Not finding the strings expected with pyparsing

Question

I'm trying to parse a string using pyparsing. Using the code below

import pyparsing as pyp

aString = "C((H2)(C(H3))) C((H1)(Cl1)) C(((C(H3))3))"

aSub = '(('+ pyp.Word('()'+pyp.srange('[A-Za-z0-9]'))+'))'
substituent = aSub('sub')

for t,s,e in substituent.scanString(aString):
    print t.sub

I get no output. However, in string aString = "C((H2)(C(H3))) C((H1)(Cl1)) C(((C(H3))3))" there are multiple occurences of ((stuff)) - specifically ((H2)(C(H3))), C((H1)(Cl1)) and C(((C(H3))3)).

My understanding of Word() was that the input (in the case of a single input, as I have) represents all possible character combinations that will successfully return a match.

Running the code

import pyparsing as pyp

aString = "C((H2)(C(H3))) C((H1)(Cl1)) C(((C(H3))3))"

aSub = '(' + pyp.Word(pyp.srange('[A-Za-z0-9]'))+')'
substituent = aSub('sub')

for t,s,e in substituent.scanString(aString):
    print t.sub

gives an output of

['(', 'H2', ')']
['(', 'H3', ')']
['(', 'H1', ')']
['(', 'Cl1', ')']
['(', 'H3', ')']

All I've changed is an additional external set of parentheses, as well as the option of parentheses inside of the string, which the desired strings have. I'm not sure why the first program gives me nothing, while the second string gives me (part of) what I want.

score 2 · Accepted Answer · edited May 23 '17 at 12:14

2

The problem is the pyparsing works left to right (source). So having the right parenthesis erases what you are searching for on the right. For instance:

aSub = '(('+ pyp.Word('()'+pyp.srange('[A-Za-z0-9]'))

returns

['((', 'H2)(C(H3)))']
['((', 'H1)(Cl1))']
['((', '(C(H3))3))']

edited May 23 '17 at 12:14

Community

1
1

answered Jan 11 '14 at 04:47

Jonathan Abbott

180
1
10

So what do I do if I don't want to have that right parentheses included? If I wanted something like `['((', 'H2)(C(H3)','))']`? Would I just add a line of code that would split that string into two and then append it to the list? – Dan Oberlam Jan 11 '14 at 07:48
@Dannnno Adding a line of code is certainly a reasonable hack, but it might not be simplest / fastest. I can't offer now a better soln. – Jonathan Abbott Jan 11 '14 at 14:06
1

If you don't want the right parenthesis included, you should not make it part of the Word expression. The problem is that Word does not know that the 2 trailing ')'s are special, so it includes them in with the rest of the word group. Some suggestions: check out pyparsing `nestedExpr`. Also, I think `pyp.alphanums` is easier to read than `pyp.srange('[A-Za-z0-9')`. – PaulMcG Jan 12 '14 at 00:24

score 1 · Answer 2 · answered Jan 12 '14 at 23:12

As suggested in the comments by Paul McGuire I found that using nestedExpr was the best choice for my situation. Using the following code

import pyparsing as pyp

aString = "C((H2)(C(H3))) C((H1)(Cl1)) C((C(H3))3)"
aList = aString.split()

for i in range(len(aList)):
    aList[i] = [pyp.nestedExpr().parseString(aList[i][1:]).asList()[0]]

print aList

I got an output of

[[[['H2'], ['C', ['H3']]]], [[['H1'], ['Cl1']]], [[['C', ['H3']], '3']]]

Which is exactly what I wanted.

Not finding the strings expected with pyparsing

2 Answers2

Linked