Subj : Re: JSErrorReport->linebuf .. Help!!
To   : netscape.public.mozilla.jseng
From : j
Date : Mon Feb 21 2005 04:06 pm

On Mon, 21 Feb 2005 13:45:24 -0800, Brendan Eich <brendan@meer.net>
wrote:

>j@j.jp wrote:
>
>> I am using JS_SetInterrupt() for my scripts, so I can step through
>> them and do a lot of behind the scenes stuff (as needed). One of the
>> passed params is: jsbytecode *pc
>> 
>> Is it possible, from here, to get the source line of code, without
>> going back to the source file manually and moving down X lines and
>> grabbing the source? (This is not possible in my application.)
>
>
>No, you can get the source line number, but it's up to you to save the 
>source so you can find the numbered line.  Why can't you save scripts 
>that you compile, in memory you allocate?  No need for files.
>
>An alternative is to decompile to canonical "pretty-printed" source, 
>compile that, and execute.  Then you can decompile and skip to the right 
>canonical line.  Venkman does this (it saves the pretty-printed source, 
>of course, to avoid repeatedly decompiling the same script).
>
>/be

I am trying to be as memory conscious as possible. Any additional
overhead is baaaad... However, this could just be a debug version as
well... I will think about that and see if it's a possibility. Some of
these scripts are in excess of a megabyte total in source; that's a
lot of bulk to be carrying around..

I looked around, but didn't spot this 'Venkman'. Perhaps I'm in the
wrong folder? Are there any quick function(s) I can call into to get
this pretty-printed source version from the jsbytecode? Are there any
function names at all that you can point me to as a starting point
(searching, researching, etc.)

Thanks again!

.