python pyparsing string: A,b,b,b,A,b,b,b* [closed]

Question

I have a crazy problem.

I am trying to use pyparsing to parse something like this: (dots are unimportant suppressed text)

...... A
B .......
B .......
...... A
B .......
B .......

What I need is something like this: (to connect A and B element into one list)

 [ [ [A],[B,B] ], [ [A],[B,B] ] , ...]

This is my code which does not work and only returns the first [A]

table = pyparsing.OneOrMore(pyparsing.Group(A + (pyparsing.OneOrMore(pyparsing.Group(B) | pyparsing.SkipTo(B).suppress()))) | pyparsing.SkipTo(A).suppress())

I have already solved this pyparsing into a list like this:

[ [A],[B],[B], [A],[B],[B] , ...]

But this was not acceptable, since the A and B elements were not directly connected in one list.

Worth mentioning is that

table1 = pyparsing.ZeroOrMore(pyparsing.Group(A) | pyparsing.SkipTo(A).suppress())
table2 = pyparsing.ZeroOrMore(pyparsing.Group(B) | pyparsing.SkipTo(B).suppress())

works and return a list of all A elements and B elements.

Answer 1

Yes, you can have OneOrMore's embedded within other OneOrMore's - it would severely limit the parsers you could write if you couldn't.

I think you might be able to adapt your existing solution if you do better grouping. See how the Groups are defined in this toy example:

test = """
...... A 
B ....... 
B ....... 
...... A 
B ....... 
B ......."""

from pyparsing import Literal, Word, printables, Group, OneOrMore

A = Literal("A")
B = Literal("B")

notAorB = Word(printables, excludeChars="AB")

parser = OneOrMore(Group(A + Group(OneOrMore(B))))
parser.ignore(notAorB)

print parser.parseString(test).asList()

Prints:

[['A', ['B', 'B']], ['A', ['B', 'B']]]

Answer 2

I think you have two options:

• rethink the way you're parsing the text

• stay happy with what you have and just clean things afterwards: You have a list like your_list=[A,B,B,A,B,B] ? You can just do

   [ [x[i], x[i+1] + x[i+2]] for i in range(len(your_list)//3)] 

  the + will concatenate your two [B] lists (represented as x[i+1] and x[i+2] ).