#!/usr/bin/perl
use warnings;
use strict;

use HTML::Template;

# oeffne das html-Template
my $template = HTML::Template->new(filename => 'test.tmpl'); # Fehlerabfrage: siehe unten

# Parameter einfuellen
$template->param(HOME => $ENV{HOME});
$template->param(PATH => $ENV{PATH});

# den nötigen Content-Type senden und dann das gefuellte Template ausgeben
print "Content-Type: text/html\n\n";
print $template->output;

Mein Home-Verzeichnis ist /home/some/directory
Mein PATH ist auf /bin;/usr/bin gesetzt.

$template->param(
  EMPLOYEE_INFO => [
    { name => 'Sam', job => 'Programmierer' },
    { name => 'Steve', job => 'Administrator' },
  ]
);
print $template->output();

# ein paar Listen von Daten, die in die Schleife gepackt werden sollen:
my @words = qw(I Am Cool);
my @numbers = qw(1 2 3);

my @loop_data = ();  # Initialisiere eine Liste, die die Schleifendaten aufnehmen soll

while (@words and @numbers) {
  my %row_data;  # verwende einen neuen Hash fuer die Daten der Zeile
  # Fuelle die Zeile
  $row_data{WORD} = shift @words;
  $row_data{NUMBER} = shift @numbers;

  # Der kritische Punkt: werfe eine Referenz auf diese Zeile
  # auf die Schleife
  push(@loop_data, \%row_data);
}

# weise die Schleifendaten dem Schleifenparameter als Referenz zu
$template->param(THIS_LOOP => \@loop_data);

$template->param(
  LOOP => [
    { name => 'Bobby',
      nicknames => [
        { name => 'the big bad wolf' },
        { name => 'He-Man' },
      ],
    },
  ],
);

my &#36template = HTML::Template->new( filename => 'file.tmpl',
               
               
            option => 'value'
               
               
          );

      my &#36t = HTML::Template->new( scalarref => &#36ref_to_template_text,
               
               
     option => 'value'
               
               
   );

      my &#36t = HTML::Template->new( arrayref => &#36ref_to_array_of_lines ,
               
               
     option => 'value'
               
               
   );

my &#36t = HTML::Template->new( filehandle => *FH, option => 'value');

my &#36t = HTML::Template->new_file('file.tmpl', option => 'value');

my &#36t = HTML::Template->new_scalar_ref(&#36ref_to_template_text,
               
               
        option => 'value');

my &#36t = HTML::Template->new_array_ref(&#36ref_to_array_of_lines,
               
               
       option => 'value');

my &#36t = HTML::Template->new_filehandle(&#36fh,
               
               
       option => 'value');

my &#36t = HTML::Template->new(type => 'filename',
               
             source => 'file.tmpl');

&#36template->param(parameterName => 'wert')

    Caching Options

        *   cache - if set to 1 the module will cache in memory the parsed
            templates based on the filename parameter and modification date
            of the file. This only applies to templates opened with the
            filename parameter specified, not scalarref or arrayref
            templates. Caching also looks at the modification times of any
            files included using <TMPL_INCLUDE> tags, but again, only if the
            template is opened with filename parameter.

            This is mainly of use in a persistent environment like
            Apache/mod_perl. It has absolutely no benefit in a normal CGI
            environment since the script is unloaded from memory after every
            request. For a cache that does work for normal CGIs see the
            'shared_cache' option below.

            Note that different new() parameter settings do not cause a
            cache refresh, only a change in the modification time of the
            template will trigger a cache refresh. For most usages this is
            fine. My simplistic testing shows that using cache yields a 90%
            performance increase under mod_perl. Cache defaults to 0.

        *   shared_cache - if set to 1 the module will store its cache in
            shared memory using the IPC::SharedCache module (available from
            CPAN). The effect of this will be to maintain a single shared
            copy of each parsed template for all instances of HTML::Template
            to use. This can be a significant reduction in memory usage in a
            multiple server environment. As an example, on one of our
            systems we use 4MB of template cache and maintain 25 httpd
            processes - shared_cache results in saving almost 100MB! Of
            course, some reduction in speed versus normal caching is to be
            expected. Another difference between normal caching and
            shared_cache is that shared_cache will work in a CGI environment
            - normal caching is only useful in a persistent environment like
            Apache/mod_perl.

            By default HTML::Template uses the IPC key 'TMPL' as a shared
            root segment (0x4c504d54 in hex), but this can be changed by
            setting the 'ipc_key' new() parameter to another 4-character or
            integer key. Other options can be used to affect the shared
            memory cache correspond to IPC::SharedCache options - ipc_mode,
            ipc_segment_size and ipc_max_size. See IPC::SharedCache for a
            description of how these work - in most cases you shouldn't need
            to change them from the defaults.

            For more information about the shared memory cache system used
            by HTML::Template see IPC::SharedCache.

        *   double_cache - if set to 1 the module will use a combination of
            shared_cache and normal cache mode for the best possible
            caching. Of course, it also uses the most memory of all the
            cache modes. All the same ipc_* options that work with
            shared_cache apply to double_cache as well. By default
            double_cache is off.

        *   blind_cache - if set to 1 the module behaves exactly as with
            normal caching but does not check to see if the file has changed
            on each request. This option should be used with caution, but
            could be of use on high-load servers. My tests show blind_cache
            performing only 1 to 2 percent faster than cache under mod_perl.

            NOTE: Combining this option with shared_cache can result in
            stale templates stuck permanently in shared memory!

        *   file_cache - if set to 1 the module will store its cache in a
            file using the Storable module. It uses no additional memory,
            and my simplistic testing shows that it yields a 50% performance
            advantage. Like shared_cache, it will work in a CGI environment.
            Default is 0.

            If you set this option you must set the "file_cache_dir" option.
            See below for details.

            NOTE: Storable using flock() to ensure safe access to cache
            files. Using file_cache on a system or filesystem (NFS) without
            flock() support is dangerous.

        *   file_cache_dir - sets the directory where the module will store
            the cache files if file_cache is enabled. Your script will need
            write permissions to this directory. You'll also need to make
            sure the sufficient space is available to store the cache files.

        *   file_cache_dir_mode - sets the file mode for newly created
            file_cache directories and subdirectories. Defaults to 0700 for
            security but this may be inconvenient if you do not have access
            to the account running the webserver.

        *   double_file_cache - if set to 1 the module will use a
            combination of file_cache and normal cache mode for the best
            possible caching. The file_cache_* options that work with
            file_cache apply to double_file_cache as well. By default
            double_file_cache is 0.

    Filesystem Options

        *   path - you can set this variable with a list of paths to search
            for files specified with the "filename" option to new() and for
            files included with the <TMPL_INCLUDE> tag. This list is only
            consulted when the filename is relative. The HTML_TEMPLATE_ROOT
            environment variable is always tried first if it exists. Also,
            if HTML_TEMPLATE_ROOT is set then an attempt will be made to
            prepend HTML_TEMPLATE_ROOT onto paths in the path array. In the
            case of a <TMPL_INCLUDE> file, the path to the including file is
            also tried before path is consulted.

            Example:

               my $template = HTML::Template->new( filename => 'file.tmpl',
                                                   path => [ '/path/to/templates',
                                                             '/alternate/path'
                                                           ]
                                                  );

            NOTE: the paths in the path list must be expressed as UNIX
            paths, separated by the forward-slash character ('/').

        *   search_path_on_include - if set to a true value the module will
            search from the top of the array of paths specified by the path
            option on every <TMPL_INCLUDE> and use the first matching
            template found. The normal behavior is to look only in the
            current directory for a template to include. Defaults to 0.

    Debugging Options

        *   debug - if set to 1 the module will write random debugging
            information to STDERR. Defaults to 0.

        *   stack_debug - if set to 1 the module will use Data::Dumper to
            print out the contents of the parse_stack to STDERR. Defaults to
            0.

        *   cache_debug - if set to 1 the module will send information on
            cache loads, hits and misses to STDERR. Defaults to 0.

        *   shared_cache_debug - if set to 1 the module will turn on the
            debug option in IPC::SharedCache - see IPC::SharedCache for
            details. Defaults to 0.

        *   memory_debug - if set to 1 the module will send information on
            cache memory usage to STDERR. Requires the GTop module. Defaults
            to 0.

    Miscellaneous Options

        *   associate - this option allows you to inherit the parameter
            values from other objects. The only requirement for the other
            object is that it have a param() method that works like
            HTML::Template's param(). A good candidate would be a CGI.pm
            query object. Example:

              my $query = new CGI;
              my $template = HTML::Template->new(filename => 'template.tmpl',
                                                 associate => $query);

            Now, $template->output() will act as though

              $template->param('FormField', $cgi->param('FormField'));

            had been specified for each key/value pair that would be
            provided by the $cgi->param() method. Parameters you set
            directly take precedence over associated parameters.

            You can specify multiple objects to associate by passing an
            anonymous array to the associate option. They are searched for
            parameters in the order they appear:

              my $template = HTML::Template->new(filename => 'template.tmpl',
                                                 associate => [$query, $other_obj]);

            The old associateCGI() call is still supported, but should be
            considered obsolete.

            NOTE: The parameter names are matched in a case-insensitve
            manner. If you have two parameters in a CGI object like 'NAME'
            and 'Name' one will be chosen randomly by associate. This
            behavior can be changed by the following option.

        *   case_sensitive - setting this option to true causes
            HTML::Template to treat template variable names
            case-sensitively. The following example would only set one
            parameter without the "case_sensitive" option:

              my $template = HTML::Template->new(filename => 'template.tmpl',
                                                 case_sensitive => 1);
              $template->param(
                FieldA => 'foo',
                fIELDa => 'bar',
              );

            This option defaults to off.

            NOTE: with case_sensitive and loop_context_vars the special loop
            variables are available in lower-case only.

        *   loop_context_vars - when this parameter is set to true (it is
            false by default) four loop context variables are made available
            inside a loop: , , , . They can
            be used with <TMPL_IF>, <TMPL_UNLESS> and <TMPL_ELSE> to control
            how a loop is output.

            In addition to the above, a  var is also made
            available when loop context variables are turned on.

            Example:

               <TMPL_LOOP NAME="FOO">
                  <TMPL_IF NAME="">
                    This only outputs on the first pass.
                  </TMPL_IF>

                  <TMPL_IF NAME="">
                    This outputs every other pass, on the odd passes.
                  </TMPL_IF>

                  <TMPL_UNLESS NAME="">
                    This outputs every other pass, on the even passes.
                  </TMPL_IF>

                  <TMPL_IF NAME="">
                    This outputs on passes that are neither first nor last.
                  </TMPL_IF>

                  This is pass number <TMPL_VAR NAME="">.

                  <TMPL_IF NAME="">
                    This only outputs on the last pass.
                  <TMPL_IF>
               </TMPL_LOOP>

            One use of this feature is to provide a "separator" similar in
            effect to the perl function join(). Example:

               <TMPL_LOOP FRUIT>
                  <TMPL_IF > and </TMPL_IF>
                  <TMPL_VAR KIND><TMPL_UNLESS >, <TMPL_ELSE>.</TMPL_UNLESS>
               </TMPL_LOOP>

            Would output (in a browser) something like:

              Apples, Oranges, Brains, Toes, and Kiwi.

            Given an appropriate param() call, of course. NOTE: A loop with
            only a single pass will get both  and  set to
            true, but not .

        *   no_includes - set this option to 1 to disallow the
            <TMPL_INCLUDE> tag in the template file. This can be used to
            make opening untrusted templates slightly less dangerous.
            Defaults to 0.

        *   max_includes - set this variable to determine the maximum depth
            that includes can reach. Set to 10 by default. Including files
            to a depth greater than this value causes an error message to be
            displayed. Set to 0 to disable this protection.

        *   global_vars - normally variables declared outside a loop are not
            available inside a loop. This option makes <TMPL_VAR>s like
            global variables in Perl - they have unlimited scope. This
            option also affects <TMPL_IF> and <TMPL_UNLESS>.

            Example:

              This is a normal variable: <TMPL_VAR NORMAL>.<P>

              <TMPL_LOOP NAME=FROOT_LOOP>
                 Here it is inside the loop: <TMPL_VAR NORMAL><P>
              </TMPL_LOOP>

            Normally this wouldn't work as expected, since <TMPL_VAR
            NORMAL>'s value outside the loop is not available inside the
            loop.

            The global_vars option also allows you to access the values of
            an enclosing loop within an inner loop. For example, in this
            loop the inner loop will have access to the value of OUTER_VAR
            in the correct iteration:

               <TMPL_LOOP OUTER_LOOP>
                  OUTER: <TMPL_VAR OUTER_VAR>
                    <TMPL_LOOP INNER_LOOP>
                       INNER: <TMPL_VAR INNER_VAR>
                       INSIDE OUT: <TMPL_VAR OUTER_VAR>
                    </TMPL_LOOP>
               </TMPL_LOOP>

        *   filter - this option allows you to specify a filter for your
            template files. A filter is a subroutine that will be called
            after HTML::Template reads your template file but before it
            starts parsing template tags.

            In the most simple usage, you simply assign a code reference to
            the filter parameter. This subroutine will recieve a single
            arguement - a reference to a string containing the template file
            text. Here is an example that accepts templates with tags that
            look like "!!!ZAP_VAR FOO!!!" and transforms them into
            HTML::Template tags:

               my $filter = sub {
                 my $text_ref = shift;
                 $$text_ref =~ s/!!!ZAP_(.*?)!!!/<TMPL_$1>/g;
               };

               # open zap.tmpl using the above filter
               my $template = HTML::Template->new(filename => 'zap.tmpl',
                                                  filter => $filter);

            More complicated usages are possible. You can request that your
            filter receieve the template text as an array of lines rather
            than as a single scalar. To do that you need to specify your
            filter using a hash-ref. In this form you specify the filter
            using the "sub" key and the desired argument format using the
            "format" key. The available formats are "scalar" and "array".
            Using the "array" format will incur a performance penalty but
            may be more convenient in some situations.

               my $template = HTML::Template->new(filename => 'zap.tmpl',
                                                  filter => { sub => $filter,
                                                              format => 'array' });

            You may also have multiple filters. This allows simple filters
            to be combined for more elaborate functionality. To do this you
            specify an array of filters. The filters are applied in the
            order they are specified.

               my $template = HTML::Template->new(filename => 'zap.tmpl',
                                                  filter => [
                                                       { sub => \&decompress,
                                                         format => 'scalar' },
                                                       { sub => \&remove_spaces,
                                                         format => 'array' }
                                                    ]);

            The specified filters will be called for any TMPL_INCLUDEed
            files just as they are for the main template file.

  param()
    param() can be called in a number of ways

    1) To return a list of parameters in the template :

       my @parameter_names = $self->param();

    2) To return the value set to a param :

       my $value = $self->param('PARAM');

    3) To set the value of a parameter :

          # For simple TMPL_VARs:
          $self->param(PARAM => 'value');

          # with a subroutine reference that gets called to get the value
          # of the scalar.  The sub will recieve the template object as a
          # parameter.
          $self->param(PARAM => sub { return 'value' });

          # And TMPL_LOOPs:
          $self->param(LOOP_PARAM =>
                       [
                        { PARAM => VALUE_FOR_FIRST_PASS, ... },
                        { PARAM => VALUE_FOR_SECOND_PASS, ... }
                        ...
                       ]
                      );

    4) To set the value of a a number of parameters :

         # For simple TMPL_VARs:
         $self->param(PARAM => 'value',
                      PARAM2 => 'value'
                     );

          # And with some TMPL_LOOPs:
          $self->param(PARAM => 'value',
                       PARAM2 => 'value',
                       LOOP_PARAM =>
                       [
                        { PARAM => VALUE_FOR_FIRST_PASS, ... },
                        { PARAM => VALUE_FOR_SECOND_PASS, ... }
                        ...
                       ],
                       ANOTHER_LOOP_PARAM =>
                       [
                        { PARAM => VALUE_FOR_FIRST_PASS, ... },
                        { PARAM => VALUE_FOR_SECOND_PASS, ... }
                        ...
                       ]
                      );

    5) To set the value of a a number of parameters using a hash-ref :

          $self->param(
                       {
                          PARAM => 'value',
                          PARAM2 => 'value',
                          LOOP_PARAM =>
                          [
                            { PARAM => VALUE_FOR_FIRST_PASS, ... },
                            { PARAM => VALUE_FOR_SECOND_PASS, ... }
                            ...
                          ],
                          ANOTHER_LOOP_PARAM =>
                          [
                            { PARAM => VALUE_FOR_FIRST_PASS, ... },
                            { PARAM => VALUE_FOR_SECOND_PASS, ... }
                            ...
                          ]
                        }
                       );

  clear_params()
    Sets all the parameters to undef. Useful internally, if nowhere else!

  output()
    output() returns the final result of the template. In most situations
    you'll want to print this, like:

       print $template->output();

    When output is called each occurrence of <TMPL_VAR NAME=name> is
    replaced with the value assigned to "name" via param(). If a named
    parameter is unset it is simply replaced with ''. <TMPL_LOOPS> are
    evaluated once per parameter set, accumlating output on each pass.

    Calling output() is guaranteed not to change the state of the Template
    object, in case you were wondering. This property is mostly important
    for the internal implementation of loops.

    You may optionally supply a filehandle to print to automatically as the
    template is generated. This may improve performance and lower memory
    consumption. Example:

       $template->output(print_to => *STDOUT);

    The return value is undefined when using the "print_to" option.

  query()
    This method allow you to get information about the template structure.
    It can be called in a number of ways. The simplest usage of query is
    simply to check whether a parameter name exists in the template, using
    the "name" option:

      if ($template->query(name => 'foo')) {
        # do something if a varaible of any type
        # named FOO is in the template
      }

    This same usage returns the type of the parameter. The type is the same
    as the tag minus the leading 'TMPL_'. So, for example, a TMPL_VAR
    parameter returns 'VAR' from query().

      if ($template->query(name => 'foo') eq 'VAR') {
        # do something if FOO exists and is a TMPL_VAR
      }

    Note that the variables associated with TMPL_IFs and TMPL_UNLESSs will
    be identified as 'VAR' unless they are also used in a TMPL_LOOP, in
    which case they will return 'LOOP'.

    "query()" also allows you to get a list of parameters inside a loop (and
    inside loops inside loops). Example loop:

       <TMPL_LOOP NAME="EXAMPLE_LOOP">
         <TMPL_VAR NAME="BEE">
         <TMPL_VAR NAME="BOP">
         <TMPL_LOOP NAME="EXAMPLE_INNER_LOOP">
           <TMPL_VAR NAME="INNER_BEE">
           <TMPL_VAR NAME="INNER_BOP">
         </TMPL_LOOP>
       </TMPL_LOOP>

    And some query calls:

      # returns 'LOOP'
      $type = $template->query(name => 'EXAMPLE_LOOP');

      # returns ('bop', 'bee', 'example_inner_loop')
      @param_names = $template->query(loop => 'EXAMPLE_LOOP');

      # both return 'VAR'
      $type = $template->query(name => ['EXAMPLE_LOOP', 'BEE']);
      $type = $template->query(name => ['EXAMPLE_LOOP', 'BOP']);

      # and this one returns 'LOOP'
      $type = $template->query(name => ['EXAMPLE_LOOP',
                                        'EXAMPLE_INNER_LOOP']);

      # and finally, this returns ('inner_bee', 'inner_bop')
      @inner_param_names = $template->query(loop => ['EXAMPLE_LOOP',
                                                     'EXAMPLE_INNER_LOOP']);

      # for non existent parameter names you get undef
      # this returns undef.
      $type = $template->query(name => 'DWEAZLE_ZAPPA');

      # calling loop on a non-loop parameter name will cause an error.
      # this dies:
      $type = $template->query(loop => 'DWEAZLE_ZAPPA');

    As you can see above the "loop" option returns a list of parameter names
    and both "name" and "loop" take array refs in order to refer to
    parameters inside loops. It is an error to use "loop" with a parameter
    that is not a loop.

    Note that all the names are returned in lowercase and the types are
    uppercase.

    Just like "param()", "query()" with no arguements returns all the
    parameter names in the template at the top level.

FREQUENTLY ASKED QUESTIONS
    In the interest of greater understanding I've started a FAQ section of
    the perldocs. Please look in here before you send me email.

    1   Q: Is there a place to go to discuss HTML::Template and/or get help?

        A: There's a mailing-list for discussing HTML::Template at
        html-template-users@lists.sourceforge.net. To join:

           [URL=http://lists.sourceforge.net/lists/listinfo/html-template-users]http://lists.sourceforge.net/lists/listinfo/html-template-users[/URL]

        If you just want to get email when new releases are available you
        can join the announcements mailing-list here:

           [URL=http://lists.sourceforge.net/lists/listinfo/html-template-announce]http://lists.sourceforge.net/lists/listinfo/html-template-announce[/URL]
    
    2   Q: Is there a searchable archive for the mailing-list?

        A: Yes, you can find an archive of the SourceForge list here:

          [URL=http://www.geocrawler.com/lists/3/SourceForge/23294/0/]http://www.geocrawler.com/lists/3/SourceForge/23294/0/[/URL]

        For an archive of the old vm.com list, setup by Sean P. Scanlon,
        see:

           [URL=http://bluedot.net/mail/archive/]http://bluedot.net/mail/archive/[/URL]

    3   Q: I want support for <TMPL_XXX>! How about it?

        A: Maybe. I definitely encourage people to discuss their ideas for
        HTML::Template on the mailing list. Please be ready to explain to me
        how the new tag fits in with HTML::Template's mission to provide a
        fast, lightweight system for using HTML templates.

        NOTE: Offering to program said addition and provide it in the form
        of a patch to the most recent version of HTML::Template will
        definitely have a softening effect on potential opponents!

    4   Q: I found a bug, can you fix it?

        A: That depends. Did you send me the VERSION of HTML::Template, a
        test script and a test template? If so, then almost certainly.

        If you're feeling really adventurous, HTML::Template has a
        publically available CVS server. See below for more information in
        the PUBLIC CVS SERVER section.

    5   Q: <TMPL_VAR>s from the main template aren't working inside a
        <TMPL_LOOP>! Why?

        A: This is the intended behavior. <TMPL_LOOP> introduces a separate
        scope for <TMPL_VAR>s much like a subroutine call in Perl introduces
        a separate scope for "my" variables.

        If you want your <TMPL_VAR>s to be global you can set the
        'global_vars' option when you call new(). See above for
        documentation of the 'global_vars' new() option.

    6   Q: Why do you use /[Tt]/ instead of /t/i? It's so ugly!

        A: Simple - the case-insensitive match switch is very inefficient.
        According to _Mastering_Regular_Expressions_ from O'Reilly Press,
        /[Tt]/ is faster and more space efficient than /t/i - by as much as
        double against long strings. //i essentially does a lc() on the
        string and keeps a temporary copy in memory.

        When this changes, and it is in the 5.6 development series, I will
        gladly use //i. Believe me, I realize [Tt] is hideously ugly.

    7   Q: How can I pre-load my templates using cache-mode and mod_perl?

        A: Add something like this to your startup.pl:

           use HTML::Template;
           use File::Find;

           print STDERR "Pre-loading HTML Templates...\n";
           find(
                sub {
                  return unless /\.tmpl$/;
                  HTML::Template->new(
                                      filename => "$File::Find::dir/$_",
                                      cache => 1,
                                     );
                },
                '/path/to/templates',
                '/another/path/to/templates/'
              );

        Note that you'll need to modify the "return unless" line to specify
        the extension you use for your template files - I use .tmpl, as you
        can see. You'll also need to specify the path to your template
        files.

        One potential problem: the "/path/to/templates/" must be EXACTLY the
        same path you use when you call HTML::Template->new(). Otherwise the
        cache won't know they're the same file and will load a new copy -
        instead getting a speed increase, you'll double your memory usage.
        To find out if this is happening set cache_debug => 1 in your
        application code and look for "CACHE MISS" messages in the logs.

    8   Q: What characters are allowed in TMPL_* NAMEs?

        A: Numbers, letters, '.', '/', '+', '-' and '_'.

    9   Q: How can I execute a program from inside my template?

        A: Short answer: you can't. Longer answer: you shouldn't since this
        violates the fundamental concept behind HTML::Template - that design
        and code should be seperate.

        But, inevitably some people still want to do it. If that describes
        you then you should take a look at HTML::Template::Expr. Using
        HTML::Template::Expr it should be easy to write a run_program()
        function. Then you can do awful stuff like:

          <tmpl_var expr="run_program('foo.pl')">

        Just, please, don't tell me about it. I'm feeling guilty enough just
        for writing HTML::Template::Expr in the first place.

    10  Q: Can I get a copy of these docs in Japanese?

        A: Yes you can. See Kawai Takanori's translation at:

           [URL=http://member.nifty.ne.jp/hippo2000/perltips/html/template.htm]http://member.nifty.ne.jp/hippo2000/perltips/html/template.htm[/URL]

    11  Q: What's the best way to create a <select> form element using
        HTML::Template?

        A: There is much disagreement on this issue. My personal preference
        is to use CGI.pm's excellent popup_menu() and scrolling_list()
        functions to fill in a single <tmpl_var select_foo> variable.

        To some people this smacks of mixing HTML and code in a way that
        they hoped HTML::Template would help them avoid. To them I'd say
        that HTML is a violation of the principle of separating design from
        programming. There's no clear separation between the programmatic
        elements of the <form> tags and the layout of the <form> tags.
        You'll have to draw the line somewhere - clearly the designer can't
        be entirely in charge of form creation.

        It's a balancing act and you have to weigh the pros and cons on each
        side. It is certainly possible to produce a <select> element
        entirely inside the template. What you end up with is a rat's nest
        of loops and conditionals. Alternately you can give up a certain
        amount of flexibility in return for vastly simplifying your
        templates. I generally choose the latter.

        Another option is to investigate HTML::FillInForm which some have
        reported success using to solve this problem.