OSX readline detection fails in the debugger

[fperez]

This is a continuation of an old bug:

https://bugs.launchpad.net/ipython/+bug/411599

Darrel Schiebel reports:

I find that the fixed loop duration is insufficient. I needed to set it to 150 (see below). At 50 or less repetitions, ipython startup consistently dies within gdb.

--- /opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/IPython/rlineimpl.py.orig 2011-05-09 18:52:30.000000000 -0400
+++ /opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/IPython/rlineimpl.py 2011-05-09 18:42:42.000000000 -0400
@@ -37,7 +37,7 @@
     # infinite loops with such code, so for now I'm taking a more conservative
     # approach. See https://bugs.launchpad.net/ipython/+bug/411599.
     #while True:
-    for i in range(10):
+    for i in range(150):
         try:
             (status, result) = commands.getstatusoutput( "otool -L %s | grep libedit" % _rl.__file__ )
             break

I tried to find a way to either re-open this defect or submit a new defect from launch pad, but it claims that it no longer knows how to submit ipython defects.

End Darrell's report

I'm wondering if the right approach shouldn't be to have an exponential backoff of the retries... It strikes me as ugly to rapid-fire 150 subcommands...

[minrk]

#475 should close this issue. I replicated the error in gdb, and after the change I can't get it to happen anymore, whether using readline or libedit.

If it does somehow fail (e.g. otool is somehow missing), the action is now just to warn that libedit detection failed, not raise an error.