Hi,
Thanks for the idea, it works well and it doesn't increase the number of
templates.
The user selects the download button, the query is run again and the output
is processed by H::T with a 'plaintext' parameter. The resulting 'excel'
output is run through HTML::Clean to remove empty lines and superfluous
spaces created by my indenting the template.
The resulting output is written to a temporary file and mailed to the user as
an attachment (using Mail::Sendmail).
On Tuesday 05 March 2002 16:40, David Ferrance wrote:
> Definitely don't use option 2-- it is (basically) equivalent to option 1
> but with more work.
>
> Methods 1 and 2 both potentially open you up to malicious bad people seeing
> other
> people's results.
>
> How hard are the queries? If you can afford to run them twice, you might
> do what I did
> in this situation. I have only one template, with this structure:
>
> <TMPL_IF NAME="EXCEL">
> <pre>
> blah,blah,csv,blah
> </pre>
> <TMPL_ELSE>
> <html>
> <p>words words words</p>
> </html>
> </TMPL_IF>
>
> So in the HTML I have a link to the same PL file, but this time with an
> extra flag to tell it
> to do $t->param(EXCEL=>1), and the output is something they can save from
> their browser
> and import into excel.
>
> None of it is terribly elegant, but if you can afford to run the query
> twice, it _does_
> simplify the code.
>
> -Dave
>
> PS I am hoping one of you posts the pretty way to do this.
>
>
> ----- Original Message -----
> From: "Martijn van den Burg" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED] >
> Sent: Tuesday, March 05, 2002 10:39 AM
> Subject: [htmltmpl] Downloading data to file. How?
>
> > Hi,
> >
> > Problem: I want the $USER to have the possibility to download the data
>
> that
>
> > is displayed (using H::T) in his browser to a file in his own directory.
> > Without the fancy HTML formatting. I'm talking about tables with numbers
> > obtained from a MySQL database.
> >
> >
> > Purpose: data download to, eg. M$ Excel, for further manipulation.
> >
> >
> > Precondition: I'd really like to reuse the same datastructures I now send
>
> to
>
> > H::T.
> >
> >
> > Possible solutions:
> >
> > (1) Always save each query output also to file on the webserver (with a
> > unique ID) using a dedicated 'ASCII' template and
>
> $template->ouput(print_to
>
> > => *STDOUT) and then provide the $USER with a d/l link for that unique
>
> file.
>
> > Instead of table formatting I'd use CSV formatting.
> >
> > Pro: relatively easy to do
> > Con: pollution of the webserver: need to remove 'stale' files
> >
> > (2) Don't use H::T and instead make separate subroutines for each
> > dataset. Downloading them to the user with something like this:
> >
> > print $query->header('-type'=>'application/octet-stream');
> > open (FILE, "<${docdir}/${file_name}") or fatal_error("$!\n");
> > foreach (@lines_of_formatted_data) {
> > print FILE "$_\n";
> > }
> > close (FILE);
> >
> >
> > (3) Use a CPAN module someone else wrote but which is as yet unknown to
>
> me.
>
> > Your input is much appreciated.
> >
> > Martijn
> >
> > --
> > +----------------------------------+
> > Martijn van den Burg
> > ASML, CS PPC Inform@tion Systems
> > Voice: +31 (0)40 268 3856
> > +----------------------------------+
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
--
+----------------------------------+
Martijn van den Burg
ASML, CS PPC Inform@tion Systems
Voice: +31 (0)40 268 3856
+----------------------------------+
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
