On Mar 05, 2010, at 05:11 AM, Martin v. Löwis wrote:

Johan Harjano ran into an interesting problem when trying to run the Django test suite under Python 3.1.

Django has doctests of the form

...

...

...

a6.headline u'Default headline'

Even when converting the doctest with 2to3, the expected output is unmodified. However, in 3.x, the expected output will change (i.e. not produce an u"" prefix anymore).

For this reason, I always recommend using print, even though...

Now, it might be possible to reformulate the test case (e.g. use print() instead of relying on repr), however, this is undesirable as a) the test should continue to test in 2.x that the result object is a unicode string, and b) it makes the test less readable.

If you really want to test that it's a unicode, shouldn't you actually test its type? (I'm not sure what would happen with that under 2to3.) Besides, the type of the string is very rarely important, so I think the u-prefix and quotes is mostly just noise.

I would like to find a solution where this gets automatically corrected, e.g. through 2to3, or through changes to doctest, or through changes of str.__repr__.

I think Michael was also talking about changes to doctest that would automatically sort dictionaries and sets. Again, it's not hard to write doctests correctly, but it's surprisingly common to implicitly rely on sort order. I think the right place to change these is in doctest. -Barry

"Martin v. Löwis" wrote: [...]

Any proposal appreciated.

I propose screaming “help me, I have written a test suite using nothing but string matching assertions, what is wrong with me?!” -Andrew.

On 3/4/2010 11:11 PM, "Martin v. Löwis" wrote:

Johan Harjano ran into an interesting problem when trying to run the Django test suite under Python 3.1.

Django has doctests of the form

...

...

...

a6.headline u'Default headline'

Even when converting the doctest with 2to3, the expected output is unmodified. However, in 3.x, the expected output will change (i.e. not produce an u"" prefix anymore).

Now, it might be possible to reformulate the test case (e.g. use print() instead of relying on repr), however, this is undesirable as a) the test should continue to test in 2.x that the result object is a unicode string, and b) it makes the test less readable.

I would like to find a solution where this gets automatically corrected, e.g. through 2to3, or through changes to doctest, or through changes of str.__repr__.

Any proposal appreciated.

What is the easiest thing that works? If 2to3 can fix string literals and fix code within doc strings, would it be difficult to fix expected strings within doc strings? On the otherhand, as Foord pointed out, the 'u' prefix is something of a CPythonism not required by the language def, so an 'ignore leading u on expected output' flag would be useful. But 'correcting'

...

...

a,b u'ah', u'hah'

is too much to expect. Doctest is both useful and fragile because it *almost* imitates human testing methods. If tests like the above were bunched, and my primary goal were to make Django tests work, and work now, without changing the style, I would be tempted to use something like the following: def u(s): #import as needed from a testutil module if type(s) is unicode: # str in 3.x print(s) else print('') #or any 'nevermatch' string

...

...

u(a6.headline) 'Default headline'

which both retains the type test and readability. This would be a project specific solution, but then, the other comments suggest that wanting to combine a type and value test this way seems to be someone project specific. Terry Jan Reedy

On Thu, Mar 4, 2010 at 8:11 PM, "Martin v. Löwis" <martin@v.loewis.de> wrote:

Johan Harjano ran into an interesting problem when trying to run the Django test suite under Python 3.1.

Django has doctests of the form

...

...

...

a6.headline u'Default headline'

Even when converting the doctest with 2to3, the expected output is unmodified. However, in 3.x, the expected output will change (i.e. not produce an u"" prefix anymore).

Now, it might be possible to reformulate the test case (e.g. use print() instead of relying on repr), however, this is undesirable as a) the test should continue to test in 2.x that the result object is a unicode string, and b) it makes the test less readable.

I would like to find a solution where this gets automatically corrected, e.g. through 2to3, or through changes to doctest, or through changes of str.__repr__.

Any proposal appreciated.

How about a heuristic rule (which you have to explicitly select) that changes u'XXX' into 'XXX' inside triply-quoted strings given certain contexts, e.g. only at the start of the line, only if there is a nearby preceding line starting with '>>>'? Exactly what context is of the right strength will have to be determined experimentally; if there are a lot of tests outputting things like [u'...'] or {u'...': u'...'} the context may have to be made more liberal. Possibly \bu('.*'|".*") would do it? The issue shows (yet again) a general problem with doctests being overspecified -- the test shouldn't care that the output starts with 'u', it should only care that the value is unicode, but there's no easy way to express this in doctests. But since these doctests exist I suggest that the practical way forward is to stick with them rather than trying to reformulate all the tests. -- --Guido van Rossum (python.org/~guido)

Martin v. Löwis wrote:

Johan Harjano ran into an interesting problem when trying to run the Django test suite under Python 3.1.

Django has doctests of the form

...

...

...

a6.headline u'Default headline'

Even when converting the doctest with 2to3, the expected output is unmodified. However, in 3.x, the expected output will change (i.e. not produce an u"" prefix anymore).

Now, it might be possible to reformulate the test case (e.g. use print() instead of relying on repr), however, this is undesirable as a) the test should continue to test in 2.x that the result object is a unicode string, and b) it makes the test less readable.

I would like to find a solution where this gets automatically corrected, e.g. through 2to3, or through changes to doctest, or through changes of str.__repr__.

Any proposal appreciated.

You can use a custom DocTestRunner that replaces sys.displayhook in its run() method and records the changed output. Something like the attached seems to do the trick. Regards, Ziga import sys import doctest import linecache import __builtin__ def greet(): """ The standard greeting, in unicode. >>> greet() u'Hello, world!' """ return u'Hello, world!' orig_displayhook = sys.displayhook def py3_displayhook(value): if not isinstance(value, unicode): return orig_displayhook(value) __builtin__._ = value s = repr(value) if s.startswith(("u'", 'u"')): s = s[1:] print >> sys.stdout, s class Runner(doctest.DocTestRunner): converted_files = {} def __init__(self, checker=None, verbose=None, optionflags=0): doctest.DocTestRunner.__init__(self, checker, False, optionflags) def run(self, test, compileflags=None, out=None, clear_globs=True): fn = test.filename if fn not in self.converted_files: self.converted_files[fn] = linecache.getlines(fn) sys.displayhook = py3_displayhook try: return doctest.DocTestRunner.run(self, test, compileflags, out, clear_globs) finally: sys.displayhook = orig_displayhook def report_failure(self, out, test, example, got): lines = [" " * example.indent + line for line in got.splitlines(True)] pos = test.lineno + example.lineno + 1 self.converted_files[test.filename][pos:pos+len(lines)] = lines def _test(): import __main__ finder = doctest.DocTestFinder() runner = Runner() for test in finder.find(__main__): runner.run(test) print "".join(runner.converted_files[__main__.__file__]) if __name__ == "__main__": _test()