From olaf at buwen.de  Thu Sep  1 09:48:56 2005
From: olaf at buwen.de (Olaf Buwen)
Date: Thu Sep  1 09:49:12 2005
Subject: [Petal] How do I use selected/checked ?
In-Reply-To: <389e9736fd2421653d3e621a13fb173c@mastersofbranding.com>
References: <337E0616-A691-4977-A5CF-96F8A1FDF06D@mastersofbranding.com>	<20050831075524.GB4176@webarchitects.co.uk>	<F3FC7C6A-3066-423B-AC98-EF0B66EB1F16@mastersofbranding.com>	<1125502219.1174.1.camel@wsmithbox>
	<389e9736fd2421653d3e621a13fb173c@mastersofbranding.com>
Message-ID: <4316C078.3050407@buwen.de>

This template:

<select>
    <option
	petal:repeat="entry options"
        petal:attributes="value entry/lang; selected entry/selected"
        petal:content="entry/title"/>
</select>

with this code:
...
print $template->process(
    options => [
        {lang => "foo", title => "Foobar"},
        {lang => "bar", title => "Barbar", selected => 'selected'},
        {lang => "baz", title => "Bazbar"},
    ]
);

gives this output:
<select>
        <option value="foo" >Foobar</option>
	<option value="bar" selected="selected">Barbar</option>
	<option value="baz" >Bazbar</option>
</select>


It is not in the docs, but AFAIK it is part of the TAL spec
that an attribute is not set when the appropriate entry is missing.

I hope this helps
Olaf
From josh at narins.net  Thu Sep  1 16:14:21 2005
From: josh at narins.net (Josh Narins)
Date: Thu Sep  1 16:14:27 2005
Subject: [Petal] Tree Tactics
In-Reply-To: <20050826195710.GB463@narins.net>
References: <20050826195710.GB463@narins.net>
Message-ID: <20050901151420.GA1918@narins.net>

> I've got a tree, with arbitrary depth, like..
> 
> {
>   text => 'This is the top',
>   sort => [5,3,7],
>   children =>
>     3 => {
>           text => 'This should be second',
>           sort => [4,6],
>           children {
>              4 => { text=>'a leaf',},
>              4 => { text=>'another leaf',},
>           },
>     5 => ...
>     7 => ...
>   }
> }
> 
> I hope you get the idea.
> 
> How do I get this to work in Petal?
> 
> I figure I should be able to do it with a single template.
> 
> myself.xml
> ==========
> 
> <table>
>   <tr>
>     <td petal:content="object/text">
>      This is dummy text 
>     </td>
>     <td petal:repeat="SOMETHING HERE">
>      <xi:include href="./myself.xml" />
>     </td>
>   </tr>
> </table>
> 
> What should that repeat be? I can do it with "each:" but then I don't
> get any order. I want the order to be that as described in the sort
> array ref.
> 
> How does anyone handle recursive trees?

No one has recursive tips?


From chris at webarchitects.co.uk  Thu Sep  1 16:19:12 2005
From: chris at webarchitects.co.uk (Chris Croome)
Date: Thu Sep  1 16:19:24 2005
Subject: [Petal] Tree Tactics
In-Reply-To: <20050901151420.GA1918@narins.net>
References: <20050826195710.GB463@narins.net>
	<20050901151420.GA1918@narins.net>
Message-ID: <20050901151912.GA20317@webarchitects.co.uk>

Hi

On Thu 01-Sep-2005 at 11:14:21AM -0400, Josh Narins wrote:
> > 
> > I figure I should be able to do it with a single template.
> 
> No one has recursive tips?

With MKDoc, for things like the sitemap, which is made up of nested
<UL>s, we use an include file -- it's not a single template, I don't
know how you would do it with a single template, but it might well
be possible...

Chris
 
-- 
Chris Croome                               <chris@webarchitects.co.uk>
web design                             http://www.webarchitects.co.uk/ 
web content management                               http://mkdoc.com/   
From raymond at print-quotes-software.com  Thu Sep  1 16:31:30 2005
From: raymond at print-quotes-software.com (Raymond Barkhouse)
Date: Thu Sep  1 16:21:52 2005
Subject: [Petal] How do I use selected/checked ?
In-Reply-To: <4316C078.3050407@buwen.de>
References: <337E0616-A691-4977-A5CF-96F8A1FDF06D@mastersofbranding.com>
	<20050831075524.GB4176@webarchitects.co.uk>
	<F3FC7C6A-3066-423B-AC98-EF0B66EB1F16@mastersofbranding.com>
	<1125502219.1174.1.camel@wsmithbox>
	<389e9736fd2421653d3e621a13fb173c@mastersofbranding.com>
	<4316C078.3050407@buwen.de>
Message-ID: <1125588690.6799.40.camel@localhost.localdomain>

On Thu, 2005-01-09 at 10:48 +0200, Olaf Buwen wrote:
> This template:
> 
> <select>
>     <option
> 	petal:repeat="entry options"
>         petal:attributes="value entry/lang; selected entry/selected"
>         petal:content="entry/title"/>
> </select>
> 
> with this code:
> ...
> print $template->process(
>     options => [
>         {lang => "foo", title => "Foobar"},
>         {lang => "bar", title => "Barbar", selected => 'selected'},
>         {lang => "baz", title => "Bazbar"},
>     ]
> );
> 
> gives this output:
> <select>
>         <option value="foo" >Foobar</option>
> 	<option value="bar" selected="selected">Barbar</option>
> 	<option value="baz" >Bazbar</option>
> </select>
> 

That's the same method I use for very dynamic forms. For just filling in
form variables from a database/Apache::Request->param/CGI->param though
you might want to consider using HTML::FillInForm.

Given a hash (
	foo => 'quibble',
	bar => 'qax',
	baz => '3',
);

Some form elements:

	<input name="foo" type="text" value="" />

        <input name="bar" type="radio" value="qux" />
        <input name="bar" type="radio" value="qax" />
        <input name="bar" type="radio" value="quz" />

	<select name="baz">
	  <option value="1">One</option>
	  <option value="2">Two</option>
	  <option value="3">Three</option>
	  <option value="4">Four</option>
	</select>

And a slight change to your perl:

	$r->print(HTML::FillInForm->new->fill(
            fdata     => \%hash, 
            scalarref => \$t->process($other_stuff),
	));

Will give you:

	<input name="foo" type="text" value="quibble" />

        <input name="bar" type="radio" value="qux" />
        <input name="bar" type="radio" value="qax" checked="checked" />
        <input name="bar" type="radio" value="quz" />

	<select name="baz">
	  <option value="1">One</option>
	  <option value="2">Two</option>
	  <option value="3" selected="selected">Three</option>
	  <option value="4">Four</option>
	</select>


Much cleaner at the expense of some processing. I'm currently working on
doing the substitution at the same time Petal is processing the XML
stream as there's no need to process it twice (as it currently
happening). If anyone else is working in that direction I'd be
interested in hearing from you.


From bruno at mkdoc.com  Thu Sep  1 17:24:31 2005
From: bruno at mkdoc.com (Bruno Postle)
Date: Thu Sep  1 17:24:44 2005
Subject: [Petal] Tree Tactics
In-Reply-To: <20050901151912.GA20317@webarchitects.co.uk>
References: <20050826195710.GB463@narins.net>
	<20050901151420.GA1918@narins.net>
	<20050901151912.GA20317@webarchitects.co.uk>
Message-ID: <20050901162431.GX31593@webarchitects.co.uk>

On Thu 01-Sep-2005 at 16:19 +0100, Chris Croome wrote:
> 
> With MKDoc, for things like the sitemap, which is made up of nested
> <UL>s, we use an include file -- it's not a single template, I don't
> know how you would do it with a single template, but it might well
> be possible...

The sitemap fragment that includes itself looks something like this:

  <?xml version="1.0"?>
  <li xmlns:petal="http://purl.org/petal/1.0/">
    <a
      href="#"
      petal:attributes="href child/URI"
      petal:content="child/Title"
    >Child Document Title</a>
    <ul
      petal:define="children child/Children"
      petal:condition="children"
      petal:repeat="child children"
    >
    <?include file="fragment.xml"?>
    </ul>
  </li>

Note that 'Children' is a method that returns a list of objects that
are already sorted.  We don't try to do the sorting in the template.

-- 
Bruno
From fergal at esatclear.ie  Thu Sep  1 22:31:14 2005
From: fergal at esatclear.ie (Fergal Daly)
Date: Thu Sep  1 22:31:25 2005
Subject: [Petal] Tree Tactics
In-Reply-To: <20050901151420.GA1918@narins.net>
References: <20050826195710.GB463@narins.net>
	<20050901151420.GA1918@narins.net>
Message-ID: <8750299605090114316271b045@mail.gmail.com>

You can create a custom modifier that takes a template name and a set of 
named arguments, it loads and executes the template, passing in the 
arguments. That's how I've done it in the past.

It's in the mailing list archives I think,

F

On 9/1/05, Josh Narins <josh@narins.net> wrote:
> 
> > I've got a tree, with arbitrary depth, like..
> >
> > {
> > text => 'This is the top',
> > sort => [5,3,7],
> > children =>
> > 3 => {
> > text => 'This should be second',
> > sort => [4,6],
> > children {
> > 4 => { text=>'a leaf',},
> > 4 => { text=>'another leaf',},
> > },
> > 5 => ...
> > 7 => ...
> > }
> > }
> >
> > I hope you get the idea.
> >
> > How do I get this to work in Petal?
> >
> > I figure I should be able to do it with a single template.
> >
> > myself.xml
> > ==========
> >
> > <table>
> > <tr>
> > <td petal:content="object/text">
> > This is dummy text
> > </td>
> > <td petal:repeat="SOMETHING HERE">
> > <xi:include href="./myself.xml" />
> > </td>
> > </tr>
> > </table>
> >
> > What should that repeat be? I can do it with "each:" but then I don't
> > get any order. I want the order to be that as described in the sort
> > array ref.
> >
> > How does anyone handle recursive trees?
> 
> No one has recursive tips?
> 
> 
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://lists.webarch.co.uk/pipermail/petal/attachments/20050901/8070ffee/attachment.html
From josh at narins.net  Fri Sep  2 12:47:38 2005
From: josh at narins.net (Josh Narins)
Date: Fri Sep  2 12:47:50 2005
Subject: [Petal] Tree Tactics
In-Reply-To: <20050901162431.GX31593@webarchitects.co.uk>
References: <20050826195710.GB463@narins.net>
	<20050901151420.GA1918@narins.net>
	<20050901151912.GA20317@webarchitects.co.uk>
	<20050901162431.GX31593@webarchitects.co.uk>
Message-ID: <20050902114738.GA6772@narins.net>

> On Thu 01-Sep-2005 at 16:19 +0100, Chris Croome wrote:
> > 
> > With MKDoc, for things like the sitemap, which is made up of nested
> > <UL>s, we use an include file -- it's not a single template, I don't
> > know how you would do it with a single template, but it might well
> > be possible...
> 
> The sitemap fragment that includes itself looks something like this:
> 
>   <?xml version="1.0"?>
>   <li xmlns:petal="http://purl.org/petal/1.0/">
>     <a
>       href="#"
>       petal:attributes="href child/URI"
>       petal:content="child/Title"
>     >Child Document Title</a>
>     <ul
>       petal:define="children child/Children"
>       petal:condition="children"
>       petal:repeat="child children"
>     >
>     <?include file="fragment.xml"?>
>     </ul>
>   </li>
> 
> Note that 'Children' is a method that returns a list of objects that
> are already sorted.  We don't try to do the sorting in the template.

Sorry for the confusion.

My own efforts would have worked had their not been bad data
in the database, which was resulting in a bad select, which 
made the tree "wrong" when displayed.

Recursively called templates work fine.


From josh at narins.net  Fri Sep  2 12:53:42 2005
From: josh at narins.net (Josh Narins)
Date: Fri Sep  2 12:53:47 2005
Subject: [Petal] Extra Utils
Message-ID: <20050902115342.GB6772@narins.net>

Writing my own "greaterthan" => 'gt' and "greaterthanorequals" => 'gte'
Petal::Utils was trivial (thanks to the authors of Equal.pm).

How does the decision to get stuff in Utils get made?

tia,
Josh
From jvanasco at mastersofbranding.com  Sat Sep  3 22:18:20 2005
From: jvanasco at mastersofbranding.com (Jonathan Vanasco)
Date: Sat Sep  3 22:19:15 2005
Subject: [Petal] & in get href
Message-ID: <E1D83405-73E8-47ED-851B-D21B363514C8@mastersofbranding.com>

petal seems to turn all & instances into &amp;

i tried using &amp; which is 'well formed xml'
no luck

then i tried changing the petal input/output formats
no luck

if i do
     tal:attributes="href structure string:a&b"

then its fine

my question is:
   must i do that?  will browsers handle &amp; well or will that  
screw up my get/post vars

From bruno at mkdoc.com  Sat Sep  3 23:34:22 2005
From: bruno at mkdoc.com (Bruno Postle)
Date: Sat Sep  3 23:34:29 2005
Subject: [Petal] & in get href
In-Reply-To: <E1D83405-73E8-47ED-851B-D21B363514C8@mastersofbranding.com>
References: <E1D83405-73E8-47ED-851B-D21B363514C8@mastersofbranding.com>
Message-ID: <20050903223422.GA1816@webarchitects.co.uk>

On Sat 03-Sep-2005 at 17:18 -0400, Jonathan Vanasco wrote:
> 
> my question is:
>   must i do that?  will browsers handle &amp; well or will that  
> screw up my get/post vars

It's fine, browsers will do the right thing if given something like
this:

  <a href="http://example.com/foo?a=b&amp;c=d">click here!</a>

-- 
Bruno
From chris at webarchitects.co.uk  Mon Sep  5 10:14:58 2005
From: chris at webarchitects.co.uk (Chris Croome)
Date: Mon Sep  5 10:15:11 2005
Subject: [Petal] & in get href
In-Reply-To: <20050903223422.GA1816@webarchitects.co.uk>
References: <E1D83405-73E8-47ED-851B-D21B363514C8@mastersofbranding.com>
	<20050903223422.GA1816@webarchitects.co.uk>
Message-ID: <20050905091458.GB26389@webarchitects.co.uk>

Hi

On Sat 03-Sep-2005 at 11:34:22PM +0100, Bruno Postle wrote:
> > 
> > will browsers handle &amp; well or will that  screw up my
> > get/post vars

More on this here:

  http://htmlhelp.com/tools/validator/problems.html.en#amp

Chris

-- 
Chris Croome                               <chris@webarchitects.co.uk>
web design                             http://www.webarchitects.co.uk/ 
web content management                               http://mkdoc.com/   
From bruno at mkdoc.com  Wed Sep  7 16:52:51 2005
From: bruno at mkdoc.com (Bruno Postle)
Date: Wed Sep  7 16:53:03 2005
Subject: [Petal] Petal-2.17 released
Message-ID: <20050907155251.GE31593@webarchitects.co.uk>

http://search.cpan.org/dist/Petal-2.17/

This release contains only one minor bugfix, documentation fixes and
some extra tests:

  2.17 Wed Sep  7 15:30:00 2005
      - test 041 check that &nbsp; becomes chr(160) not chr(194) (William McKee)
      - Petal::Canonicalizer::XML Fix bug where \-escaped semicolons were not recognised (William McKee)
      - test 090 for ${1} appearance when Petal::I18N->process
      - pod fixes (Kurt Stephens, Tom Metro, Bruno Postle)

Diff between this and the last release:
http://search.cpan.org/diff/Petal-2.16-Petal-2.17.diff

-- 
Bruno
From bruno at mkdoc.com  Fri Sep  9 17:10:47 2005
From: bruno at mkdoc.com (Bruno Postle)
Date: Fri Sep  9 17:10:55 2005
Subject: [Petal] BUG: newlines within attributes?
Message-ID: <20050909161047.GA12903@mkdoc.com>

It seems that Petal processes these two differently:

  <div petal:content="string:
    plum">xxx</div>

  <div petal:content="string: pear">xxx</div>

The first one gets 'half processed'.  I was under the impression 
that Petal should be whitespace insensitive here.

Attached a test case, Could somebody else confirm that I'm really 
looking at such an obvious bug?

The following one-liner fixes it for me.  Are there any valid 
reasons for preserving newlines within attributes?

Index: lib/Petal/Parser.pm
===================================================================
RCS file: /var/spool/cvs/Petal/lib/Petal/Parser.pm,v
retrieving revision 1.5
diff -u -r1.5 Parser.pm
--- lib/Petal/Parser.pm 9 Jul 2004 11:09:34 -0000       1.5
+++ lib/Petal/Parser.pm 9 Sep 2005 16:07:22 -0000
@@ -114,6 +114,8 @@
     local %_ = %{shift()};
     delete $_{'/'};

+    foreach my $key (keys %_) { $_{$key} =~ s/\n/ /g }
+
     # process the Petal namespace
     my $ns = (scalar @NameSpaces) ? $NameSpaces[$#NameSpaces] : $Petal::NS;
     foreach my $key (keys %_)

-- 
Bruno
-------------- next part --------------
#!/usr/bin/perl
use warnings;
use strict;
use lib ('lib');
use Test::More 'no_plan';
use Petal;

$Petal::BASE_DIR     = './t/data/';
$Petal::DISK_CACHE   = 0;
$Petal::MEMORY_CACHE = 0;
$Petal::TAINT        = 1;

my $string = Petal->new ('attribute-newline.xml')->process ();
like ($string, '/>apple</', '');
like ($string, '/>orange</', '');
like ($string, '/>plum</', '');
like ($string, '/>pear</', '');

-------------- next part --------------
<stuff xmlns:petal="http://purl.org/petal/1.0/">

<div petal:define="foo
string: apple">$foo</div>
<div petal:define="bar string: orange">$bar</div>

<div petal:content="string:
plum">xxx</div>
<div petal:content="string: pear">xxx</div>

</stuff>
From mark at thinkfoo.com  Fri Sep  9 18:13:43 2005
From: mark at thinkfoo.com (Mark Holland)
Date: Fri Sep  9 18:13:56 2005
Subject: [Petal] BUG: newlines within attributes?
In-Reply-To: <20050909161047.GA12903@mkdoc.com>
References: <20050909161047.GA12903@mkdoc.com>
Message-ID: <4321C2C7.9090603@thinkfoo.com>

Hi Bruno,

I can confirm I get the bug with my unpatched version of Petal-2.16. The 
patch fixes it. I've also recompiled and tested our templates with the 
patched version and they seem fine.

I, and the guys doing Petal here, don't tend to split Petal attributes 
across lines, so I'm sure it's something that could go unnoticed...

Cheers,
~mark

Bruno Postle wrote:

> It seems that Petal processes these two differently:
>
>  <div petal:content="string:
>    plum">xxx</div>
>
>  <div petal:content="string: pear">xxx</div>
>
> The first one gets 'half processed'.  I was under the impression that 
> Petal should be whitespace insensitive here.
>
> Attached a test case, Could somebody else confirm that I'm really 
> looking at such an obvious bug?
>
> The following one-liner fixes it for me.  Are there any valid reasons 
> for preserving newlines within attributes?
>
> Index: lib/Petal/Parser.pm
> ===================================================================
> RCS file: /var/spool/cvs/Petal/lib/Petal/Parser.pm,v
> retrieving revision 1.5
> diff -u -r1.5 Parser.pm
> --- lib/Petal/Parser.pm 9 Jul 2004 11:09:34 -0000       1.5
> +++ lib/Petal/Parser.pm 9 Sep 2005 16:07:22 -0000
> @@ -114,6 +114,8 @@
>     local %_ = %{shift()};
>     delete $_{'/'};
>
> +    foreach my $key (keys %_) { $_{$key} =~ s/\n/ /g }
> +
>     # process the Petal namespace
>     my $ns = (scalar @NameSpaces) ? $NameSpaces[$#NameSpaces] : 
> $Petal::NS;
>     foreach my $key (keys %_)
>
>------------------------------------------------------------------------
>
>#!/usr/bin/perl
>use warnings;
>use strict;
>use lib ('lib');
>use Test::More 'no_plan';
>use Petal;
>
>$Petal::BASE_DIR     = './t/data/';
>$Petal::DISK_CACHE   = 0;
>$Petal::MEMORY_CACHE = 0;
>$Petal::TAINT        = 1;
>
>my $string = Petal->new ('attribute-newline.xml')->process ();
>like ($string, '/>apple</', '');
>like ($string, '/>orange</', '');
>like ($string, '/>plum</', '');
>like ($string, '/>pear</', '');
>
>  
>
>------------------------------------------------------------------------
>
><stuff xmlns:petal="http://purl.org/petal/1.0/">
>
><div petal:define="foo
>string: apple">$foo</div>
><div petal:define="bar string: orange">$bar</div>
>
><div petal:content="string:
>plum">xxx</div>
><div petal:content="string: pear">xxx</div>
>
></stuff>
>  
>

From bruno at mkdoc.com  Sat Sep 10 21:22:21 2005
From: bruno at mkdoc.com (Bruno Postle)
Date: Sat Sep 10 21:22:26 2005
Subject: [Petal] BUG: newlines within attributes?
In-Reply-To: <4321C2C7.9090603@thinkfoo.com>
References: <20050909161047.GA12903@mkdoc.com> <4321C2C7.9090603@thinkfoo.com>
Message-ID: <20050910202221.GF2645@webarchitects.co.uk>

On Fri 09-Sep-2005 at 18:13 +0100, Mark Holland wrote:
>
> I, and the guys doing Petal here, don't tend to split Petal attributes 
> across lines, so I'm sure it's something that could go unnoticed...

We have lots of stuff like this:

  <div petal:define="a string: ay;
                     b string: bee;
                     c string: cee;">

..but the Petal canonicaliser splits the attribute on the 
semi-colons and turns it into three separate statements, which then 
pass through the system happily.

The bug seems to be that the canonicaliser assumes that the 
code-generator can accept tags that span multiple lines.  Whereas 
the code-generator assumes that all canonicalised tags are on a 
single line.

Since I can think of no reason why we shouldn't allow newlines to 
pass through, this patch fixes the code-generator and is less 
intrusive than the previous one:

Index: lib/Petal/CodeGenerator.pm
===================================================================
RCS file: /var/spool/cvs/Petal/lib/Petal/CodeGenerator.pm,v
retrieving revision 1.46
diff -u -r1.46 CodeGenerator.pm
--- lib/Petal/CodeGenerator.pm  9 Jul 2004 11:09:34 -0000       1.46
+++ lib/Petal/CodeGenerator.pm  10 Sep 2005 20:06:10 -0000
@@ -162,9 +162,9 @@

     foreach $token (@{$tokens})
     {
-        if ($token =~ /$PI_RE/)
+        if ($token =~ /$PI_RE/s)
         {
-           ($token_name) = $token =~ /$PI_RE/;
+           ($token_name) = $token =~ /$PI_RE/s;
            my @atts1 = $token =~ /(\S+)\=\"(.*?)\"/gos;
            my @atts2 = $token =~ /(\S+)\=\'(.*?)\'/gos;
            %token_hash = (@atts1, @atts2);

-------------- next part --------------
#!/usr/bin/perl
use warnings;
use strict;
use lib ('lib');
use Test::More 'no_plan';
use Petal;

$Petal::BASE_DIR     = './t/data/';
$Petal::DISK_CACHE   = 0;
$Petal::MEMORY_CACHE = 0;
$Petal::TAINT        = 1;

my $string = Petal->new ('attribute-newline.xml')->process ();
like ($string, '/>apple</', '');
like ($string, '/>orange</', '');
like ($string, '/>plum</', '');
like ($string, '/>pear</', '');
like ($string, '/>ay bee cee</', '');
like ($string, '/>ex why zed</', '');

-------------- next part --------------
<stuff xmlns:petal="http://purl.org/petal/1.0/">

<div petal:define="foo
string: apple">$foo</div>
<div petal:define="bar string: orange">$bar</div>

<div petal:content="string:
plum">xxx</div>
<div petal:content="string: pear">xxx</div>

<div petal:define="
         a string: ay;
         b string: bee;
         c string: cee;
      "><span petal:replace="a" /> <span petal:replace="b" /> <span petal:replace="c" /></div>

<div petal:define="
         x
         string: ex;
         y string:
         why;
         z string: zed;
      "><span petal:replace="x" /> <span petal:replace="y" /> <span petal:replace="z" /></div>

</stuff>
From josh at narins.net  Fri Sep 23 17:11:35 2005
From: josh at narins.net (Josh Narins)
Date: Fri Sep 23 17:11:57 2005
Subject: [Petal] decode: help
Message-ID: <20050923161135.GA6628@narins.net>

What exactly is it supposed to mean, the phrase "decode:" ?

I also don't have it working as advertised. Not sure whether it is a bug
or not.

extra = 
  [
   {
    c=>1,
    short=>'One Thing',
    selected=>'selected',
   },
   {
    c=>2,
    short=>'Another',
   }
  ]

<select name="c">
  <option petal:repeat="e extra" petal:content="e/short" 
          petal:attributes="value e/c; 
                         selected decode: e/selected 'selected' 'selected'" />
</select>

The selected attribute is correct for c=1, but I get selected="0" for
c=2. That doesn't seem right.

Thanks in advance,
Josh
From wsmith at hbslabs.com  Fri Sep 23 18:34:51 2005
From: wsmith at hbslabs.com (Warren Smith)
Date: Fri Sep 23 18:37:17 2005
Subject: [Petal] decode: help
In-Reply-To: <20050923161135.GA6628@narins.net>
References: <20050923161135.GA6628@narins.net>
Message-ID: <43343CBB.5010806@hbslabs.com>

Josh,

Yes there is a bug. Here's the fix:

cvs diff: Diffing .
cvs diff: Diffing lib
cvs diff: Diffing lib/Petal
cvs diff: Diffing lib/Petal/Utils
Index: lib/Petal/Utils/Decode.pm
===================================================================
RCS file: /home/wsmith/cvs/Petal-Utils/lib/Petal/Utils/Decode.pm,v
retrieving revision 1.1.1.1
diff -r1.1.1.1 Decode.pm
30a31
 >   return undef;
cvs diff: Diffing t
Index: t/25__Decode.t
===================================================================
RCS file: /home/wsmith/cvs/Petal-Utils/t/25__Decode.t,v
retrieving revision 1.1.1.1
diff -r1.1.1.1 25__Decode.t
17a18,19
 > my $str3 = "checked";
 > my $str4 = undef;
22a25,26
 >       str3 => $str3,
 >       str4 => $str4,
47a52,54
 > like($out, qr/decode16: <input[^>]*checked="checked"/, 'decode16');
 > unlike($out, qr/decode17: <input[^>]*checked=/, 'decode17');
 >
cvs diff: Diffing t/data
Index: t/data/25__decode.html
===================================================================
RCS file: /home/wsmith/cvs/Petal-Utils/t/data/25__decode.html,v
retrieving revision 1.1.1.1
diff -r1.1.1.1 25__decode.html
40a41,44
 >       <!--? conditional attributes -->
 >       decode16: <input type="checkbox" tal:attributes="checked 
decode:$str3 'checked' 'checked'" />
 >       decode17: <input type="checkbox" tal:attributes="checked 
decode:$str4 'checked' 'checked'" />
 >





>What exactly is it supposed to mean, the phrase "decode:" ?
>
>I also don't have it working as advertised. Not sure whether it is a bug
>or not.
>
>extra = 
>  [
>   {
>    c=>1,
>    short=>'One Thing',
>    selected=>'selected',
>   },
>   {
>    c=>2,
>    short=>'Another',
>   }
>  ]
>
><select name="c">
>  <option petal:repeat="e extra" petal:content="e/short" 
>          petal:attributes="value e/c; 
>                         selected decode: e/selected 'selected' 'selected'" />
></select>
>
>The selected attribute is correct for c=1, but I get selected="0" for
>c=2. That doesn't seem right.
>
>Thanks in advance,
>Josh
>  
>

From bruno at mkdoc.com  Fri Sep 30 20:13:35 2005
From: bruno at mkdoc.com (Bruno Postle)
Date: Fri Sep 30 20:13:42 2005
Subject: [Petal] Problem with win32 file paths
Message-ID: <20050930191335.GA3886@webarchitects.co.uk>

I was under the impression that Petal plays nicely on Windows, can 
anyone make sense of this bug report:

  http://rt.cpan.org/NoAuth/Bug.html?id=14750

-- 
Bruno