obj, uintN argc, jsval *argv, jsval *rval)
{
jsuint length;
jsval *vector;
/* If called without new, replace obj with a new Array object. */
if (!cx->fp->constructing) {
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
obj = js_NewObject(cx, &js_ArrayClass, NULL, NULL);
if (!obj)
return JS_FALSE;
*rval = OBJECT_TO_JSVAL(obj);
}
if (argc == 0) {
length = 0;
vector = NULL;
} else if (cx->version != JSVERSION_1_2 &&
argc == 1 && ValueIsLength(cx, argv[0], &length))
{
/*
* Only use 1 arg as first element for version 1.2; for any other
* version (including 1.3), follow ECMA and use it as a length.
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
*/
vector = NULL;
} else {
length = (jsuint) argc;
vector = argv;
}
return InitArrayObject(cx, obj, length, vector);
}
JSObject *
js_InitArrayClass(JSContext *cx, JSObject *obj)
{
JSObject *proto;
proto = JS_InitClass(cx, obj, NULL, &js_ArrayClass, Array, 1,
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
NULL, array_methods, NULL, NULL);
/* Initialize the Array prototype object so it gets a length property. */
if (!proto || !InitArrayObject(cx, proto, 0, NULL))
js.c jsemit.c jsemit.h jsgc.c jsinterp.c jsopcode.c jsopcode.def - Switch improvements: - JSOP_CONDSWITCH is a 1 byte nop, not variable length with the same kind of immediate operand as JSOP_LOOKUPSWITCH (which is useless except for decompilation). New scheme uses SRC_COMMA notes on each JSOP_CASE opcode, usually 2 bytes per note, and a typically-1-byte 2nd offset on SRC_SWITCH: 1 + 2 * ncases vs. the previous JSOP_LOOKUPSWITCH immediate, which consumed: 4 * ncases bytes after the switch opcode just for decompilation. - SRC_SWITCH has two offsets, first to end of switch as before, the second to first case if JSOP_CONDSWITCH, for decompilation. - Optimize switches with all-constant cases using JSOP_TABLESWITH, or if that can't be used, JSOP_LOOKUPSWITCH, before falling back on ECMAv2's JSOP_CONDSWITCH. - Use cx->gcDisabled when evaluating case exprs at compile time for old, pre-ECMAv2 switches, to prevent branch-callback-based GC invocations from ripping apart the unrooted temporary script for each case expr. - Fixed up stale SRC_SWITCH comments in jsemit.h. jsemit.c jsemit.h - TREE_CONTEXT_INIT to match ATOM_LIST_INIT, not English word order. - Reorganized JSCodeGenerator to sort of match argument order to js_InitCodeGenerator. - Got rid of confusing CG_RESET* macros and used memset(cg, 0, sizeof *cg) and non-zero-default init in js_InitCodeGenerator. js_ResetCodeGenerator just releases the code and temp arena pools and leaves the cg in a state where it must be re-initialized (as before, but more obvious). - In the same spirit, don't do partial "resets" or src and trynotes in their js_FinishTaking*Notes functions -- those are friends of jsscript.c and are not general, idempotent functions. jsapi.c jsapi.h jsarray.c jsatom.c jsatom.h jscntxt.c jsemit.c jsmsg.def jsnum.c jsobj.c jsopcode.c jsregexp.c jsscan.c jsstr.c jsxdrapi. - Use PR_snprintf rather than sprintf always, so we don't have to worry about systems with 64-bit longs that overflow 12-byte buffers and open Morris-Worm-type security holes. - Trim extra spaces, fix hanging indentation, and similar anal retention. - Renamed JSMSG_BAD_PROTO_SORT to JSMSG_BAD_SORT_ARG cuz that's what it is complaining about. - SRC_CATCHGUARD still lived in comments, but it's SRC_CATCH in code. jscntxt.c jscntxt.h jsinterp.c - Packed nearby JSPackedBools and added a new one: gcDisabled, for use by jsemit.c's pre-ECMAv2 switch case expr eval. - Rip out old js_InterpreterHooks stuff from original liveconnect (moja). - Remove javaData and savedErrors from JSContext. Leaving it to fur or shaver to remove javaData from jsscript.h.
1998-09-08 09:39:51 +04:00
return NULL;
return proto;
}
JSObject *
js_NewArrayObject(JSContext *cx, jsuint length, jsval *vector)
{
JSObject *obj;
obj = js_NewObject(cx, &js_ArrayClass, NULL, NULL);
if (!obj)
return NULL;
if (!InitArrayObject(cx, obj, length, vector)) {
cx->newborn[GCX_OBJECT] = NULL;
return NULL;
}
return obj;
}