How do I compare two complex data structures?

Question

I have some nested datastructures, each something like:

[ ('foo', [ {'a':1, 'b':2},
                 {'a':3.3, 'b':7} ]),
  ('bar', [ {'a':4, 'd':'efg', 'e':False} ])   ]

I need to compare these structures, to see if there are any differences. Short of writing a function to explicitly walk the structure, is there an existing library or method of doing this kind of recursive comparison?

Do you need to just see if they are equal or not, or actually find where they differ? — , May 04 '10 at 10:48
To see where they differ, have a look at: http://stackoverflow.com/a/26171760/5069869 — Bernhard, Jun 22 '16 at 13:18

score 6 · Accepted Answer · answered May 04 '10 at 10:50

6

The built-in aggregation types (list, tuple, dict, etc.) already support equality and relational comparison. For types you create, you need to implement the rich comparison methods.

answered May 04 '10 at 10:50

Ignacio Vazquez-Abrams

740,318
145
1,296
1,325

score 3 · Answer 2 · answered May 04 '10 at 15:02

Your example data structures will already do appropriate equality testing, because you are using built-in data types which properly implement __eq__ and __ne__, including recursing into nested values.

If you want to include your own classes, you need to implement both of these methods (note that implementing __eq__ does not imply that if you do a != comparison your __eq__ will be called, you must implement __ne__, too).

score 0 · Answer 3 · answered Sep 11 '14 at 05:20

0

If I don't need to use the diff within Python itself, I would probably end up hacking it. Convert both to yaml, and run a diff on them. :D

answered Sep 11 '14 at 05:20

Sandip Bhattacharya

974
9
11

How do I compare two complex data structures?

3 Answers3