Parsing multilevel text list

Question

I need to parse text lists:

1 List name

1 item
2 item
3 item

2 List name

1 item
2 item
3 item

3 List name

1 item
2 item
3 item

I was trying to use regular expression to split first level list:

import re
def re_show(pat, s):
 print re.compile(pat, re.S).sub("{\g<0>}", s),'\n'

s = '''
1 List name

1 item
2 item
3 item

2 List name

1 item
2 item
3 item

3 List name

1 item
2 item
3 item
'''

re_show('\n\d+.*?(?=\n\n\d+.*?\n\n)', s)

But it doesn't work. Instead of this:

{
1 List name

1 item
2 item
3 item}
{
2 List name

1 item
2 item
3 item}
{
3 List name

1 item
2 item
3 item}

I've got this:

{
1 List name}
{
1 item
2 item
3 item}
{
2 List name}
{
1 item
2 item
3 item}

3 List name

1 item
2 item
3 item

What would you recommend to solve this task?

Thanks for your answers. I've learned many new features of Python.

I think, I will use "state machine" approach as described here

Answer 1

I suspect I'm missing the point, but isn't this simply a question of looking for List?

Answer 2

here's one way using dictionary

f=open("myfile")
d={}
e=0
for line in f:
    line=line.rstrip()
    if "List" in line:
      e=e+1
      d.setdefault(e,[])
    d[e].append(line)
f.close()
for i ,j in d.iteritems():
    print i,j

Answer 3

class ListParser:

 def __init__(self, s):
  self.str = s.split("\n")
  print self.str
  self.answer = []

 def parse(self):
  self.nextLine()
  self.topList()
  return

 def topList(self):
  while(len(self.str) > 0):
   self.topListItem()

 def topListItem(self):
  l = self.nextLine()
  print "TOP: " + l
  l = self.nextLine()
  if l != '':
   raise Exception("expected blank line but found '%s'" % l)
  sub = self.sublist()

 def nextLine(self):
  return self.str.pop(0)

 def sublist(self):
  while True:
   l = self.nextLine()
   if l == '':
    return # end of sublist marked by blank line
   else:
    print "SUB: " + l

parser = ListParser(s)
parser.parse() 
print "done"

prints

TOP: 1 List name
SUB: 1 item
SUB: 2 item
SUB: 3 item
TOP: 2 List name
SUB: 1 item
SUB: 2 item
SUB: 3 item
TOP: 3 List name
SUB: 1 item
SUB: 2 item
SUB: 3 item
done

Answer 4

Not sure how fast it is, but this works as long as the spacing doesn't change.

>>>items = text.strip('\n').split('\n\n')
>>>dict((x, y.splitlines()) for x, y in zip(items[::2],items[1::2]))
{'3 List three': ['1 item', '2 item', '3 item'], '2 List two': ['1 item', '2 item', '3 item'], '1 List one': ['1 item', '2 item', '3 item']}

Answer 5

Do you have control over the list format? With just a little editing, you could turn that into config file format, and use the ConfigParser module.

Otherwise, how about with a little recursion?

from collections import defaultdict

def fill_data(data, key, sequence, pred):
    """Recursively fill the data dictionary"""
    for item in sequence:
        # if the pred is true, add it to the list
        if pred(item):
            data[key].append(item)
        # otherwise recurse, with item as key
        else:
            return fill_data(data, item, sequence, pred)
    return data

# a key->list dictionary
data = defaultdict(list)
# Get the text as a sequence of non-empty lines
lines = (l for l in s.splitlines() if l.strip())

def is_data_line(line):
    """Is this line a data line (i.e. two items)?"""
    return len(line.split()) == 2

result = fill_data(data, None, lines, is_data_line )

print dict(result)

Output (prettified):

{'2 List name': 
    ['1 item', '2 item', '3 item'], 
 '3 List name': 
    ['1 item', '2 item', '3 item'], 
 '1 List name': 
    ['1 item', '2 item', '3 item']}