Should not merge objects with same type name into one statistics entry in typestats

[Kevin Xiao]

Classes in different may have same class name, like

module.a.A
module.b.A

But `typestats` merges these objects into one entry key `"A"`, making the result of `show_growth` confusing.

I think it's better to use the full qualified name of a class as key is better,
{
"module.a.A": 12,
"module.b.A": 3
}

[Wichert Akkerman (wichert)]

I ran into this as well. I modified my copy of objgraph to always use a small utility function to build the dotted name for a type:

def _typename(o):
    t = type(o)
    m = getattr(t, '__module__' '')
    if m:
        return '%s.%s' % (m, t.__name__)
    else:
        return t.__name__

And modified all code using type(o).__name__ to use _typename(o) instead.

[Marius Gedminas (mgedmin)]

Apologies for not noticing this sooner. (Launchpad didn't send me any email, possibly because I didn't configure the project page correctly.)

The suggestion makes sense, I'll make it so.

[Marius Gedminas (mgedmin)]

Can we please continue this on GitHub? Their bugtracker is nicer.

https://github.com/mgedmin/objgraph/issues/4

[Marius Gedminas (mgedmin)]

Fixed in master, by adding an optional shortnames argument, with default value that keeps backwards-compatibility.

You can now use objgraph.show_growth(shortnames=False) to see fully-qualified type names.