creating nested dictionary from flat list with python - algorithm

Question

i have a list of file in this form:
base/images/graphs/one.png
base/images/tikz/two.png
base/refs/images/three.png
base/one.txt
base/chapters/two.txt
i would like to convert them to a nested dictionary of this sort:
{ "name": "base" , "contents":
[{"name": "images" , "contents":
[{"name": "graphs", "contents":[{"name":"one.png"}] },
{"name":"tikz", "contents":[{"name":"two.png"}]}
]
},
{"name": "refs", "contents":
[{"name":"images", "contents": [{"name":"three.png"}]}]
},
{"name":"one.txt", },
{"name": "chapters", "contents":[{"name":"two.txt"}]
]
}
trouble is, my attempted solution, given some input like images/datasetone/grapha.png" ,"images/datasetone/graphb.png" each one of them will end up in a different dictionary named "datasetone" however i'd like both to be in the same parent dictionary as they are in the same directory, how do i create this nested structure without duplicating parent dictionaries when there's more than one file in a common path?
here is what i had come up with and failed:
def path_to_tree(params):
start = {}
for item in params:
parts = item.split('/')
depth = len(parts)
if depth > 1:
if "contents" in start.keys():
start["contents"].append(create_base_dir(parts[0],parts[1:]))
else:
start ["contents"] = [create_base_dir(parts[0],parts[1:]) ]
else:
if "contents" in start.keys():
start["contents"].append(create_leaf(parts[0]))
else:
start["contents"] =[ create_leaf(parts[0]) ]
return start
def create_base_dir(base, parts):
l={}
if len(parts) >=1:
l["name"] = base
l["contents"] = [ create_base_dir(parts[0],parts[1:]) ]
elif len(parts)==0:
l = create_leaf(base)
return l
def create_leaf(base):
l={}
l["name"] = base
return l
b=["base/images/graphs/one.png","base/images/graphs/oneb.png","base/images/tikz/two.png","base/refs/images/three.png","base/one.txt","base/chapters/two.txt"]
d =path_to_tree(b)
from pprint import pprint
pprint(d)
In this example you can see we end up with as many dictionaries named "base" as there are files in the list, but only one is necessary, the subdirectories should be listed in the "contents" array.

Answer 1

This does not assume that all paths start with the same thing, so we need a list for it:
from pprint import pprint
def addBits2Tree( bits, tree ):
if len(bits) == 1:
tree.append( {'name':bits[0]} )
else:
for t in tree:
if t['name']==bits[0]:
addBits2Tree( bits[1:], t['contents'] )
return
newTree = []
addBits2Tree( bits[1:], newTree )
t = {'name':bits[0], 'contents':newTree}
tree.append( t )
def addPath2Tree( path, tree ):
bits = path.split("/")
addBits2Tree( bits, tree )
tree = []
for p in b:
print p
addPath2Tree( p, tree )
pprint(tree)
Which produces the following for your example path list:
[{'contents': [{'contents': [{'contents': [{'name': 'one.png'},
{'name': 'oneb.png'}],
'name': 'graphs'},
{'contents': [{'name': 'two.png'}],
'name': 'tikz'}],
'name': 'images'},
{'contents': [{'contents': [{'name': 'three.png'}],
'name': 'images'}],
'name': 'refs'},
{'name': 'one.txt'},
{'contents': [{'name': 'two.txt'}], 'name': 'chapters'}],
'name': 'base'}]

Answer 2

Omitting the redundant name tags, you can go on with :
import json
result = {}
records = ["base/images/graphs/one.png", "base/images/tikz/two.png",
"base/refs/images/three.png", "base/one.txt", "base/chapters/two.txt"]
recordsSplit = map(lambda x: x.split("/"), records)
for record in recordsSplit:
here = result
for item in record[:-1]:
if not item in here:
here[item] = {}
here = here[item]
if "###content###" not in here:
here["###content###"] = []
here["###content###"].append(record[-1])
print json.dumps(result, indent=4)
The # characters are used for uniqueness (there could be a folder which name was content in the hierarchy). Just run it and see the result.
EDIT : Fixed a few typos, added the output.