I am getting error with pipes in 3.08 Beta2 with the locate
stage, with several different pipes. at org.netrexx.njpipes.stages.locate.run(locate.java:316) at org.netrexx.njpipes.pipes.ThreadPool.run(ThreadPool.java:60) at java.lang.Thread.run(Unknown Source) Exiting due to uncaught Exception java.lang.String cannot be cast to netrexx.lang.Rexx This is a simple test pipe that gives the error: pipe (locate_fail_test ) This should show just 2. This test works OK: pipe (locate_ok_test ) It does give 20 Jeff _______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/ |
This works:
pipe gen 5 | rexx locate /2/ | console 2 I have never been too sure about when pipes generates a Rexx and when it is a String, but I am conditioned to stick a ‘rexx’ in between if I see that exception. René
_______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/ |
… and we might want to review that - what the original plan was, and how we can make it less astonishing for the innocent bystanders.
_______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/ |
I have thought about this, and changed the ‘gen’ stage to emit a Rexx type integer.
The change does not break anything for those pipes that have the ‘rexx’ cast in them. This might be a remnant from the ‘must be possible to integrate with java source’ principle, or it might have come about since we build the netrexx translator with mode binary and unwittlngly changed this for pipes also. I don’t know. I thank you for the bug report and think we should emit Rexx type as much as possible. best regards, René.
_______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/ |
I never did understand why the CASTING a String to Rexx can not be done by the NetRexx Compiler automatically in case of question, i.e. when Option BINARY is in effect, and a Rexx Builtin Function is used. Maybe this might be an issue for the next release, to be
discussed on the Source-Forge Discussion list ? Kindly, Thomas Schneider. ==========================================================================================
Am 26.08.2019 um 21:25 schrieb René
Jansen:
I have thought about this, and changed the ‘gen’ stage to emit a Rexx type integer. _______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
Thomas Schneider, Vienna, Austria (Europe) :-)
www.thsitc.com www.db-123.com |
In reply to this post by rvjansen
René, Thank you. We / I should take a look at locate, too. To make it accept java.lang.Strings. (And maybe do something more polite when it is fed some other object.) [Some other day, as I am on a cross country road trip right now.] Jeff On 8/26/2019 12:25 PM, René Jansen
wrote:
I have thought about this, and changed the ‘gen’ stage to emit a Rexx type integer. _______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/ |
In reply to this post by ThSITC
Hi Thomas,
I’ll answer on the developer’s list. René.
_______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/ |
In reply to this post by Jeff Hennick-3
I took a look at the locate source that I have. It
accepts an object in, works with a cast of it to Rexx to decide if
it has a match, and then tries to put out the original object. In
putting it out, it uses the system-wide output() method,
that apparently works only with type Rexx. I don't think I have
the source for output(). If this analysis is right, the
"correct" fix is going to be in output(), so locate
could work with anything that has a toString method. On 8/26/2019 6:50 PM, Jeff Hennick
wrote:
_______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/ |
In reply to this post by rvjansen
Hi Rene: 1.) Awaiting Your reply on NetRexx-Discuss (not yet found there) ... 2.) Wouldn't it be a good idea to SPLIT the IBM-NetRexx list from the PIPES List as well, so that, after the upcoming release of PIPES at the upcoming RexxLA meeting, the IBM-NetRexx list is only used for the Language issues ? I have No problems with this, however, but the latest communications have been very heavily PIPES-involved :-( Anyway, that shall be Your decision, not mine. I do adore NetRexx, and did hear and read about PIPES, when I did
work on IBM CMS (but never used PIPES, it's a pity) . Kindly, Thomas. =================================================================================================== Am 27.08.2019 um 04:31 schrieb René
Jansen:
Hi Thomas, _______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
Thomas Schneider, Vienna, Austria (Europe) :-)
www.thsitc.com www.db-123.com |
Hi Thomas,
yes, I need to answer that still. I agree that the discussions have been pipes-heavy, but that was because that part of the package is in development for this release; the goal is to bring it as close as possible to the CMS version. I do want to keep it one list, but I will move development-oriented discussion to its proper list. Not everyone has made the move, however. It is incredible, and indeed a pity, that your work on the CMS platform did not include PIPES. I worked from sep 2018- march 2019 in a VM shop, and with nearly everything that I did I needed Pipes; it is very closely integrated in most system maintenance execs with Rexx. Previously I played with the BatchPipes package on TSO, in the late nineties; with the upcoming release (which is real close now - found a few bugs, included some extra features, including the NetRexx workspace, which now runs pipelines brilliantly) you can catch up. best regards, René.
_______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/ |
Hi Rene, thanks for the prompt answer. As already noted, everything is OK for me, andf I do understand that when You are going to release PIPES that ibm-netRexx does become PIPES Heavy. Keep pls. up your brilliant work! Thomas. ================================================================================================== Am 29.08.2019 um 01:32 schrieb René
Jansen:
Hi Thomas, _______________________________________________ Ibm-netrexx mailing list [hidden email] Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
Thomas Schneider, Vienna, Austria (Europe) :-)
www.thsitc.com www.db-123.com |
Free forum by Nabble | Edit this page |