html-template-users — Discussion of HTML::Template and related modules

Re: [htmltmpl] how to determine if tmpl_var exists in a loop?

[Karen J. C. <si...@ph...>]

On Sat, 17 Apr 2004, Puneet Kishor wrote:

PK>Interesting. I have approached this always with the attitude that I
PK>have to separate programming from display as much as possible. I've

Yes, but is the data part of the "programming," or part of the "display?"
In the case of this project, the data definition is part of the
programming for a *different* set of scripts, but this script is the UI
for settings maintenance (for/by the user).  Because the
nontechnical-user--friendliness is the paramount issue, it's the HTML
designer who gets to decide how the stuff is presented as far as layout,
the order things get handled on, and so on, so the script has to follow
the pattern set by the display in this case.

PK>sure... the user's actions would arrive to you (the cgi) as cgi params.
PK>Most all decisions can be made based on those. In fact, even before
PK>that, quite a few illogical user acts can be trapped with judicious use
PK>of javascript.

I'm not so concerned about illogical, but rather about hostile.  Because
it's going to be a public Internet app, I have to assume that some users
are going to look at the code and submit data that deliberately violates
it.  So the script has to know what list selections are permissible
(presently I feed those *from* the script via TMPL_LOOP), things like
that.  Client-side checking absolutely can't be trusted to do that.

-- 
Karen J. Cravens  si...@ph...

Re: [htmltmpl] how to determine if tmpl_var exists in a loop?

[Puneet K. <pk...@ei...>]

On Apr 17, 2004, at 11:25 AM, Karen J. Cravens wrote:

> On Sat, 17 Apr 2004, Puneet Kishor wrote:
>
> PK>What is/are a typical circumstance(s) where I would not already know
> PK>what vars exist in a template and would want to use query to find 
> out?
>
> It's the split between template and programming thing.  I've considered

On Apr 17, 2004, at 11:18 AM, Roger Burton West wrote:

> On Sat, Apr 17, 2004 at 11:10:35AM -0500, Puneet Kishor wrote:
>
>> What is/are a typical circumstance(s) where I would not already know
>> what vars exist in a template and would want to use query to find out?
>
> The most obvious case is the one in which someone else is being given
> the HTML design part of the job.

Interesting. I have approached this always with the attitude that I 
have to separate programming from display as much as possible. I've 
just finished a relatively complicated app -- a single cgi script about 
2000 lines long, and about 45 different templates. My approach has been 
to create the templates in such a way that they could exist as a 
standalone website even if there was no H-T involved. In other words, 
if a designer-type decided to load the entire templates folder with its 
stylesheets and javascripts and whatnot in a designer-type program 
(GoLive or Dreamweaver, etc.), a complete, well-formed website would 
manifest, of course, with tmpl_vars intact. It may look a bit kooky 
since tables won't be fleshed out, and a few unavoidable tmpl_ifs won't 
be logic-ed out, but nonetheless -- it would be a website with all its 
links and scripts and behaviors and whatnot.



On Apr 17, 2004, at 11:25 AM, Karen J. Cravens wrote:

>
> I can't speak to the original poster's exact situation, but I've been
> mulling over something along these lines.  Problem *I* have is that the
> script really needs to know more than just the parm name... it should 
> know
> whether the end-user obeyed the form length, actually picked one of the
> SELECT options or hand-built a bogus submission, things like that.

sure... the user's actions would arrive to you (the cgi) as cgi params. 
Most all decisions can be made based on those. In fact, even before 
that, quite a few illogical user acts can be trapped with judicious use 
of javascript.

On Apr 17, 2004, at 11:18 AM, Roger Burton West wrote:

> A secondary case is general debugging - if the template passes the
> syntax check but doesn't show the expected results, it's helpful to be
> able to show H::T's idea of its internal structure.
>
ok. This might make sense. I've usually just rolled my own debug 
breakpoints simply because I am really scared of the perl debugger.

Thanks folks, for your insights.

Re: [htmltmpl] how to determine if tmpl_var exists in a loop?

[petersm <pe...@ve...>]

Puneet Kishor <pk...@ei...>
> 
> I am curious too... I've been using H-T for about a year and a half, 
> and I've never had to query the template. My assumption is that params 
> in the template are assigned by me, so I would automatically know what 
> vars exist in the template.
> 
> What is/are a typical circumstance(s) where I would not already know 
> what vars exist in a template and would want to use query to find out?
> 
> Thanks.

Querying the template allows you to be able to create another layer of
abstraction between the code and the template. We use it to create a custom
'TemplateLoader' that will look at a given template, and by the naming
convention that we use on the vars and loops it will actually create the SQL
statements, query the database and load the templates. The only thing we have
to remember is to use the nameing convention for the vars. This works for
about 95% of the occassions that we use H::T and then the other %5 we use it
more straight forward. 

The benefits of this are that if we want to add some more data to the template
(an extra column or field or something) we just add the correct TMPL_VAR to
the template and then the TemplateLoader does the rest. We don't have to
change the code.

If you're curiosity is piqued, then you can check out how we do this at
cheesepizza.venzia.com. The module in question is named
'CheesePizza::TemplateLoader'.

Michael Peters
Venzia

Re: [htmltmpl] how to determine if tmpl_var exists in a loop?

[Karen J. C. <si...@ph...>]

On Sat, 17 Apr 2004, Puneet Kishor wrote:

PK>What is/are a typical circumstance(s) where I would not already know
PK>what vars exist in a template and would want to use query to find out?

It's the split between template and programming thing.  I've considered
this myself... suppose you're wanting to build a form, with a great deal
of customization involved layout/graphics wise (so that none of the
form-building tools are really adequate).  Suppose further that the script
doesn't really care (much) what's in the form, because its only job is to
save the settings in a file.

I can't speak to the original poster's exact situation, but I've been
mulling over something along these lines.  Problem *I* have is that the
script really needs to know more than just the parm name... it should know
whether the end-user obeyed the form length, actually picked one of the
SELECT options or hand-built a bogus submission, things like that.

I haven't quite figured out whether I should write something that analyzes
a read-in template to produce verification code (and if so, whether it
should happen on the fly or not), use a meta-template to produce both the
verification code and HTML fragments, or what.

-- 
Karen J. Cravens  si...@ph...

Re: [htmltmpl] how to determine if tmpl_var exists in a loop?

[Roger B. W. <ro...@fi...>]

On Sat, Apr 17, 2004 at 11:10:35AM -0500, Puneet Kishor wrote:

>What is/are a typical circumstance(s) where I would not already know 
>what vars exist in a template and would want to use query to find out?

The most obvious case is the one in which someone else is being given
the HTML design part of the job.

A secondary case is general debugging - if the template passes the
syntax check but doesn't show the expected results, it's helpful to be
able to show H::T's idea of its internal structure.

Roger

Re: [htmltmpl] how to determine if tmpl_var exists in a loop?

[Puneet K. <pk...@ei...>]

On Apr 17, 2004, at 10:39 AM, <al...@lo...> wrote:

> I don't know what you are trying to do,
> but that doesn't seem like a good thing.
> It feels like you're making the code overly complex.
> You should think it through thoroughly to see
> if theres a better way.
>
> Alan.
>
>
> ----- Original Message -----
> From: "Kapoor, Nishikant" <nis...@xc...>
> To: "H::T (E-mail)" <htm...@li...>
> Sent: Friday, April 16, 2004 6:39 AM
> Subject: [htmltmpl] how to determine if tmpl_var exists in a loop?
>
>
> I have used
>
> if ($template->query(name => 'tmplVar')) {}
>
> to check if a tmpl_var exists in the template or not.
>
> How/what do I use for a tmpl_var inside a tmpl_loop?
>
>

I am curious too... I've been using H-T for about a year and a half, 
and I've never had to query the template. My assumption is that params 
in the template are assigned by me, so I would automatically know what 
vars exist in the template.

What is/are a typical circumstance(s) where I would not already know 
what vars exist in a template and would want to use query to find out?

Thanks.

Re: [htmltmpl] how to determine if tmpl_var exists in a loop?

[<al...@lo...>]

I don't know what you are trying to do,
but that doesn't seem like a good thing.
It feels like you're making the code overly complex.
You should think it through thoroughly to see
if theres a better way.

Alan.


----- Original Message -----=20
From: "Kapoor, Nishikant" <nis...@xc...>
To: "H::T (E-mail)" <htm...@li...>
Sent: Friday, April 16, 2004 6:39 AM
Subject: [htmltmpl] how to determine if tmpl_var exists in a loop?


I have used=20

if ($template->query(name =3D> 'tmplVar')) {}

to check if a tmpl_var exists in the template or not.

How/what do I use for a tmpl_var inside a tmpl_loop?

Thanks
Nishi



-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id638&op=3Dick
_______________________________________________
Html-template-users mailing list
Htm...@li...
https://lists.sourceforge.net/lists/listinfo/html-template-users

RE: [htmltmpl] how to determine if tmpl_var exists in a loop?

[Kapoor, N. <nis...@xc...>]

Sorry, I should have checked the docs first. Anyway, thanks for the =
help.=20

This is exactly what I was looking for

  if ($template->query(name =3D> ['tmplLoop', 'tmplVar'])) {}

Thanks
Nishi

> -----Original Message-----
> From: Roger Burton West [mailto:ro...@fi...]
> Sent: Friday, April 16, 2004 9:02 AM
> To: H::T (E-mail)
> Subject: Re: [htmltmpl] how to determine if tmpl_var exists in a loop?
>=20
>=20
> On Fri, Apr 16, 2004 at 08:39:05AM -0500, Kapoor, Nishikant wrote:
> >I have used=20
> >if ($template->query(name =3D> 'tmplVar')) {}
> >to check if a tmpl_var exists in the template or not.
> >How/what do I use for a tmpl_var inside a tmpl_loop?
>=20
> Recurse or append parameters. See the documentation, under:
>=20
>        "query()" also allows you to get a list of parameters
>        inside a loop (and inside loops inside loops).
>=20
> which explains it very clearly.
>=20
> Roger
>=20
>=20
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> =
administration.http://ads.osdn.com/?ad_id=3D1470&alloc_id=3D3638&op=3Dcli=
ck
> _______________________________________________
> Html-template-users mailing list
> Htm...@li...
> https://lists.sourceforge.net/lists/listinfo/html-template-users
>=20
>=20

Re: [htmltmpl] how to determine if tmpl_var exists in a loop?

[Roger B. W. <ro...@fi...>]

On Fri, Apr 16, 2004 at 08:39:05AM -0500, Kapoor, Nishikant wrote:
>I have used 
>if ($template->query(name => 'tmplVar')) {}
>to check if a tmpl_var exists in the template or not.
>How/what do I use for a tmpl_var inside a tmpl_loop?

Recurse or append parameters. See the documentation, under:

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

which explains it very clearly.

Roger

[htmltmpl] how to determine if tmpl_var exists in a loop?

[Kapoor, N. <nis...@xc...>]

I have used=20

if ($template->query(name =3D> 'tmplVar')) {}

to check if a tmpl_var exists in the template or not.

How/what do I use for a tmpl_var inside a tmpl_loop?

Thanks
Nishi

Re: [htmltmpl] [question] efficient use of HTML::Template->new

[Sam T. <sa...@tr...>]

On Wed, 14 Apr 2004, C Hagstrom wrote:

> But my concern is that the approach would not
> be "server friendly". That said, I'm thinking that the overhead
> for template calls is so small, I need not be concerned.

I'm not sure what you mean by "server friendly", but the only way to
know anything about performance is to measure it.  I recommend you get
comfortable with Benchmark.pm and Devel::DProf.  Using these tools you
can find out exactly how long it takes to call new() on your
templates.

But what the heck, here's a couple hints anyway: new() takes much more
time on larger and more complicated templates.  You can reduce this
time to almost nothing using caching and pre-compiling your templates.

-sam

Re: [htmltmpl] [question] efficient use of HTML::Template->new

[Mathew R. <mat...@re...>]

If you use the caching options of H::T, then wont need to care where you =
call new(), as H::T will automatically use a previously parsed version =
of the template from the cache, ie. the impact of call new() on a =
previously parsed template, is nearly zero.

Read up on the caching options within the H::T perldoc; you should find =
one of the type which would suite your application environment.

regards,
Mathew


> Wondering if anyone has input regarding the use of:
>=20
> HTML::Template->new
>=20
> It would seem (to me) that from a performance standpoint,
> only calling a template as it is needed is the most efficient
> approach.
>=20
> My question has to do with the actual impact of
> HTML::Template->new
>=20
> In a script, when I only use a given template one time,
> I only call it in the one location where it's needed
>=20
> When the template is used globally (every time the
> script is fired off), I call the template at the beginning
> of the script - one time.
>=20
> My question has to do with templates that are potentially
> used in many areas of the script, even though only once.
> In other words, there may be various ways the script
> can be triggered that will require a given template (and
> many ways the script can be called where the template
> isn't needed).
>=20
> To avoid code clutter, what I'd like to do is call all my
> templates in one spot:
>=20
> my $temp1 =3D HTML::Template->new( etc ...
> my $temp2 =3D HTML::Template->new( etc ...
> my $temp3 =3D HTML::Template->new( etc ...
>=20
> But my concern is that the approach would not
> be "server friendly". That said, I'm thinking that the overhead
> for template calls is so small, I need not be concerned.
>=20
> Hopefully, some on the list can enlighten me.
>=20
> Carl Hagstrom
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> =
administration.http://ads.osdn.com/?ad_id=3D1470&alloc_id=3D3638&op=3Dcli=
ck
> _______________________________________________
> Html-template-users mailing list
> Htm...@li...
> https://lists.sourceforge.net/lists/listinfo/html-template-users
> 

[htmltmpl] [question] efficient use of HTML::Template->new

[C H. <hag...@ep...>]

Wondering if anyone has input regarding the use of:

HTML::Template->new

It would seem (to me) that from a performance standpoint,
only calling a template as it is needed is the most efficient
approach.

My question has to do with the actual impact of
HTML::Template->new

In a script, when I only use a given template one time,
I only call it in the one location where it's needed

When the template is used globally (every time the
script is fired off), I call the template at the beginning
of the script - one time.

My question has to do with templates that are potentially
used in many areas of the script, even though only once.
In other words, there may be various ways the script
can be triggered that will require a given template (and
many ways the script can be called where the template
isn't needed).

To avoid code clutter, what I'd like to do is call all my
templates in one spot:

my $temp1 = HTML::Template->new( etc ...
my $temp2 = HTML::Template->new( etc ...
my $temp3 = HTML::Template->new( etc ...

But my concern is that the approach would not
be "server friendly". That said, I'm thinking that the overhead
for template calls is so small, I need not be concerned.

Hopefully, some on the list can enlighten me.

Carl Hagstrom

Re: [htmltmpl] [Question] H::T and DBI => stuffing query info into <TMPL_LOOP>'s

[Mathew R. <mat...@re...>]

> >What has Perl's string interpolation got to do with DBI?
> >
> >"SELECT * FROM $working_tbl WHERE first=3D'one' ORDER BY Location"
> >
> >The variable $working_tbl will be interpolated by Perl, long before =
DBI=20
> >ever gets to see the SQL statement, due to the double-quote.
>=20
> $working_tbl may come from some kind of user input, which opens you up =
to=20
> SQL injection attacks.  Most databases don't allow placeholders for =
the=20
> table name.  It's best to use DBI->quote() whenever you absolutely =
must=20
> interpolate a variable into an SQL statement.

no - you should never take any kind of user input without doing input =
validation -> otherwsie you are asking for trouble.

Also, DBI->quote() will not help with the example written above, as Perl =
will interpolate the vairable into to string, due to the double-quotes.  =
You are correct that most databases wont use variable table names, but =
the syntax was not written like:

"SELECT * FROM ? WHERE ..."

Mathew

Re: [htmltmpl] [Question] H::T and DBI => stuffing query info into <TMPL_LOOP>'s

[Timm M. <tm...@ag...>]

At 09:21 AM 4/14/04 +1000, Mathew Robertson wrote:
> > > I can get the results to print directly with this code:
> > >
> > > $sth = $dbh->prepare ("SELECT * FROM $working_tbl WHERE first='one'
> > > ORDER BY Location");
> > > $sth->execute ();
> > > while ( my $hash_ref = $sth->fetchrow_hashref ) {
> > >         print "ID: $hash_ref->{'ID'} -- Name: $hash_ref->{'Name'}
> > > (etc....) <br>\n";
> > > }
>
> > this is cool, but do note that the DBI docs generally advise against the
> > use of a variable table name (well... I understand them to be advising
> > against) as it doesn't allow for creating a plan and optimizing the
> > query. Anyway...
>
>What has Perl's string interpolation got to do with DBI?
>
>"SELECT * FROM $working_tbl WHERE first='one' ORDER BY Location"
>
>The variable $working_tbl will be interpolated by Perl, long before DBI 
>ever gets to see the SQL statement, due to the double-quote.

$working_tbl may come from some kind of user input, which opens you up to 
SQL injection attacks.  Most databases don't allow placeholders for the 
table name.  It's best to use DBI->quote() whenever you absolutely must 
interpolate a variable into an SQL statement.


>Mathew
>
>
>-------------------------------------------------------
>This SF.Net email is sponsored by: IBM Linux Tutorials
>Free Linux tutorial presented by Daniel Robbins, President and CEO of
>GenToo technologies. Learn everything from fundamentals to system
>administration.
>_______________________________________________
>Html-template-users mailing list
>Htm...@li...
>https://lists.sourceforge.net/lists/listinfo/html-template-users

Re: [htmltmpl] problems with scalarref

[Jeff C. <oka...@po...>]

On Tue, 13 Apr 2004 14:01:35 -0400 (EDT) , Sam Tregar <sa...@tr...>  wrote:

>On Tue, 13 Apr 2004, Jeff Capes wrote:

>

>> HTML::Template->new() : Syntax error in <TMPL_*> tag at

>> /fake/path/for/non/file/template : 1. at

>> /usr/lib/perl5/site_perl/5.8.0/HTML/Template.pm line 2243.

>

>Any chance you're running Perl 5.8.0 on Redhat 8 or Redhat 9?  If so,

>you might have a broken Perl install.



Red Hat 9.



>

>> my $tmpl;

>>

>> $tmpl = &get_template($tmpl);

>

>What do you see in your error log if you insert this here:

>

>  print STDERR "-----TEMPLATE START----\n\n$tmpl\n\n---TEMPLATE END----\n";

>

>I'm interested to see if $tmpl contains what you think it does.



It did.



I've tested a workaround for the problem by setting LANG=en_GB.

The workaround worked, so I edited /etc/sysconfig/i18n to make it permanent. This also fixed Date::Manip.



Thanks to everyone who helped!


___________________________________________________ 
Reduce your company's IT costs today with Officemaster. Sign up for a free trial!

http://www.officemaster.net

Re: [htmltmpl] [Question] H::T and DBI => stuffing query info into <TMPL_LOOP>'s

[Puneet K. <pk...@ei...>]

On Apr 13, 2004, at 6:21 PM, Mathew Robertson wrote:

>>> I can get the results to print directly with this code:
>>>
>>> $sth = $dbh->prepare ("SELECT * FROM $working_tbl WHERE first='one'
>>> ORDER BY Location");
>>> $sth->execute ();
>>> while ( my $hash_ref = $sth->fetchrow_hashref ) {
>>>         print "ID: $hash_ref->{'ID'} -- Name: $hash_ref->{'Name'}
>>> (etc....) <br>\n";
>>> }
>
>> this is cool, but do note that the DBI docs generally advise against 
>> the
>> use of a variable table name (well... I understand them to be advising
>> against) as it doesn't allow for creating a plan and optimizing the
>> query. Anyway...
>
> What has Perl's string interpolation got to do with DBI?
>
> "SELECT * FROM $working_tbl WHERE first='one' ORDER BY Location"
>
> The variable $working_tbl will be interpolated by Perl, long before 
> DBI ever gets to see the SQL statement, due to the double-quote.
>
>

you are correct, and my advise was half-baked. The assertion I made 
holds true when doing multiple queries using bind variables, and hence, 
I try to stay away from variable-izing table names... that is where my 
advise came from and it really doesn't apply to Carl's case. But, as 
they say... you gotta do what you gotta do.

Re: [htmltmpl] [Question] H::T and DBI => stuffing query info into <TMPL_LOOP>'s

[Mathew R. <mat...@re...>]

> > I can get the results to print directly with this code:
> >=20
> > $sth =3D $dbh->prepare ("SELECT * FROM $working_tbl WHERE =
first=3D'one'=20
> > ORDER BY Location");
> > $sth->execute ();
> > while ( my $hash_ref =3D $sth->fetchrow_hashref ) {
> >         print "ID: $hash_ref->{'ID'} -- Name: $hash_ref->{'Name'} =20
> > (etc....) <br>\n";
> > }

> this is cool, but do note that the DBI docs generally advise against =
the=20
> use of a variable table name (well... I understand them to be advising =

> against) as it doesn't allow for creating a plan and optimizing the=20
> query. Anyway...

What has Perl's string interpolation got to do with DBI?=20

"SELECT * FROM $working_tbl WHERE first=3D'one' ORDER BY Location"

The variable $working_tbl will be interpolated by Perl, long before DBI =
ever gets to see the SQL statement, due to the double-quote.

Mathew

Re: [htmltmpl] problems with scalarref

[Sam T. <sa...@tr...>]

On Tue, 13 Apr 2004, Jeff Capes wrote:

> HTML::Template->new() : Syntax error in <TMPL_*> tag at
> /fake/path/for/non/file/template : 1. at
> /usr/lib/perl5/site_perl/5.8.0/HTML/Template.pm line 2243.

Any chance you're running Perl 5.8.0 on Redhat 8 or Redhat 9?  If so,
you might have a broken Perl install.

> my $tmpl;
>
> $tmpl = &get_template($tmpl);

What do you see in your error log if you insert this here:

  print STDERR "-----TEMPLATE START----\n\n$tmpl\n\n---TEMPLATE END----\n";

I'm interested to see if $tmpl contains what you think it does.

-sam

Re: Summary => RE: [htmltmpl] [Question] H::T and DBI

[Puneet K. <pk...@ei...>]

C Hagstrom wrote:

> Thanks to all for the suggestions
> (the speed and number of the responses was great)
> 
> I'm going to run with Puneet's suggestion:
> 
>   # Fetch the results as a ref to an AofH
>   my $res = $sth->fetchall_arrayref({});
>   # Assign the ref to a var in the template
>   $t->param(RES => $res);
> 
> for the simple reason:
> 
> "Specify the column names you want to retrieve,
> and the column names automatically become
> the names of the TMPL_VARs."
> 
> This seems the simplest way to go for my application
>  From what I understand, working with hash_ref 's
> is supposed to be inefficient (performance-wise),
> but in my case, I'm pursuing such a light
> weight database that I don't think it's an issue.


ya, possibly. As you say, given your situation (and, I daresay, for most 
common situations), simplicity weighs heavier than speed.

You can always achieve more speed by tweaking other things (faster 
database, better db structure, cached templates, mod_perl or some other 
persistent perl, faster hardware and more memory).

I always, always, err toward simplicity.

Good luck.

Summary => RE: [htmltmpl] [Question] H::T and DBI

[C H. <hag...@ep...>]

Thanks to all for the suggestions
(the speed and number of the responses was great)

I'm going to run with Puneet's suggestion:

   # Fetch the results as a ref to an AofH
   my $res = $sth->fetchall_arrayref({});
   # Assign the ref to a var in the template
   $t->param(RES => $res);

for the simple reason:

"Specify the column names you want to retrieve,
and the column names automatically become
the names of the TMPL_VARs."

This seems the simplest way to go for my application
 From what I understand, working with hash_ref 's
is supposed to be inefficient (performance-wise),
but in my case, I'm pursuing such a light
weight database that I don't think it's an issue.


Carl Hagstrom
hag...@ep...	

Re: [htmltmpl] [Question] H::T and DBI => stuffing query info into <TMPL_LOOP>'s

[Keith J. <kja...@ey...>]

Here is a code snippet that uses fetchrow_hashref and pushs the records
into an array to pass to a template...

------------------------------------
my $sth = $dbh->prepare(qq(select * from shopping_cart));
$sth->execute();
while($hash_ref = $sth->fetchrow_hashref()){
	push(@records, $hash_ref);
}

use HTML::Template;
my $template = new HTML::Template(filename=>cart.html);
$template->param(records => \@records);
print $template->output();
-------------------------------------

Hope that helps....



On Tue, 2004-04-13 at 10:38, C Hagstrom wrote:
> I'm hoping for some conceptual help here....
> I have tried to find references to my question without
> success ( the list-archive link appears not to be working).
> 
> System info: Linux - Perl - MySql
> 
> I'm fairly comfortable at this stage with H::T,
> and am also fairly comfortable with the DBI/MySql
> interface.
> 
> I have a script that queries the database, and pulls
> information from numerous records ... in other words,
> it grabs and displays all the record information
> for a selected group of records.
> 
> I can get the results to print directly with this code:
> 
> $sth = $dbh->prepare ("SELECT * FROM $working_tbl WHERE first='one' ORDER 
> BY Location");
> $sth->execute ();
> while ( my $hash_ref = $sth->fetchrow_hashref ) {
>          print "ID: $hash_ref->{'ID'} -- Name: 
> $hash_ref->{'Name'}  (etc....) <br>\n";
> }
> 
> And before I became an H::T convert, I would use
> the above approach, and craft the HTML output
> directly in the script.
> 
> Now that I am an H::T convert, I'm looking for guidance
> on how best to handle the database output so I can
> take advantage of the <TMPL_LOOP> feature in H::T
> 
> I tried googling, and did find a few references on this topic,
> but I had difficulty understanding them. I'm no wizard when
> it comes to the object oriented side of things, or MySql
> retrievals in general.
> 
> I have found that (for me) retrieving data using fetchrow_hashref
> is easier 'cause the info can be gleaned using the field names,
> where (it seems to me), fetchrow_array is based on the numerical
> order/array-reference of the items, and is easier for me to screw up.
> 
> At any rate, I think I'm close to understanding it, but am hoping
> someone can explain the process in a pretty basic way. I think
> I understand the fundamentals, but am really struggling
> trying to understand how the info gets tossed into some
> sort of container so it can be sent out to a <TEMPL_LOOP>
> call in the related template.
> 
> 
> Carl Hagstrom
> 
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> _______________________________________________
> Html-template-users mailing list
> Htm...@li...
> https://lists.sourceforge.net/lists/listinfo/html-template-users
-- 
Keith Jackson <kja...@ey...>

Re: [htmltmpl] problems with scalarref

[Roger B. W. <ro...@fi...>]

On Tue, Apr 13, 2004 at 12:43:55PM +0100, Jeff Capes wrote:
>I'm having problems with scalarref.

I'm not. Running Perl 5.6.1, H::T 2.5, and your code (modified only to
change the filename), I get:

### HTML::Template Debug ### In _parse:
### HTML::Template Debug ### /fake/path/for/non/file/template : line 2 :
LOOP employee_info start
### HTML::Template Debug ### /fake/path/for/non/file/template : line 4 :
parsed VAR name
### HTML::Template Debug ### /fake/path/for/non/file/template : line 6 :
parsed VAR job
### HTML::Template Debug ### /fake/path/for/non/file/template : line 8 :
LOOP end
### HTML::Template Debug ### In output

which is as it should be.

Roger

Re: [htmltmpl] [Question] H::T and DBI => stuffing query info into <TMPL_LOOP>'s

[petersm <pe...@ve...>]

C Hagstrom wrote:
> I'm hoping for some conceptual help here....
> I have tried to find references to my question without
> success ( the list-archive link appears not to be working).
> 
> 
> I can get the results to print directly with this code:
> 
> $sth = $dbh->prepare ("SELECT * FROM $working_tbl WHERE first='one' ORDER 
> BY Location");
> $sth->execute ();
> while ( my $hash_ref = $sth->fetchrow_hashref ) {
>          print "ID: $hash_ref->{'ID'} -- Name: 
> $hash_ref->{'Name'}  (etc....) <br>\n";
> }
> 
> 
> Now that I am an H::T convert, I'm looking for guidance
> on how best to handle the database output so I can
> take advantage of the <TMPL_LOOP> feature in H::T


you can use something like this to get the structure from DBI

$sth = $dbh->prepare("SELECT FirstName, LastName FROM Person WHERE ...");
#this will get an array of hashrefs
my $result = $sth->fetchall_arrayref({});

then you would put that structure into your template like so...
my $tmpl->HTML::Template->new(..);
$tmpl->param('Person_Loop' => $result);

and then your template could look like this

<TMPL_LOOP Person_Loop>
  Hello, My name is <TMPL_VAR FirstName> <TMPL_VAR LastName>.<br />
</TMPL_LOOP>

Does that help?

Michael Peters
Venzia

Re: [htmltmpl] [Question] H::T and DBI => stuffing query info into <TMPL_LOOP>'s

[Puneet K. <pk...@ei...>]

C Hagstrom wrote:

> I'm hoping for some conceptual help here....
..
> 
> I have a script that queries the database, and pulls
> information from numerous records ... in other words,
> it grabs and displays all the record information
> for a selected group of records.
> 
> I can get the results to print directly with this code:
> 
> $sth = $dbh->prepare ("SELECT * FROM $working_tbl WHERE first='one' 
> ORDER BY Location");
> $sth->execute ();
> while ( my $hash_ref = $sth->fetchrow_hashref ) {
>         print "ID: $hash_ref->{'ID'} -- Name: $hash_ref->{'Name'}  
> (etc....) <br>\n";
> }
> 

this is cool, but do note that the DBI docs generally advise against the 
use of a variable table name (well... I understand them to be advising 
against) as it doesn't allow for creating a plan and optimizing the 
query. Anyway...


> And before I became an H::T convert, I would use
> the above approach, and craft the HTML output
> directly in the script.
> 
> Now that I am an H::T convert, I'm looking for guidance
> on how best to handle the database output so I can
> take advantage of the <TMPL_LOOP> feature in H::T

I have posted several times about how to do this using a very simple 
approach -- the fetchall_arrayref({}) method in DBI. This fetches a ref 
to an array of hashes... exactly what the TMPL_LOOP wants.

   # Create the sql query
   my $sql = <<SQL;
   SELECT col1, col2, col3
   FROM my_working_tbl
   WHERE first = 'one'
   ORDER BY Location
   SQL

   # Create the statement handle and execute it
   my $sth = $dbh->prepare(qq{$sql});
   $sth->execute;

   # Fetch the results as a ref to an AofH
   my $res = $sth->fetchall_arrayref({});

   # Assign the ref to a var in the template
   $t->param(RES => $res);

Then in your template --

   <table>
   <tr><th>Col1</th><th>Col2</th><th>Col2</th></tr>
   <TMPL_LOOP RES>
   <tr>
	<td><TMPL_VAR COL1></td>
	<td><TMPL_VAR COL2></td>
	<td><TMPL_VAR COL3></td>
   </tr>
   </TMPL_LOOP>
   </table>

Keep in mind -- you can't use SELECT * (which is inefficient for the 
database anyway). Specify the column names you want to retrieve, and the 
column names automatically become the names of the TMPL_VARs.


I hope this helps. It really doesn't get any simpler and basic than this.

Let us know if you need further help.

Ya, the archive seems to be down, but when it comes back up (if it 
does), there are a lot of good discussions there.

Good luck.