aniani/elinks
1
0
Fork 0
mirror of https://github.com/rkd77/elinks.git synced 2024-12-04 14:46:47 -05:00
Code

Bug 755: A hack to make SMJS preserve integers as private data.

Browse Source 
The previous code cast the integer (long actually) to void * and gave
that to JS_SetPrivate.  This did not work because JS_SetPrivate
expects pointers to be aligned and replaces the least significant bit
with a tag.

[ From attachment 379 of bug 755.  This is less clean than attachment
  380 (aka commit 871a1befad in ELinks
  0.12.GIT) but avoids the need to raise the SpiderMonkey version
  requirement in the ELinks 0.11 stable branch.  --KON ]
This commit is contained in:
Kalle Olavi Niemitalo 2007-06-24 17:27:51 +03:00 committed by Kalle Olavi Niemitalo
parent c16a37a9ad
commit cd855a269c
1 changed files with 23 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
src/ecmascript/spidermonkey/form.c
View File

@ -60,7 +60,23 @@ static JSBool input_set_property(JSContext *ctx, JSObject *obj, jsval id, jsval
/* Each @input_class object must have a @form_class parent. */
static const JSClass input_class = {
"input", /* here, we unleash ourselves */
JSCLASS_HAS_PRIVATE, /* struct form_state * */
/* In instances of @input_class, the private data is not
* actually a pointer, although SMJS assumes so. Rather, it
* is an integer used as an index to view_state.form_info[].
* This allows ELinks to reallocate form_info[] without
* keeping track of SMJS objects that refer to its elements.
*
* JS_SetPrivate converts private pointers to jsval, and
* JS_GetPrivate converts back. These conversions assume that
* private pointers are aligned. Therefore, ELinks must not
* cast the integer directly to void *. Instead, ELinks takes
* advantage of the fact that the jsval format for private
* pointers is the same as for integers (presumably to make GC
* ignore the pointers). So when ELinks is initializing the
* private data, it converts the integer to a jsval and from
* there to a pointer, which SMJS then converts back to the
* same jsval. */
JSCLASS_HAS_PRIVATE,
JS_PropertyStub, JS_PropertyStub,
input_get_property, input_set_property,
JS_EnumerateStub, JS_ResolveStub, JS_ConvertStub, JS_FinalizeStub
@ -125,9 +141,10 @@ static const JSFunctionSpec input_funcs[] = {
static struct form_state *
input_get_form_state(JSContext *ctx, JSObject *obj, struct view_state *vs)
{
int n = (int)(long)JS_GetInstancePrivate(ctx, obj,
void *private = JS_GetInstancePrivate(ctx, obj,
(JSClass *) &input_class,
NULL);
int n = JSVAL_TO_INT(PRIVATE_TO_JSVAL(private));
return &vs->form_info[n];
}
@ -504,10 +521,11 @@ get_input_object(JSContext *ctx, JSObject *jsform, long number)
/* FIXME: That is NOT correct since the real containing element
* should be its parent, but gimme DOM first. --pasky */
JSObject *jsinput = JS_NewObject(ctx, (JSClass *) &input_class, NULL, jsform);
void *private = JSVAL_TO_PRIVATE(INT_TO_JSVAL(number));
JS_DefineProperties(ctx, jsinput, (JSPropertySpec *) input_props);
JS_DefineFunctions(ctx, jsinput, (JSFunctionSpec *) input_funcs);
JS_SetPrivate(ctx, jsinput, (void *)number); /* to @input_class */
JS_SetPrivate(ctx, jsinput, private); /* to @input_class */
return jsinput;;
}