fix for pprint #660 #661
fix for pprint #660 #661
Conversation
| @@ -192,7 +193,7 @@ def __len__(self): | |||
| """ | |||
| return len(self._callbacks) | |||
|
|
|||
| def __repr__(self): | |||
| def __str__(self): | |||
There was a problem hiding this comment.
Based on these definitions, I believe these two function definitions should be reversed.
__str__() should be the formatted and nicely printable representation.
https://docs.python.org/2/library/functions.html#str
__repr__() should be the string which could be passed back to eval()
https://docs.python.org/2/library/functions.html#repr
There was a problem hiding this comment.
"Based on these definitions, I believe these two function definitions should be reversed.
str() should be the formatted and nicely printable representation."
I kept it normal as earlier too facts used to be printed in a line only. If we want it to be formated, then we need only repr.
"repr() should be the string which could be passed back to eval()
https://docs.python.org/2/library/functions.html#repr"
The kind of facts we have(containing objects inside dict), that cant be EVALuated.
For me it use to throw SyntexError when used eval with fats sring.
There was a problem hiding this comment.
Yes, the string from repr() can not yet be passed to eval() successfully, but that's because of the string representation of the jnpr.junos.version_info objects. We can fix that when I address #651. If we keep the function the __repr__() function as it was, and add your new function as __str__(), then we most closely mimic the previous behavior when dev.facts was a true dict.
lib/jnpr/junos/factcache.py
Outdated
| for key in sorted(self): | ||
| if not key.startswith('_'): | ||
| self.get(key) | ||
| return pformat(self._cache) |
There was a problem hiding this comment.
Your implementation causes the "internal" facts (the facts whose names begin with an underscore) to also be printed. In addition, I believe the whole method implementation can be reduced to:
return pformat(dict(self))
There was a problem hiding this comment.
Thanks Stacy, This is even what I wanted.
There was a problem hiding this comment.
I discussed further with @vnitinv and we agreed that the current implementation in this pull request most closely mimics the previous behavior. Therefore, we should NOT switch the __repr__() and __str__() functions as I had previously suggested.
Now print print and pprint should work as expected.