CODEWITHSUNDEEP
parsingpython-3.xtext-indent
user3426681
user3426681

Reputation: 151

Parsing nested indented text into lists

Parsing nested indented text into lists

Hi,

maybe someone can give me a start help.

I have nested indented txt similar to this. I should parse that into a nested list structure like

TXT = r"""
Test1
    NeedHelp
        GotStuck
            Sometime
            NoLuck
    NeedHelp2
        StillStuck
        GoodLuck
"""

Nested_Lists = ['Test1', 
    ['NeedHelp', 
        ['GotStuck', 
            ['Sometime', 
            'NoLuck']]], 
    ['NeedHelp2', 
        ['StillStuck', 
        'GoodLuck']]
]

Nested_Lists = ['Test1', ['NeedHelp', ['GotStuck', ['Sometime', 'NoLuck']]], ['NeedHelp2', ['StillStuck', 'GoodLuck']]]

Any help for python3 would be appriciated

Upvotes: 3

Views: 2363

Answers (4)

deargle
deargle

Reputation: 517

Riffing off of this answer, if entire lines want to be retained and if those lines consist of more than just variable names, t.type == NAME can be substituted with t.type == NEWLINE, and that if-statement can append the stripped line instead of the t.string. Something like this:

from tokenize import NEWLINE, INDENT, DEDENT, tokenize

def parse(file):
    stack = [[]]
    lastindent = len(stack)

    def push_new_list():
        stack[-1].append([])
        stack.append(stack[-1][-1])
        return len(stack)

    for t in tokenize(file.readline):
        if t.type == NEWLINE:
            if lastindent != len(stack):
                stack.pop()
                lastindent = push_new_list()
            stack[-1].append(t.line.strip()) # add entire line to current list
        elif t.type == INDENT:
            lastindent = push_new_list()
        elif t.type == DEDENT:
            stack.pop()
    return stack[-1]

Otherwise, the lines get split on any token, where a token includes spaces, parentheses, brackets, etc.

Upvotes: 0

SZIEBERTH Ádám
SZIEBERTH Ádám

Reputation: 4206

I hope you can understand my solution. If not, ask.

def nestedbyindent(string, indent_char=' '):
    splitted, i = string.splitlines(), 0
    def first_non_indent_char(string):
        for i, c in enumerate(string):
            if c != indent_char:
                return i
        return -1
    def subgenerator(indent):
        nonlocal i
        while i < len(splitted):
            s = splitted[i]
            title = s.lstrip()
            if not title:
                i += 1
                continue
            curr_indent = first_non_indent_char(s)
            if curr_indent < indent:
                break
            elif curr_indent == indent:
                i += 1
                yield title
            else:
                yield list(subgenerator(curr_indent))
    return list(subgenerator(-1))

>>> nestedbyindent(TXT)
['Test1', ['NeedHelp', ['GotStuck', ['Sometime', 'NoLuck']],
'NeedHelp2',['StillStuck', 'GoodLuck']]]

Upvotes: 4

Kei Minagawa
Kei Minagawa

Reputation: 4521

Here is the answer that is very non-Pythonic and verbose way. But it seems to work.

TXT = r"""
Test1
    NeedHelp
        GotStuck
            Sometime
            NoLuck
    NeedHelp2
        StillStuck
        GoodLuck
"""

outString = '['
level = 0
first = 1
for i in TXT.split("\n")[1:]:
    count = 0
    for j in i:
        if j!=' ':
            break
        count += 1
    count /= 4 #4 space = 1 indent
    if i.lstrip()!='':
        itemStr = "'" + i.lstrip() + "'"
    else:
        itemStr = ''
    if level < count:
        if first:
            outString += '['*(count - level) + itemStr
            first = 0
        else:
            outString += ',' + '['*(count - level) + itemStr
    elif level > count:
        outString += ']'*(level - count) + ',' + itemStr
    else:
        if first:
            outString += itemStr
            first = False
        else:
            outString += ',' + itemStr
    level = count
if len(outString)>1:
    outString = outString[:-1] + ']'
else:
    outString = '[]'

output = eval(outString)
#['Test1', ['NeedHelp', ['GotStuck', ['Sometime', 'NoLuck']], 'NeedHelp2', ['StillStuck', 'GoodLuck']]]

Upvotes: 0

jfs
jfs

Reputation: 414875

You could exploit Python tokenizer to parse the indented text:

from tokenize import NAME, INDENT, DEDENT, tokenize

def parse(file):
    stack = [[]]
    lastindent = len(stack)

    def push_new_list():
        stack[-1].append([])
        stack.append(stack[-1][-1])
        return len(stack)

    for t in tokenize(file.readline):
        if t.type == NAME:
            if lastindent != len(stack):
                stack.pop()
                lastindent = push_new_list()
            stack[-1].append(t.string) # add to current list
        elif t.type == INDENT:
            lastindent = push_new_list()
        elif t.type == DEDENT:
            stack.pop()
    return stack[-1]

Example:

from io import BytesIO
from pprint import pprint
pprint(parse(BytesIO(TXT.encode('utf-8'))), width=20)

Output

['Test1',
 ['NeedHelp',
  ['GotStuck',
   ['Sometime',
    'NoLuck']]],
 ['NeedHelp2',
  ['StillStuck',
   'GoodLuck']]]

Upvotes: 8

Related Questions