CODEWITHSUNDEEP
pythonpython-2.7parsingtext-parsingpyparsing
XYZ
XYZ

Reputation: 191

preserving text structure information - pyparsing

Using pyparsing, is there a way to extract the context you are in during recursive descent. Let me explain what I mean. I have the following code:

import pyparsing as pp

openBrace = pp.Suppress(pp.Literal("{"))
closeBrace = pp.Suppress(pp.Literal("}"))
ident = pp.Word(pp.alphanums + "_" + ".")
comment = pp.Literal("//") + pp.restOfLine
messageName = ident
messageKw = pp.Suppress(pp.Keyword("msg"))
text = pp.Word(pp.alphanums + "_" + "." + "-" + "+")
otherText = ~messageKw + pp.Suppress(text)
messageExpr = pp.Forward()
messageExpr << (messageKw + messageName + openBrace +
                pp.ZeroOrMore(otherText) + pp.ZeroOrMore(messageExpr) +
                pp.ZeroOrMore(otherText) + closeBrace).ignore(comment)
testStr = "msg msgName1 { some text msg msgName2 { some text } some text }"
print messageExpr.parseString(testStr)

which produces this output: ['msgName1', 'msgName2']

In the output, I would like to keep track of the structure of embedded matches. What I mean is that, for example, I would like the following output with the test string above: ['msgName1', 'msgName1.msgName2'] to keep track of the hierarchy in the text. However, I am new to pyparsing and have yet to find a way yet to extract the fact that "msgName2" is embedded in the structure of "msgName1."

Is there a way to use the setParseAction() method of ParserElement to do this, or maybe using naming of results?

Helpful advice would be appreciated.

Upvotes: 4

Views: 220

Answers (1)

XYZ
XYZ

Reputation: 191

Thanks to Paul McGuire for his sagely advice. Here are the additions/changes I made, which solved the problem:

msgNameStack = []

def pushMsgName(str, loc, tokens):
    msgNameStack.append(tokens[0])
    tokens[0] = '.'.join(msgNameStack)

def popMsgName(str, loc, tokens):
    msgNameStack.pop()

closeBrace = pp.Suppress(pp.Literal("}")).setParseAction(popMsgName)
messageName = ident.setParseAction(pushMsgName)

And here is the complete code:

import pyparsing as pp

msgNameStack = []


def pushMsgName(str, loc, tokens):
    msgNameStack.append(tokens[0])
    tokens[0] = '.'.join(msgNameStack)


def popMsgName(str, loc, tokens):
    msgNameStack.pop()

openBrace = pp.Suppress(pp.Literal("{"))
closeBrace = pp.Suppress(pp.Literal("}")).setParseAction(popMsgName)
ident = pp.Word(pp.alphanums + "_" + ".")
comment = pp.Literal("//") + pp.restOfLine
messageName = ident.setParseAction(pushMsgName)
messageKw = pp.Suppress(pp.Keyword("msg"))
text = pp.Word(pp.alphanums + "_" + "." + "-" + "+")
otherText = ~messageKw + pp.Suppress(text)
messageExpr = pp.Forward()
messageExpr << (messageKw + messageName + openBrace +
                pp.ZeroOrMore(otherText) + pp.ZeroOrMore(messageExpr) +
                pp.ZeroOrMore(otherText) + closeBrace).ignore(comment)

testStr = "msg msgName1 { some text msg msgName2 { some text } some text }"
print messageExpr.parseString(testStr)

Upvotes: 2

Related Questions