Re: libapl apl_exec results displayed as ?

2022-07-25 Thread Dr . Jürgen Sauermann
Hi enztec, not entirely sure since I haven't written libapl myself. However, looking at the code it seems to me as if apl_exec returns an APL error code (most likely as per Error.def) if it fails. To capture the APL result I suppose you h

Re: libapl apl_exec results displayed as ?

2022-07-25 Thread Dr . Jürgen Sauermann
Hi again, just realized that there is a nice "info libapl" written by Dr. Dirk Laurie (the author of libapl) which explains the callback in section 2.4. Best Regards, Jürgen On 7/25/22 11:28 AM, Dr. Jürgen Sauermann wrote:

Re: libapl apl_exec results displayed as ?

2022-07-25 Thread Chris Moller
Sorry, but I have no idea how the interface between freepascal and C libraries works.  If your ⎕io is set to 1, the default, there's certainly no obvious reason for an ASCII 0 to be emitted to stdout.  Does freepascal start foreign code in a separate process?  Without reopening stdout? On 7/2

Re: libapl apl_exec results displayed as ?

2022-07-25 Thread Dr . Jürgen Sauermann
Hi, 0 is an error code, pretty much ⎕ET on page 287 of the APL language reference. The major error code, i.e. ↑⎕ET, is the same as in APL2, the minor error code 1↓⎕ET may differ. It has no relation to ⎕IO. I can't currently see how the result c

Re: libapl apl_exec results displayed as ?

2022-07-25 Thread Chris Moller
Yeah, I don't understand a path from a result code integer to a stdout character, which is why I suspect there's something going on in the freepascal-C interface.  Many years ago, I worked with a Java-to-C interface which was not straight-forward.  Maybe the freepascal-C interface is similarly