comment on

The story

I am working on a prototype DSL and I have crafted a small-ish converter that reads us validated XML and throws up transformed SQL code based on the source markup.

You can find the entire code for reference at the LUARM/ITPSL SVN repository

A typical XML input markup is given below:

<?xml version="1.0"?>
<itpslsig xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
    <itpslheader>
        <signid> 5938724b6b41a834ac695529dd104ed0 </signid>
            <signdate>
                      <year>2010</year>
                      <month>12</month>
                      <day>20</day>
                </signdate>
            <ontology>
            <reason>intentional</reason>
                  <revision>1.0</revision>
                  <user_role>ordinary_users</user_role>
                  <detectby>multi</detectby>
                  <multihost>no</multihost>
                  <hostlist>proteas,dionisos,slart,cn1,panoptis</hostl
+ist>
                  <weightmatrix>3,10,20,70</weightmatrix>
                  <os>linux</os>
                  <osver>2.6</osver>
                  <keywords>DoS software install DoS loiq </keywords>
                  <synopsis> This signature predicts the usage of the 
+Low Orbit Ion Cannon tool for DDoS attacks.
            </synopsis>
            </ontology>
    </itpslheader>
    <itpslbody>
        <mainblock>
            <mainop>as_a_result_of</mainop>
                <subblock>
                    <subop>AND</subop>
                    <fileexists>
                        <filename>loiq</filename>
                        <type>executable</type>
                        <location>OR (#userhome#/*,/site/*,/tmp/*,/tem
+p/*)</location>
                        <singlefile>yes</singlefile>
                        <ownedbyuser>johnc</ownedbyuser>
                    </fileexists>
                    <fileexists>
                        <filename>loiq.pro</filename>
                        <type>textdata</type>
                        <location>OR(#userhome#/*,/site/*,/tmp/*,/temp
+/*)</location>
                        <ownedbyuser>johnc</ownedbyuser>
                        <singlefile>yes</singlefile>
                    </fileexists>
                    <fileexists>
                        <filename>loiq.qrc</filename>
                        <type>textdata</type>
                        <location>OR(#userhome#/*,site/*,/tmp/*,/temp/
+*)</location>
                        <singlefile>yes</singlefile>
                        <ownedbyuser>johnc</ownedbyuser>
                    </fileexists>                            
                </subblock>
                <subblock>
                    <subop>single</subop>
                    <userexec>
                        <username>johnc</username>
                        <name>OR (file-roller,tar,bunzip2)</name>
                        <path>OR(/usr/bin/,/usr/local/bin)</path>
                        <singleprocess>yes</singleprocess>
                        <argumentlist>loiq*.bz2</argumentlist>
                        <pattern>any</pattern>
                    </userexec>
                </subblock>
                <subblock>
                    <subop>single</subop>
                    <fileexists>
                        <filename>*</filename>
                        <type>any</type>
                        <location>OR (#userhome#/.mozilla/*,#userhome#
+/.opera)</location>
                        <singlefile>yes</singlefile>
                        <withcontents>
                            <stringsearch>"http://sourceforge.net/proj
+ects/loiq"</stringsearch>
                        </withcontents>    
                        <ownedbyuser>johnc</ownedbyuser>
                    </fileexists>    
                </subblock>
        </mainblock>
    </itpslbody>
</itpslsig>
[download]

In short, the program creates an XML::Twig::XPath-ed structure with handlers. The problems start with calling the 'parsesubs' subroutine:

my $twig = new XML::Twig::XPath( TwigHandlers => {
    #ITPSL header parsing data
    "/itpslsig/itpslheader/ontology/weightmatrix" => \&getwm, 
    "/itpslsig/itpslheader/ontology/detectby" => \&getdetectmethods,
    "/itpslsig/itpslheader/ontology/os" => \&getos,
    "/itpslsig/itpslheader/ontology/osver" => \&getosver,
    #ITPSL body parsing data
    "/itpslsig/itpslbody/mainblock/mainop" => \&getmainop,
    "/itpslsig/itpslbody/mainblock" => \&getnoofsubblocks,
    "/itpslsig/itpslbody/mainblock/subblock" => \&parsesubs,
    
});
# parse, handling nodes on the way
$twig->parsefile( shift @ARGV );
[download]

Part of the 'parsesubs' subroutine calls via if statements other subroutines to localize the parsing of specific markup directives. The anonymous array reference ($sblockstack) is my crude way of having a stack of parsed directives (an array of arrays, each row representing the parsed directives per sub-block).
The problem
When I execute the code on a Linux based Perl 5.12 system, what I get (excluding harmless warnings and with the help of Data::Dumper) is the following:

Called parsesubs 
Directive is fileexists 
Pushing to sblockstack [0] [0] 
Directive is fileexists 
Pushing to sblockstack [0] [1] 
Directive is fileexists 
Pushing to sblockstack [0] [2] 
Called parsesubs 
Directive is userexec 
Pushing to sblockstack [1] [0] 
Called parsesubs 
Directive is fileexists 
Pushing to sblockstack [2] [0] 
$VAR1 = [
          [
            '##STARTOP:fileexists',
            '##operand:type:executable',
            '##operand:location:OR (#userhome#/*,/site/*,/tmp/*,/temp/
+*)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq.pro',
            '##operand:type:textdata',
            '##operand:location:OR(#userhome#/*,/site/*,/tmp/*,/temp/*
+)',
            '##operand:ownedbyuser:johnc',
            '##operand:singlefile:yes',
            '##operand:filename:loiq.qrc',
            '##operand:type:textdata',
            '##operand:location:OR(#userhome#/*,site/*,/tmp/*,/temp/*)
+',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##ENDOFOP##',
            '##operand:filename:loiq',
            '##operand:type:executable',
            '##operand:location:OR (#userhome#/*,/site/*,/tmp/*,/temp/
+*)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq.pro',
            '##operand:type:textdata',
            '##operand:location:OR(#userhome#/*,/site/*,/tmp/*,/temp/*
+)',
            '##operand:ownedbyuser:johnc',
            '##operand:singlefile:yes',
            '##operand:filename:loiq.qrc',
            '##operand:type:textdata',
            '##operand:location:OR(#userhome#/*,site/*,/tmp/*,/temp/*)
+',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq',
            '##operand:type:executable',
            '##operand:location:OR (#userhome#/*,/site/*,/tmp/*,/temp/
+*)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq.pro',
            '##operand:type:textdata',
            '##operand:location:OR(#userhome#/*,/site/*,/tmp/*,/temp/*
+)',
            '##operand:ownedbyuser:johnc',
            '##operand:singlefile:yes',
            '##operand:filename:loiq.qrc',
            '##operand:type:textdata',
            '##operand:location:OR(#userhome#/*,site/*,/tmp/*,/temp/*)
+',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq',
            '##operand:type:executable',
            '##operand:location:OR (#userhome#/*,/site/*,/tmp/*,/temp/
+*)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq.pro',
            '##operand:type:textdata',
            '##operand:location:OR(#userhome#/*,/site/*,/tmp/*,/temp/*
+)',
            '##operand:ownedbyuser:johnc',
            '##operand:singlefile:yes',
            '##operand:filename:loiq.qrc',
            '##operand:type:textdata',
            '##operand:location:OR(#userhome#/*,site/*,/tmp/*,/temp/*)
+',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:*',
            '##operand:type:any',
            '##operand:location:OR (#userhome#/.mozilla/*,#userhome#/.
+opera)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc'
          ],
          [
            '##STARTOP:fileexists',
            '##ENDOFOP##'
          ],
          [
            '##STARTOP:fileexists',
            '##ENDOFOP##'
          ]
        ];
$VAR2 = [
          [
            '##STARTOP:userexec',
            '##operand:name:OR (file-roller,tar,bunzip2)',
            '##operand:path:OR(/usr/bin/,/usr/local/bin)',
            '##operand:singleprocess:yes',
            '##operand:argumentlist:loiq*.bz2',
            '##operandpattern:any',
            '##ENDOFOP##'
          ]
        ];
$VAR3 = [
          [
            '##STARTOP:fileexists',
            '##ENDOFOP##'
          ]
        ];
[download]

Obviously the Data::Dumper output tells me that I have a problem in parsing the whole thing properly. Each directive is not terminated properly ( '##ENDOFOP##' string) and certain directives are not entered into their proper place on the array of arrays. A proper output I would expect/want to have by Data::Dumper would be like the following:

...
Directive is fileexists 
Pushing to sblockstack [2] [0] 
$VAR1 = [
          [
            '##STARTOP:fileexists',
            '##operand:type:executable',
            '##operand:location:OR (#userhome#/*,/site/*,/tmp/*,/temp/
+*)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq',
            '##operand:type:textdata',
            '##ENDOFOP##'
          ],

          [
            '##STARTOP:fileexists',
            '##operand:type:executable',
            '##operand:location:OR (#userhome#/*,/site/*,/tmp/*,/temp/
+*)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq.pro',
            '##operand:type:textdata',
            '##ENDOFOP##'
          ],

          [
            '##STARTOP:fileexists',
            '##operand:type:executable',
            '##operand:location:OR (#userhome#/*,/site/*,/tmp/*,/temp/
+*)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:loiq.qrc',
            '##operand:type:textdata',
            '##ENDOFOP##'

          ]
       
        ];

$VAR2 = [
          [
            '##STARTOP:userexec',
            '##operand:name:OR (file-roller,tar,bunzip2)',
            '##operand:path:OR(/usr/bin/,/usr/local/bin)',
            '##operand:singleprocess:yes',
            '##operand:argumentlist:loiq*.bz2',
            '##operandpattern:any',
            '##ENDOFOP##'
          ]
        ];

$VAR3 = [
          [  
            '##STARTOP:fileexists',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc',
            '##operand:filename:*',
            '##operand:type:any',
            '##operand:location:OR (#userhome#/.mozilla/*,#userhome#/.
+opera)',
            '##operand:singlefile:yes',
            '##operand:ownedbyuser:johnc'
           ]
         ];
[download]

The Question
What am I doing wrong with this program structure? Can somebody suggest a structure that works?
Many thanks!

In reply to XML::Twig blues by gmagklaras

Posts are HTML formatted. Put <p> </p> tags around your paragraphs. Put <code> </code> tags around your code and data!

Titles consisting of a single word are discouraged, and in most cases are disallowed outright.

Read Where should I post X? if you're not absolutely sure you're posting in the right place.

Please read these before you post! —

Posts may use any of the Perl Monks Approved HTML tags:

a, abbr, b, big, blockquote, br, caption, center, col, colgroup, dd, del, details, div, dl, dt, em, font, h1, h2, h3, h4, h5, h6, hr, i, ins, li, ol, p, pre, readmore, small, span, spoiler, strike, strong, sub, summary, sup, table, tbody, td, tfoot, th, thead, tr, tt, u, ul, wbr

You may need to use entities for some characters, as follows. (Exception: Within code tags, you can put the characters literally.)

	For:		Use:
	&		`&`
	<		`<`
	>		`>`
	[		`[`
	]		`]`

Link using PerlMonks shortcuts! What shortcuts can I use for linking?

See Writeup Formatting Tips and other pages linked from there for more info.