<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<p><br>
</p>
<br>
<div class="moz-cite-prefix">On 9/25/16 4:41 PM, Cantor, Scott
wrote:<br>
</div>
<blockquote cite="mid:9066579D-3173-4D40-BFDE-439111D63BCF@osu.edu"
type="cite">
<pre wrap="">On 9/25/16, 4:26 PM, "dev on behalf of Tom Zeller" <a class="moz-txt-link-rfc2396E" href="mailto:dev-bounces@shibboleth.netonbehalfoftzeller@dragonacea.biz"><dev-bounces@shibboleth.net on behalf of tzeller@dragonacea.biz></a> wrote:
</pre>
<blockquote type="cite">
<pre wrap=""> Just added another test to verify that the
IdentifiableBeanPostProcessor does indeed run on singleton beans in a
simple Spring application context, not sure what SWF is doing.
</pre>
</blockquote>
<pre wrap="">
It's deeper. From what I can tell, SWF doesn't run them at all, singletons or prototypes, I just had to get past the singletons to find the rest breaking.
Spring is supposed to auto-install and run those post processors and it does except inside SWF. But the weird thing is the action post-processor does run, obviously. One difference is one of them hooks pre-init and the other post-init. Maybe that's involved, don't know yet.
This is worth bottoming out either way since we have this stuff all declared and it turns out to not work.
</pre>
</blockquote>
<br>
<br>
I haven't looked into this really, but it occurs to me that: IIRC,
every SWF flow creates its own Spring ApplictionContext (I think as
a child of the MVC DispatcherServlet one). I didn't personally even
know this until I wrote some code once to output some debug logging
of the Spring context tree.<br>
<br>
Anyway, could that be related? I don't think bean post-processors
are inherited by child contexts, are they? I'm not 100% sure. But
if they're not, and if you aren't doing something to explicitly wire
the post-processor(s) into the bean set of every concrete flow, then
that could be the reason. <br>
<br>
</body>
</html>