Converting "true" (JSON) to Python equivalent "True"
PythonJsonDictionaryBooleanPython Problem Overview
The Train status API I use recently added two additional key value pairs (has_arrived, has_departed)
in the JSON object, which caused my script to crash.
Here's the dictionary:
{
"response_code": 200,
"train_number": "12229",
"position": "at Source",
"route": [
{
"no": 1,
"has_arrived": false,
"has_departed": false,
"scharr": "Source",
"scharr_date": "15 Nov 2015",
"actarr_date": "15 Nov 2015",
"station": "LKO",
"actdep": "22:15",
"schdep": "22:15",
"actarr": "00:00",
"distance": "0",
"day": 0
},
{
"actdep": "23:40",
"scharr": "23:38",
"schdep": "23:40",
"actarr": "23:38",
"no": 2,
"has_departed": false,
"scharr_date": "15 Nov 2015",
"has_arrived": false,
"station": "HRI",
"distance": "101",
"actarr_date": "15 Nov 2015",
"day": 0
}
]
}
Not surprisingly, I got the following error:
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
NameError: name 'false' is not defined
If I am not mistaken, I think this is because the boolean value in the JSON response is false
/true
whereas Python recognizes False
/True
.
Is there any way around it?
PS: I tried converting the JSON response of has_arrived
to string and then converting it back to a boolean value, only to find out that I'll always get a True
value if there's any character in the string.
I am kinda stuck here.
Python Solutions
Solution 1 - Python
Even though Python's object declaration syntax is very similar to Json syntax, they're distinct and incompatible. As well as the True
/true
issue, there are other problems (eg Json and Python handle dates very differently, and python allows single quotes and comments while Json does not).
Instead of trying to treat them as the same thing, the solution is to convert from one to the other as needed.
Python's native json library can be used to parse (read) the Json in a string and convert it into a python object, and you already have it installed...
# Import the library
import json
# Define a string of json data
data_from_api = '{"response_code": 200, ...}'
data = json.loads(data_from_api)
# info is now a python dictionary (or list as appropriate) representing your Json
You can convert python objects to json too...
info_as_json = json.dumps(info)
Example:
# Import the json library
import json
# Get the Json data from the question into a variable...
data_from_api = """{
"response_code": 200,
"train_number": "12229",
"position": "at Source",
"route": [
{
"no": 1, "has_arrived": false, "has_departed": false,
"scharr": "Source",
"scharr_date": "15 Nov 2015", "actarr_date": "15 Nov 2015",
"station": "LKO", "actdep": "22:15", "schdep": "22:15",
"actarr": "00:00", "distance": "0", "day": 0
},
{
"actdep": "23:40", "scharr": "23:38", "schdep": "23:40",
"actarr": "23:38", "no": 2, "has_departed": false,
"scharr_date": "15 Nov 2015", "has_arrived": false,
"station": "HRI", "distance": "101",
"actarr_date": "15 Nov 2015", "day": 0
}
]
}"""
# Convert that data into a python object...
data = json.loads(data_from_api)
print(data)
And a second example showing how the True/true conversion happens. Note also the changes to quotation and how the comment is stripped...
info = {'foo': True, # Some insightful comment here
'bar': 'Some string'}
# Print a condensed representation of the object
print(json.dumps(info))
> {"bar": "Some string", "foo": true}
# Or print a formatted version which is more human readable but uses more bytes
print(json.dumps(info, indent=2))
> {
> "bar": "Some string",
> "foo": true
> }
Solution 2 - Python
You can also do a cast to boolean with the value. For example, assuming that your data is called "json_data":
value = json_data.get('route')[0].get('has_arrived') # this will pull "false" into *value
boolean_value = bool(value == 'true') # resulting in False being loaded into *boolean_value
It's kind of hackey, but it works.
Solution 3 - Python
Instead of doing eval
on the answer, use the json
module.
Solution 4 - Python
It is possible to utilize Python's boolean value for int, str, list etc.
For example:
bool(1) # True
bool(0) # False
bool("a") # True
bool("") # False
bool([1]) # True
bool([]) # False
In Json file, you can set
"has_arrived": 0,
Then in your Python code
if data["has_arrived"]:
arrived()
else:
not_arrived()
The issue here is not to confuse 0 indicated for False and 0 for its value.
Solution 5 - Python
"""
String to Dict (Json): json.loads(jstr)
Note: in String , shown true, in Dict shown True
Dict(Json) to String: json.dumps(jobj)
"""
>>> jobj = {'test': True}
>>> jstr = json.dumps(jobj)
>>> jobj
{'test': True}
>>> jstr
'{"test": true}'
>>> json.loads(jstr)
{'test': True}
Solution 6 - Python
I would like to add one more thing that would work for people who are reading from a file.
with open('/Users/mohammed/Desktop/working_create_order.json')as jsonfile:
data = json.dumps(jsonfile.read())
Then follow the above accepted answer i.e.
data_json = json.loads(data)
Solution 7 - Python
{ "value": False } or { "key": false } is not a valid json https://jsonlint.com/
Solution 8 - Python
json.loads cant parse the pythons boolean type (False, True). Json wants to have small letters false, true
my solution:
python_json_string.replace(": True,", ": true,").replace(": False,", ": false,")