NLTK python variable in CFG

Question

This is my current NLTK python grammar:

cfg_2 = CFG.fromstring("""
S -> ADVP VP P
S -> VP NP P
S -> VP P
ADVP -> RB
VP -> VB NP
VP -> VB PP
VP -> VB
VP -> VP CC ADVP VP
NP -> DT NML
VP -> VB NP PP
VP -> VB CC VB
NP -> NP A NP CC NP
NP -> DT NN
NP -> NN
NP -> PRP
NP -> NML NN
PP -> IN NP
NML -> NN CC NN
NML -> CD 
  DT -> 'the'|'a'
  NN -> 'carrots'|'celery'|'courgette'|'garlic'|'onion'|'leek'|'oregano'|'vegetables'|'bowl'|'ends'|'slices'|'beans'|'cannellini'|'water'|'potato'
  VB -> 'chop'|'trim'|'peel'|'add'|'cut'|'wash'|'quarter'|'scrub'|'dice'|'drain'
  CC -> 'and'
  RB -> 'roughly'|'finely'|'lengthways'|'now'|'then'
  JJ -> 'large'
  A -> ','
  P -> '.'
  IN -> 'off'|'under'|'into'|'to'
  CD -> '2'
  PRP -> 'it'
""".format(noun_string))

As you can see, my list of NN, nouns, is quite long. Therefore, I would like to put all my nouns in a list and then add that list to the CFG. I've tried something like this, but this doesn't work because the CFG itself is a string:

noun_list = ['carrots','celery','courgette'
             ,'garlic','onion','leek','oregano',
             'vegetables','bowl','ends','slices','beans','cannellini','water','potato']

noun_string = "'|'".join(noun_list)
print(noun_string)

cfg_2 = CFG.fromstring("""
S -> ADVP VP P
S -> VP NP P
S -> VP P
ADVP -> RB
VP -> VB NP
VP -> VB PP
VP -> VB
VP -> VP CC ADVP VP
NP -> DT NML
VP -> VB NP PP
VP -> VB CC VB
NP -> NP A NP CC NP
NP -> DT NN
NP -> NN
NP -> PRP
NP -> NML NN
PP -> IN NP
NML -> NN CC NN
NML -> CD 
  DT -> 'the'|'a'
  NN -> {}
  VB -> 'chop'|'trim'|'peel'|'add'|'cut'|'wash'|'quarter'|'scrub'|'dice'|'drain'
  CC -> 'and'
  RB -> 'roughly'|'finely'|'lengthways'|'now'|'then'
  JJ -> 'large'
  A -> ','
  P -> '.'
  IN -> 'off'|'under'|'into'|'to'
  CD -> '2'
  PRP -> 'it'
""".format(noun_string))

Is there any other way to add a list of nouns to my CFG?

Answer 1

Put something like <noun_list_goes_here> in your big string.

Then do:

noun_string = "'|'".join(noun_list)
big_string = big_string.replace('<noun_list_goes_here>', f"'{noun_string}'")`