Data::Validate::Struct - Validate recursive Hash Structures

 use Data::Validate::Struct;
 my $validator = new Data::Validate::Struct($reference);
 if ( $validator->validate($config_hash_reference) ) {
   print "valid\n";
 }
 else {
   print "invalid " . $validator->errstr() . "\n";
 }

This module validates a config hash reference against a given hash structure in contrast to Data::Validate in which you have to check each value separately using certain methods.

This hash could be the result of a config parser or just any hash structure. Eg. the hash returned by XML::Simple could be validated using this module. You may also use it to validate CGI input, just fetch the input data from CGI, map it to a hash and validate it.

Data::Validate::Struct uses some of the methods exported by Data::Validate, so you need to install it too.

int

Match a simple integer number.

hex

Match a hex value.

oct

Match an octagonal value.

number

Match a decimal number, it may contain , or . and may be signed.

word

Match a single word, _ and - are tolerated.

line

Match a line of text - no newlines are allowed.

text

Match a whole text(blob) including newlines. This expression is very loosy, consider it as an alias to any.

regex

Match a perl regex using the operator qr(). Valid examples include:

 qr/[0-9]+/
 qr([^%]*)
 qr{\w+(\d+?)}

Please note, that this doesn't mean you can provide here a regex against config options must match.

Instead this means that the config options contains a regex.

eg:

 <cfg>
   grp  = qr/root|wheel/
 </cfg>

regex would match the content of the variable 'grp' in this example.

To add your own rules for validation, use the type() method, see below.

uri

Match an internet URI.

ipv4

Match an IPv4 address.

cidrv4

The same as above including cidr netmask (/24), IPv4 only, eg:

 10.2.123.0/23

Note: shortcuts are not supported for the moment, eg:

 10.10/16

will fail while it is still a valid IPv4 cidr notation for a network address (short for 10.10.0.0/16). Must be fixed in Regex::Common.

ipv6

Match an IPv6 address. Some examples:

 3ffe:1900:4545:3:200:f8ff:fe21:67cf
 fe80:0:0:0:200:f8ff:fe21:67cf
 fe80::200:f8ff:fe21:67cf
 ff02:0:0:0:0:0:0:1
 ff02::1

quoted

Match a text quoted with single quotes, eg:

 'barbara is sexy'

hostname

Match a valid hostname, it must qualify to the definitions in RFC 2396.

resolvablehost

Match a hostname resolvable via dns lookup. Will fail if no dns is available at runtime.

path

Match a valid absolute path, it won't do a stat() system call. This will work on any operating system at runtime. So this one:

 C:\Temp

will return TRUE if running on WIN32, but FALSE on FreeBSD!

fileexists

Look if value is a file which exists. Does a stat() system call.

user

Looks if the given value is an existent user. Does a getpwnam() system call.

group

Looks if the given value is an existent group. Does a getgrnam() system call.

port

Match a valid tcp/udp port. Must be a digit between 0 and 65535.

vars

Matches a string of text containing variables (perl style variables though) eg:

 $user is $attribute
 I am $(years) old
 Missing ${points} points to succeed

In some rare situations you might require a negative match. So a test shall return TRUE if a particular value does NOT match the given type. This might be usefull to prevent certain things.

To achieve this, you just have to prepend one of the below mentioned types with the keyword no.

Example:

 $ref = { path => 'novars' }

This returns TRUE if the value of the given config hash does NOT contain ANY variables.

The expected structure must be a standard perl hash reference. This hash may look like the config you are validating but instead of real-live values it contains types that define of what type a given value has to be.

In addition the hash may be deeply nested. In this case the validated config must be nested the same way as the reference hash.

Example:

 $reference = { user => 'word', uid => 'int' };

The following config would be validated successful:

 $config = { user => 'HansDampf',  uid => 92 };

this one not:

 $config = { user => 'Hans Dampf', uid => 'nine' };
                          ^                ^^^^
                          |                |
                          |                +----- is not a number
                          +---------------------- space not allowed

For easier writing of references you yould use a configuration file parser like Config::General or Config::Any, just write the definition using the syntax of such a module, get the hash of it and use this hash as validation reference.

You can also match against nested structures. Data::Validate::Struct iterates into the given config hash the same way as the reference hash looks like.

If the config hash doesn't match the reference structure, perl will throw an error, which Data::Validate::Struct catches and returns FALSE.

Given the following reference hash:

 $ref = {
 'b1' => {
          'b2' => {
                   'b3' => {
                            'item' => 'int'
                           }
                  }
          }
 }

Now if you validate it against the following config hash it will return TRUE:

 $cfg = {
 'b1' => {
	  'b2' => {
		   'b3' => {
			    'item' => '100'
			   }
		  }
	 }
 }

If you validate it for example against this hash, it will return FALSE:

 $cfg = {
 'b1' => {
	  'b2' => {
		    'item' => '100'
		   }
	 }
 }

validate($config)

$config must be a hash reference you'd like to validate.

It returns a true value if the given structure looks valid.

If the return value is false (0), then the error message will be written to the variable $!.

type(%types)

You can enhance the validator by adding your own rules. Just add one or more new types using a simple hash using the type() method. Values in this hash can be regexes or anonymous subs.

Example:

 $v3->type(
  (
  address => qr(^\w+\s\s*\d+$),
  list    =>
    sub {
      my $list = $_[0];
      my @list = split /\s*,\s*/, $list;
      if (scalar @list > 1) {
	return 1;
      }
      else {
	return 0;
      }
    }
  )
 );

In this example we add 2 new types, 'list' and 'address', which are really simple. 'address' is a regex which matches a word followed by an integer. 'list' is a subroutine which gets called during evaluation for each option which you define as type 'list'.

Such subroutines must return a true value in order to produce a match.

You can also add reversive types, which are like all other types but start with no. The validator does a negative match in such a case, thus you will have a match if a variable does not match the type. The builtin type 'novars' is such a type.

Regexes will be executed exactly as given. No flags or ^ or $ will be used by the module. Eg. if you want to match the whole value from beginning to the end, add ^ and $, like you can see in our 'address' example above.

debug()

Enables debug output which gets printed to STDERR.

errstr()

Returns the last error, which is useful to notify the user about what happened.

Arrays must be handled in a special way, just define an array with two elements and the second empty. The config will only validated against the first element in the array.

We assume all elements in an array must have the same structure.

Example for array of hashes:

 $reference = {
                [
                  {
                     user => 'word',
                     uid => 'int'
                  },
                  {} # empty 2nd element
                ]
               };

Example of array of values:

 $reference = {
  var => [ 'int', '' ]
 }

Take a look to t/run.t for lots of examples.

No environment variables will be used.

I recommend you to read the following documentations, which are supplied with perl:

perlreftut Perl references short introduction.

perlref Perl references, the rest of the story.

perldsc Perl data structures intro.

perllol Perl data structures: arrays of arrays.

Data::Validate common data validation methods.

Copyright (c) 2007 Thomas Linden

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

Some implementation details as well as the API may change in the future. This will no more happen if entering a stable release (starting with 1.00).

To submit use http://rt.cpan.org.

None known.

To debug Data::Validate::Struct use debug() or the perl debugger, see perldebug.

For example to debug the regex matching during processing try this:

 perl -Mre=debug yourscript.pl

Data::Validate::Struct depends on the module Data::Validate, Regexp::Common, File::Spec and File::stat.

• Add support for type mixes, such as:
  
  

   { name => 'int|number' }
  
  

• Add support for ranges, in fact Regexp::Common or Data::Validate already supports this, but Data::Validate::Struct currently doesn't support parameters for types.
• Perhaps add code validation too, for example we could have a type 'perl' which tries to evaluate the given value. On the other side this may lead to security holes - so I might never do it.
• Plugin System
• Possibly add support for grammars. This might be much more powerful than regular expressions, say:
  
  

   { name => 'expr OP expr | expr' }
  
  

  
  
  or something like this.

Thomas Linden <tlinden |AT| cpan.org>

Thanks to David Cantrell for his helpful hints.

0.06