Then the question is how to avoid breaking the structure on copy
I'm using whitespace to vertically align some toml config and when editing said config with your great library noticed that alignment disappears, below is a small repro
content = """
[keys.a]
# ↓ this whitespace disappears
[keys .'a'.'c']
'd' = 'e'
"""
def testtoml():
from tomlkit import dumps
from tomlkit import parse
doc = parse(content)
for mode in doc["keys"]: # self to self shouldn't change anything?
doc["keys"][mode] = doc["keys"][mode]
assert dumps(doc) == doc
# print(dumps(doc))
testtoml()
It seems that the latter whitespace matches the first encountered in [keys.a] (if I add whitespace to [keys.a] then it propagates to further as well
Is it possible to preserve the difference in whitespace so that editing wouldn't "import" whitespace from upper keys, but retain its own?
upd: or maybe it's because
https://github.com/sdispater/tomlkit/issues/255#issuecomment-1407551898
Then the question is how to avoid breaking the structure on copy
I'm using whitespace to vertically align some toml config and when editing said config with your great library noticed that alignment disappears, below is a small repro
It seems that the latter whitespace matches the first encountered in
[keys.a]
(if I add whitespace to[keys.a]
then it propagates to further as wellIs it possible to preserve the difference in whitespace so that editing wouldn't "import" whitespace from upper keys, but retain its own?
(tested this on a master branch of your library)