CODEWITHSUNDEEP
pythonjsonfile-handling
Zlatan Omerovic
Zlatan Omerovic

Reputation: 4097

Python file input (write mode) issue with JSON

I'm learning Python and I'm following official documentation from:

Section: 7.2.2. Saving structured data with json for Python 3

I'm testing the json.dump() function to dump my python set into a file pointer:

>>> response = {"success": True, "data": ["test", "array", "response"]}
>>> response
{'success': True, 'data': ['test', 'array', 'response']}
>>> import json
>>> json.dumps(response)
'{"success": true, "data": ["test", "array", "response"]}'
>>> f = open('testfile.txt', 'w', encoding='UTF-8')
>>> f
<_io.TextIOWrapper name='testfile.txt' mode='w' encoding='UTF-8'>
>>> json.dump(response, f)

The file testfile.txt already exists in my working directory and even if it didn't, statement f = open('testfile.txt', 'w', encoding='UTF-8') would have re-create it, truncated.

The json.dumps(response) converts my response set into a valid JSON object, so that's fine.

Problem is when I use the json.dumps(response, f) method, which actually updates my testfile.txt, but it gets truncated.

I've managed to do a reverse workaround like:

>>> f = open('testfile.txt', 'w', encoding='UTF-8')
>>> f.write(json.dumps(response));
56
>>>

After which the contents of my testfile.txt become as expected:

{"success": true, "data": ["test", "array", "response"]}

Even, this approach works too:

>>> json.dump(response, open('testfile.txt', 'w', encoding='UTF-8'))

Why does this approach fail?:

>>> f = open('testfile.txt', 'w', encoding='UTF-8')
>>> json.dump(response, f)

Note that I don't get any errors from the console; just a truncated file.

Upvotes: 0

Views: 542

Answers (1)

Mark Tolonen
Mark Tolonen

Reputation: 177674

It looks like you aren't exiting the interactive prompt to check the file. Close the file to flush it:

f.close()

It will close if you exit the interactive prompt as well.

Upvotes: 2

Related Questions