If you have the time, it would be great if you could fix this. Otherwise I'll
file a bug and get to it when I can...
On 2011-02-18, at 17:27, André Bargull wrote:
> No, a #passthrough block is ok for this method, or alternatively `#pragma
> "debugBacktrace=false"`. But adding just another #pragma makes the method
> even less readable, so #passthrough should be cleaner. Do you want to make
> the changes (in ecma-forin.lzl and e4x-foreach.lzl), or should I send out a
> changeset?
>
> On 2/18/2011 11:18 PM, P T Withington wrote:
>> Wow. Wonder what the compiler thinks it's doing?
>>
>> Any reason not to just put the proxy code in a passthrough?
>>
>> On 2011-02-18, at 17:12, André Bargull wrote:
>>
>>> Reproducible for me as well in backtrace mode.
>>> ---
>>> override flash_proxy function isAttribute (name:*) :Boolean {
>>> // FIXME [20110215 anba]: workaround for LPP-9760
>>> #pragma "throwsError=false"
>>> #pragma "catchFunctionExceptions=false"
>>> #pragma "debug=false"
>>> #pragma "debugSWF9=false"
>>> throw new Error("invalid operation");
>>> }
>>> ---
>>>
>>> The function is compiled to:
>>> ---
>>> /* -*- file: ecma-forin.lzl#99 -*- */
>>> flash_proxy override {
>>> /* -*- file: ecma-forin.lzl#99 -*- */
>>> function isAttribute (name_$0:*):Boolean {
>>> /* -*- file: -*- */
>>> var $1 = Debug;
>>> var $2 = $1.backtraceStack;
>>> try {
>>> var $3 = ["name", name_$0];
>>> $3.callee = "isAttribute";
>>> $3["this"] = this;
>>> $3.filename = "ecma-forin.lzl";
>>> $3.lineno = 99;
>>> $2.push($3);
>>> if ($2.length> $2.maxDepth) {
>>> $1.stackOverflow()
>>> };
>>> /* -*- file: ecma-forin.lzl#105 -*- */
>>> throw $3.lineno = 105, new Error("invalid operation")
>>> }
>>> /* -*- file: -*- */
>>> finally {
>>> if ($2) {
>>> $2.length--
>>> }}}
>>> ;
>>> isAttribute["_dbg_filename"] = "ecma-forin.lzl";
>>> isAttribute["_dbg_lineno"] = 99
>>> }
>>> ---
>>>
>>> On 2/18/2011 11:01 PM, P T Withington wrote:
>>>> Ok, with a clean everything, I finally realized this is only happening if
>>>> I enable backtrace in swf10:
>>>>
>>>> test/smoke/smokecheck.lzx?lzoptions=incremental(false)%2Cdebug(true)%2Cproxied(false)%2Cruntime(swf10)%2Cbacktrace(true)
>>>>
>>>> Thoughts?
>>>>
>>>> On 2011-02-17, at 16:45, André Bargull wrote:
>>>>
>>>>> This is a flex compiler error. I don't see this error, do you have any
>>>>> local changes which may produce bad ActionScript intermediate files?
>>>>>
>>>>> On 2/17/2011 9:35 PM, P T Withington wrote:
>>>>>> Anyone else seeing this?
>>>>>>
>>>>>>> Compilation Errors
>>>>>>>
>>>>>>> org.openlaszlo.sc.CompilerError: e4x-foreach.lzl: 125: Error: Syntax
>>>>>>> error: expected a definition keyword (such as function) after attribute
>>>>>>> flash_proxy, not override, in line: flash_proxy override {
>>>>>>> e4x-foreach.lzl: 125: Error: Syntax error: expecting rightbrace before
>>>>>>> leftbrace, in line: flash_proxy override {
>>>>>>> e4x-foreach.lzl: 128: Error: Syntax error: finally is unexpected, in
>>>>>>> line: finally {
>>>>>>> ecma-forin.lzl: 105: Error: Syntax error: expected a definition keyword
>>>>>>> (such as function) after attribute flash_proxy, not override, in line:
>>>>>>> flash_proxy override {
>>>>>>> ecma-forin.lzl: 105: Error: Syntax error: expecting rightbrace before
>>>>>>> leftbrace, in line: flash_proxy override {
>>>>>>> ecma-forin.lzl: 108: Error: Syntax error: finally is unexpected, in
>>>>>>> line: finally {
>>>>>>
>>>>
>>>>
>>
>>
