[qi]parsing of recursivly quoted text

[Anton Potapov]

Hi all,
Is it possible to parse the the recursively quoted text using spirit.qi ?

I.e.  make following test pass:

        non_quote %= lexeme[*( (char_ - '\"')  ) >>  -(skip[qouted_word] )];
        qouted_word   %=   '\"'>> skip[non_quote]>>'\"' ;

        {
                std::string msg = "\"\"b\"\"";
                std::string result;
                std::string::iterator it = msg.begin(); BOOST_CHECK(qi::phrase_parse(it,msg.end(),qouted_word,charset::space,result));
                BOOST_CHECK_EQUAL(result,"\"b\"");
                BOOST_CHECK_EQUAL(std::string(it,msg.end()),"");
        }

What i missed ?

The full text is attached.

Regards, Anton


------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Spirit-general mailing list
Spirit-general@...
https://lists.sourceforge.net/lists/listinfo/spirit-general

[OvermindDL1]

2009/11/5 Anton Potapov <balanco@...>:

> Hi all,
> Is it possible to parse the the recursively quoted text using spirit.qi ?
>
> I.e.  make following test pass:
>
>        non_quote       %= lexeme[*( (char_ - '\"')  ) >>  -(skip[qouted_word] )];
>        qouted_word     %=   '\"'>> skip[non_quote]>>'\"' ;
>
>        {
>                std::string msg = "\"\"b\"\"";
>                std::string result;
>                std::string::iterator it = msg.begin();          BOOST_CHECK(qi::phrase_parse(it,msg.end(),qouted_word,charset::space,result));
>                BOOST_CHECK_EQUAL(result,"\"b\"");
>                BOOST_CHECK_EQUAL(std::string(it,msg.end()),"");
>        }
>
> What i missed ?

That is easy to do, but the question is, how would you know when the
string starts and stops then?  If you know your whole text is a string
and you are just stripping quotes there are much easier ways to do
that, otherwise, your grammar would become... interesting.  That is
why C/C++/Python/Lua/Java/etc.../everything has you quote out a quote,
like with most \" puts in a literal quote instead of a delimiter.
Would be better to have more information on what you are actually
trying to do then.

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Spirit-general mailing list
Spirit-general@...
https://lists.sourceforge.net/lists/listinfo/spirit-general

[Anton Potapov]

-----Original Message-----
From: OvermindDL1 <overminddl1@...>
To: Anton Potapov <balanco@...>,
Spirit General Mailing List <spirit-general@...>
Date: Thu, 5 Nov 2009 05:55:01 -0700
Subject: Re: [Spirit-general] [qi]parsing of recursivly quoted text

> 2009/11/5 Anton Potapov <balanco@...>:
> > Hi all,
> > Is it possible to parse the the recursively quoted text using spirit.qi ?
> >
> > I.e. ?make following test pass:
> >
> > ? ? ? ?non_quote ? ? ? %= lexeme[*( (char_ - '\"') ?) >> ?-(skip[qouted_word] )];
> > ? ? ? ?qouted_word ? ? %= ? '\"'>> skip[non_quote]>>'\"' ;
> >
> > ? ? ? ?{
> > ? ? ? ? ? ? ? ?std::string msg = "\"\"b\"\"";
> > ? ? ? ? ? ? ? ?std::string result;
> > ? ? ? ? ? ? ? ?std::string::iterator it = msg.begin(); ? ? ? ? ?BOOST_CHECK(qi::phrase_parse(it,msg.end(),qouted_word,charset::space,result));
> > ? ? ? ? ? ? ? ?BOOST_CHECK_EQUAL(result,"\"b\"");
> > ? ? ? ? ? ? ? ?BOOST_CHECK_EQUAL(std::string(it,msg.end()),"");
> > ? ? ? ?}
> >
> > What i missed ?
>
> That is easy to do, but the question is, how would you know when the
> string starts and stops then?  If you know your whole text is a string
> and you are just stripping quotes there are much easier ways to do
> that,

the text I'm trying to parse is *kind* of file path, and things that know upfront is that:
1. the full text begins and end with a quote (").
2. if there are quotes between then they are paired.
eg.

> otherwise, your grammar would become... interesting.  That is
> why C/C++/Python/Lua/Java/etc.../everything has you quote out a quote,
> like with most \" puts in a literal quote instead of a delimiter.
> Would be better to have more information on what you are actually
> trying to do then.
>

actually I'm trying to parse third party tool output. Which looks like that:
name="c:\exchange\test.zip " "files.txt"", info=""

Regards, Anton


------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Spirit-general mailing list
Spirit-general@...
https://lists.sourceforge.net/lists/listinfo/spirit-general

[Dennison, Ted]

Anton Potapov
> the text I'm trying to parse is *kind* of file path, and things that
know
> upfront is that:
> 1. the full text begins and end with a quote (").
> 2. if there are quotes between then they are paired.
> eg.

Ahh. So it isn't really recursive...they can only be nested one deep.

Still, I don't see how you (even as a human being) would be able to tell
the difference between a string nested in another string and a line with
two disjoint strings on it.


> actually I'm trying to parse third party tool output. Which looks like
> that:
> name="c:\exchange\test.zip " "files.txt"", info=""

OK. This doesn't make sense at all. You have an odd number of quotes up
there, so something isn't matched. Perhaps it should have started:
name=""...?

Assuming that's what you meant, you can probably work with a grammar
that assumes a production like

LINE ::= IDENTIFIER '=' '"' VALUE '"' EOL

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Spirit-general mailing list
Spirit-general@...
https://lists.sourceforge.net/lists/listinfo/spirit-general

[Anton Potapov]

-----Original Message-----
From: "Dennison, Ted" <Ted.Dennison@...>
To: "Anton Potapov" <balanco@...>,
"Spirit General Mailing List" <spirit-general@...>
Date: Thu, 5 Nov 2009 08:43:32 -0600
Subject: RE: [Spirit-general][qi]parsing of recursivly quoted text

> Anton Potapov
> > the text I'm trying to parse is *kind* of file path, and things that
> know
> > upfront is that:
> > 1. the full text begins and end with a quote (").
> > 2. if there are quotes between then they are paired.
> > eg.
>
> Ahh. So it isn't really recursive...they can only be nested one deep.
>
> Still, I don't see how you (even as a human being) would be able to tell
> the difference between a string nested in another string and a line with
> two disjoint strings on it.

In my case there is an assumption that text is *always* string nested in string.
 
>
> > actually I'm trying to parse third party tool output. Which looks like
> > that:
> > name="c:\exchange\test.zip " "files.txt"", info=""
>
> OK. This doesn't make sense at all. You have an odd number of quotes up
> there, so something isn't matched. Perhaps it should have started:
> name=""...?
Sorry, looks like typo :). Intended string was :  
  name="c:\exchange\test.zip > "files.txt"", info=""
>
> Assuming that's what you meant, you can probably work with a grammar
> that assumes a production like
>
> LINE ::= IDENTIFIER '=' '"' VALUE '"' EOL
not quite right. I would say something like :
LINE ::= IDENTIFIER '=' '"' VALUE '"' ',' IDENTIFIER '=' '"' VALUE '"'  EOL
and VALUE can have quoted strings inside.

regards, anton

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Spirit-general mailing list
Spirit-general@...
https://lists.sourceforge.net/lists/listinfo/spirit-general