[with patch, positive review] trace no longer works in 3.0.2

[fwclarke]

----------------------------------------------------------------------
| SAGE Version 3.0.1, Release Date: 2008-05-05                       |
| Type notebook() for the GUI, and license() for information.        |
----------------------------------------------------------------------
sage: trace('2 + 2')
> <string>(1)<module>()
(Pdb) s
--Return--
> <string>(1)<module>()->4
(Pdb) s
4

But now

----------------------------------------------------------------------
| SAGE Version 3.0.2, Release Date: 2008-05-24                       |
| Type notebook() for the GUI, and license() for information.        |
----------------------------------------------------------------------

sage: trace('2 + 2')
> <string>(1)<module>()
(Pdb) s
NameError: "name 'Integer' is not defined"
> <string>(1)<module>()
(Pdb) s
--Return--
> <string>(1)<module>()->None
(Pdb) s
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)

/Users/mafwc/<ipython console> in <module>()

/Users/mafwc/sage-3.0.2/local/lib/python2.5/site-packages/sage/misc/trace.py in trace(code, preparse)
     41     """
     42     code = preparser.preparse(code)
---> 43     return pdb.runeval(code)
     44 
     45     # this could also be useful; it drops

/Users/mafwc/sage-3.0.2/local/lib/python2.5/pdb.py in runeval(expression, globals, locals)
   1144 
   1145 def runeval(expression, globals=None, locals=None):
-> 1146     return Pdb().runeval(expression, globals, locals)
   1147 
   1148 def runctx(statement, globals, locals):

/Users/mafwc/sage-3.0.2/local/lib/python2.5/bdb.py in runeval(self, expr, globals, locals)
    383         try:
    384             try:
--> 385                 return eval(expr, globals, locals)
    386             except BdbQuit:
    387                 pass

/Users/mafwc/<string> in <module>()

NameError: name 'Integer' is not defined

It seems that in 3.0.2 nothing is defined inside trace.

[mhansen]

I've attached a patch which grabs the interpreters namespace from IPython and passes it in as the globals to pdb.runeval. I'm not sure why trace worked before / why it's not working now though.

[fwclarke]

With the patch trace('2 + 2') works now, but strangely each instance of trace seems to start by recalling the last :

----------------------------------------------------------------------
| SAGE Version 3.0.2, Release Date: 2008-05-24                       |
| Type notebook() for the GUI, and license() for information.        |
----------------------------------------------------------------------
Loading SAGE library. Current Mercurial branch is: fwc
sage: trace('2 + 2')
> <string>(1)<module>()
(Pdb) s
--Return--
> <string>(1)<module>()->4
(Pdb) s
4
sage: trace('2 + 2')
> <string>(1)<module>()->4
(Pdb) s
--Return--
> <string>(1)<module>()->4
(Pdb) s
4
sage: trace('3 + 3')
> <string>(1)<module>()->4
(Pdb) s
--Return--
> <string>(1)<module>()->6
(Pdb) s
6
sage: trace('3 + 3')
> <string>(1)<module>()->6
(Pdb) s
--Return--
> <string>(1)<module>()->6
(Pdb) s
6

And there are still some things trace can't do:

sage: trace("P.<y> = QQ[]")
------------------------------------------------------------
   File "<string>", line 1
     P = QQ['y']; (y,) = P._first_ngens(Integer(1))
       ^
SyntaxError: invalid syntax

But there's no problem with

sage: P = QQ['y']; (y,) = P._first_ngens(Integer(1))
sage: P
Univariate Polynomial Ring in y over Rational Field
sage: P == parent(y)
True

[mhansen]

Why did this get positive review? It looks broken.

[mhansen]

I don't know what's going on recalling the last value produced by trace, but that seems to be something within pdb. Trace has never worked on (and probably shouldn't) work on statements (like assignment ones) -- only on expressions.

[was]

done while refereeing.

[was]

While refereeing trac_3345.patch I:

1. fix the issue with running assignments, and
2. added some mean serious doctests
3. Added a proper error message when one uses trace in the notebook

[was]

To apply this patch you should apply trac_3345.patch and trac_3345-part2.patch

[mabshoff]

When applying the two patches William recommended I am seeing the following failure on sage.math:

sage -t -long devel/sage/sage/misc/trace.py    
**********************************************************************
File "/scratch/mabshoff/release-cycle/sage-3.0.6.alpha1/tmp/trace.py", line 49:
    sage: print s.before
Expected:
    /rings/arith.py(...)factor()
    ...
    ipdb> c
    2 * 5
    <BLANKLINE>
Got:
    -3.0.6.alpha1/local/lib/python2.5/site-packages/
**********************************************************************
1 items had failures:
   1 of   9 in __main__.example_1
***Test Failed*** 1 failures.
For whitespace errors, see the file /scratch/mabshoff/release-cycle/sage-3.0.6.alpha1/tmp/.doctest_trace.py
         [3.8 s]
exit code: 1024

Am I doing something wrong?

Cheers,

Michael

[mhansen]

Positive review for all four trac_3345* patches.

[mabshoff]

Merged trac_3345.patch, trac_3345-part2.patch, trac_3345-part3.patch and trac_3345-part4.patch in Sage 3.0.6.rc0.

Cheers,

Michael