yakovsh

<!--
Recently I had a weird problem while using Tomcat - sessions would get lost when using Microsoft's Internet Explorer.
Since Tomcat stores the session id as JSESSIONID cookie, the session would get lost when the cookie is not stored.
Same behavior occurs with other web servers including IIS and Resin.

After further research, it seems that IE does not store cookies at default privacy settings unless a P3P policy is
provided for the site. In order to bypass this problem, a basic P3P policy needs to defined and included in the site.
I used IBM's P3P editor to generate a basic P3P XML policy file and the following snippet of HTML code to enable it.

This is still true as of 2011: [http://blog.toppingdesign.com/2011/11/18/ie-iframe-p3p-cookies-oh-my/]

yakovsh

/*
 * One of the recent problem that I ran across at work is an easy way to do highlighting for form fields.
 * Of course the best way would be CSS 2
 * (as shown here [http://www.cssdrive.com/index.php/examples/exampleitem/focus_pseudo_class/]) but alas,
 * IE does not support that.
 *
 * The alternative is JavaScript onFocus and onBlur method. However, one thing which I also wanted to avoid
 * is changing over 200 different pages in our system to add those in. So instead, I wrote a simple piece of
 * Javascript which attaches the event handlers on the fly. The best part about this is that the actual form
 * does not need to be changed - just include the Javascript on the bottom of your webpage. The code is as follows

yakovsh

<!--
  Here is how I am wrapping text via XSLT as opposed to CSS and JavaScript. The reason why I am prefering to do this
  server side via XSLT as opposed to JavaScript client-side is because if I will be doing client-side JavaScripts,
  I would have to escape the stuff during the HTML generation anyway in XSLT server-side. So in the same processing
  power I might as well wrap it instead. Of course you can also do it via some fancy JavaScript client side by going
  through the entire table and chopping up the text in each row, but that would make things look rather interesting
  to the user (tables automatically re-arranging themselves).
  
  Anyway, here is a simple recursive template based on the code posted by Andrew Welch:
  https://web.archive.org/web/20070116081723/http://www.xslt.com/html/xsl-list/2002-03/msg00334.html

yakovsh

/*
  One of our customers had recently requested a feature to make all links look the same, visited or unvisited.
  While debates and complaints have been raised about whether they should look the same, in our case this was
  necessary. A web application such as ours usually dynamically generated all the content. That means that
  links may stay static but their content changes. Therefore, visited links make no sense in a dynamic system.
  
  In any case, the way I made them look the same is via the following piece of CSS
*/

A:link {color: blue}

yakovsh

<!--
    While working with XSLT templates, I came across an interesting problem. I am using an XSLT template
    to transform an XML file into HTML. However, for debugging purposes I need to see the original XML
    and since the generation process is done on a web server (like Resin does), it is not easy to get it.
    The solution: display the original XML file inside the output HTML itself. As it turns out, this was
    not easy since it requires to change all "<" and ">" to use entities like "<" and ">". In XSLT,
    the solution looks as follows (another solution would be to use JavaScript to escape this client-side)
-->

<xsl:template match="*">

yakovsh

# During a recent project, I ran into an issue when handling Unicode data in metadata headers in Amazon S3.
# Apparently, Amazon adds on "?UTF-8?B?" in front of any Unicode data and "?=" in end of the data.
# I could not find any existing standard that describes this or why it is done, but I surmise this probably
# has to do with Base-64 encoding and how it handles Unicode.
#
# As per @rawnsley:
# apparently this is because HTTP headers must only be encoded in ASCII: http://stackoverflow.com/a/4410331/671393
#
# An easy Perl hack to get around this is as following (assuming you are using MIME::Base64 module):

yakovsh

#!/usr/bin/perl
#
# Here is a short script that can mass delete files in an Amazon S3 bucket. It is limited to a 1,000 keys at a time
#

use Net::Amazon::S3;

my $s3 = Net::Amazon::S3->new({
  aws_access_key_id     => 'ACCESS_ID',
  aws_secret_access_key => 'ACCESS_KEY',

yakovsh

#
# Here is a short way to cleanup bad HTML input and convert to XML with Perl:
#

use HTML::TreeBuilder;
use XML::LibXML;

$html_code = '';

my $builder = HTML::TreeBuilder->new();

yakovsh

# Here is another way to cleanup bad HTML with Perl, and convert to XML:
# This approach relies on the HTML::DOMbo module to do the actual conversion
# between HTML and XML, and HTML::TreeBuilder for parsing.

use HTML::DOMbo;
use HTML::TreeBuilder;
use XML::LibXML;

$html_code = '';

yakovsh

# Quick way to fix the following error in Perl:
#
#    :1: parser error : Input is not proper UTF-8, indicate encoding !
#    Bytes: 0xA0 0x20 0xA0 0x3C
#
# Use this command:
#
use Encode:
$string1 = decode("UTF-8", $input);