Can you do fuzzy comparison between integers, floats, etc?

[Asaf Peleg]

Can I use testfixtures to ensure a dictionary/json object contains the correct type versus doing an exact match, or should i look elsewhere?

[Chris Withers]

On 31/03/2016 16:11, Asaf Peleg wrote:

Can I use testfixtures to ensure a dictionary/json object contains the correct type versus doing an exact match, or should i look elsewhere?

Why would you not want to make explicit assertions about the values?

You *can* do want you're asking for, it's just not a great idea:

from testfixtures import Comparison as C, compare

compare(actual, expected=dict(thing=C(int), other_thing=C(str)))

cheers,

Chris

[Daniel Fortunov]

If you're trying to do schema validation of JSON input you may want to look at schema.py: https://github.com/keleshev/schema

--
You received this message because you are subscribed to the Google Groups "Simplistix Open Source Software" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simplistix+...@googlegroups.com.
To post to this group, send email to simpl...@googlegroups.com.
Visit this group at https://groups.google.com/group/simplistix.
For more options, visit https://groups.google.com/d/optout.

[Chris Withers]

For that sort of thing, I've been massively impressed with Voluptuous:

https://github.com/alecthomas/voluptuous

cheers,

Chris

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please visit http://www.symanteccloud.com
______________________________________________________________________

[Asaf Peleg]

We generally uses testfixtures.compare to confirm that the responses from our RESTful api matches up with an expected response.  Usually we want to compare an exact value but sometimes we just want to do general validation like you mentioned via your links.

Take the following json response for example...

{

   'message': 'Your report was successfully created'

   'report_url': 'http://somedomain.com/downloads/your_report_201604010001.pdf'

   'time-elapsed': 23

}

We'd like to confirm that key time-elapsed was present in the response and that the value of it was an integer but it's not necessary to test exactly how long it took to generate the report.  

As for the message, we want to confirm that the message matches exactly as expected.

All this being said I will take a closer look at the libraries you provided to determine if they would be a better general fit for our use case.