update for newer catalyst unicode changes

[catagits/Catalyst-View-TT.git] / lib / Catalyst / View / TT.pm

diff --git a/lib/Catalyst/View/TT.pm b/lib/Catalyst/View/TT.pm
index 39725f9..db72b25 100644 (file)
--- a/lib/Catalyst/View/TT.pm
+++ b/lib/Catalyst/View/TT.pm
@@ -453,15 +453,42 @@ If you are calling C<render> directly then you can specify dynamic paths by
 having a C<additional_template_paths> key with a value of additional directories
 to search. See L<CAPTURING TEMPLATE OUTPUT> for an example showing this.
 
-=head2 Unicode
+=head2 Unicode (pre Catalyst v5.90080)
 
 B<NOTE> Starting with L<Catalyst> v5.90080 unicode and encoding has been
 baked into core, and the default encoding is UTF-8.  The following advice
-is for older versions of L<Catalyst>
+is for older versions of L<Catalyst>.
 
 Be sure to set C<< ENCODING => 'utf-8' >> and use
 L<Catalyst::Plugin::Unicode::Encoding> if you want to use non-ascii
-characters (encoded as utf-8) in your templates.
+characters (encoded as utf-8) in your templates.  This is only needed if
+you actually have UTF8 literals in your templates and the BOM is not
+properly set.  Setting encoding here does not magically encode your
+template output.  If you are using this version of L<Catalyst> you need
+to all the Unicode plugin, or upgrade (preferred)
+
+=head2 Unicode (Catalyst v5.90080+)
+
+This version of L<Catalyst> will automatically encode your body output
+to UTF8. This means if your variables contain multibyte characters you don't
+need top do anything else to get UTF8 output.  B<However> if your templates
+contain UTF8 literals (like, multibyte characters actually in the template
+text), then you do need to either set the BOM mark on the template file or
+instruct TT to decode the templates at load time via the ENCODING configuration
+setting.  Most of the time you can just do:
+
+    MyApp::View::HTML->config(
+        ENCODING => 'UTF-8');
+
+and that will just take care of everything.  This configuration setting will
+force L<Template> to decode all files correctly, so that when you hit
+the finalize_encoding step we can properly encode the body as UTF8.  If you
+fail to do this you will get double encoding issues in your output (but again,
+only for the UTF8 literals in your actual template text.)
+
+Again, this ENCODING configuration setting only instructs template toolkit
+how (and if) to decode the contents of your template files when reading them from
+disk.  It has no other effect.
 
 =head2 RENDERING VIEWS