python - When using setResultsName with listAllMatches true, some matched items are nested -

based on grammar:

from pyparsing import *  g = quotedstring.setparseaction( removequotes ) eg = suppress('-') + quotedstring.setparseaction( removequotes ) choice = or( [ g.setresultsname("out",listallmatches=true),                    eg.setresultsname("in",listallmatches=true) ] ) grammar = zeroormore( choice ) + suppress(restofline)  = world.parsestring( ' "ali"  -"baba"  "holy cow"   -"smoking beaute"  ' ) print a.dump() 

i have discovered tokens satisfy eg nonterminal wrapped in list. difference g has leading `suppress('-')'.

['ali', 'baba', 'holy cow', 'smoking beaute'] - in: [['baba'], ['smoking beaute']] - out: ['ali', 'holy cow'] 

how make them behave same ? want achieve result below:

['ali', 'baba', 'holy cow', 'smoking beaute'] - in: ['baba', 'smoking beaute'] - out: ['ali', 'holy cow'] 

it's been while since i've looked @ issue - problem and's return tokens lists, if contain single value.

here ungrouper can clear you, i'll include in next pyparsing release:

ungroup = lambda expr : tokenconverter(expr).setparseaction(lambda t:t[0]) eg = ungroup(suppress('-') + quotedstring.setparseaction( removequotes )) 

with test code, these results:

['ali', 'baba', 'holy cow', 'smoking beaute'] - in: ['baba', 'smoking beaute'] - out: ['ali', 'holy cow']