Here's a post-processing solution if you have too many different types of objects contributing to the JSON to attempt the JSONEncoder method and too many varying types to use a regex. This function collapses whitespace after a specified level, without needing to know the specifics of the data itself.
def collapse_json(text, indent=12):
"""Compacts a string of json data by collapsing whitespace after the
specified indent level
NOTE: will not produce correct results when indent level is not a multiple
of the json indent level
"""
initial = " " * indent
out = [] # final json output
sublevel = [] # accumulation list for sublevel entries
pending = None # holder for consecutive entries at exact indent level
for line in text.splitlines():
if line.startswith(initial):
if line[indent] == " ":
# found a line indented further than the indent level, so add
# it to the sublevel list
if pending:
# the first item in the sublevel will be the pending item
# that was the previous line in the json
sublevel.append(pending)
pending = None
item = line.strip()
sublevel.append(item)
if item.endswith(","):
sublevel.append(" ")
elif sublevel:
# found a line at the exact indent level *and* we have sublevel
# items. This means the sublevel items have come to an end
sublevel.append(line.strip())
out.append("".join(sublevel))
sublevel = []
else:
# found a line at the exact indent level but no items indented
# further, so possibly start a new sub-level
if pending:
# if there is already a pending item, it means that
# consecutive entries in the json had the exact same
# indentation and that last pending item was not the start
# of a new sublevel.
out.append(pending)
pending = line.rstrip()
else:
if pending:
# it's possible that an item will be pending but not added to
# the output yet, so make sure it's not forgotten.
out.append(pending)
pending = None
if sublevel:
out.append("".join(sublevel))
out.append(line)
return "\n".join(out)
For example, using this structure as input to json.dumps with an indent level of 4:
text = json.dumps({"zero": ["first", {"second": 2, "third": 3, "fourth": 4, "items": [[1,2,3,4], [5,6,7,8], 9, 10, [11, [12, [13, [14, 15]]]]]}]}, indent=4)
here's the output of the function at various indent levels:
>>> print collapse_json(text, indent=0)
{"zero": ["first", {"items": [[1, 2, 3, 4], [5, 6, 7, 8], 9, 10, [11, [12, [13, [14, 15]]]]], "second": 2, "fourth": 4, "third": 3}]}
>>> print collapse_json(text, indent=4)
{
"zero": ["first", {"items": [[1, 2, 3, 4], [5, 6, 7, 8], 9, 10, [11, [12, [13, [14, 15]]]]], "second": 2, "fourth": 4, "third": 3}]
}
>>> print collapse_json(text, indent=8)
{
"zero": [
"first",
{"items": [[1, 2, 3, 4], [5, 6, 7, 8], 9, 10, [11, [12, [13, [14, 15]]]]], "second": 2, "fourth": 4, "third": 3}
]
}
>>> print collapse_json(text, indent=12)
{
"zero": [
"first",
{
"items": [[1, 2, 3, 4], [5, 6, 7, 8], 9, 10, [11, [12, [13, [14, 15]]]]],
"second": 2,
"fourth": 4,
"third": 3
}
]
}
>>> print collapse_json(text, indent=16)
{
"zero": [
"first",
{
"items": [
[1, 2, 3, 4],
[5, 6, 7, 8],
9,
10,
[11, [12, [13, [14, 15]]]]
],
"second": 2,
"fourth": 4,
"third": 3
}
]
}
