[Bashdb-pydb] breakpoint bug (and patch)

[R. B. <ro...@pa...>]

DaveS writes:
 > According to the docs for 'break', when called with a function name it
 > should set the breakpoint at the first executable line of the function.
 > Right now "break sub1" will set the breakpoint at the definition of the
 > function, not the execution of it (like when setting a breakpoint by line
 > number).

Okay, will investigate when I get chance but I'm not sure when that
will be. Doing lots of Ruby/Rails nowadays.

 > It looks like the function name is not being set when the breakpoint is
 > parsed in get_brkpt_lineno (fns.py).  I've attached a patch that does the
 > right thing.
 > 
 > --- fns.py	2007-03-13 19:25:03.087000000 -0700
 > +++ fns.py.orig	2007-03-13 19:23:37.321375000 -0700
 > @@ -356,7 +356,6 @@
 >              code = func.func_code
 >              # use co_name to identify the bkpt (function names
 >              #could be aliased, but co_name is invariant)
 > -            funcname = arg
 >              lineno = code.co_firstlineno
 >              filename = code.co_filename
 >          except:

Although I agree there seems to be a bug, I'm not convinced this is
the right thing. The comment seems to indicate that code.co_name
should be used. So what are the merits/demerits of using that co_name
versus arg?

Could you make up a little test case that shows the bug, and maybe why
arg is better than co_name or vice versa? 

Thanks.