Installing Eclipse Plugin Error

Okay so today I was trying to play with installing the CheckStyle plugin for my Eclipse 3.4 environment. This is a relatively painless process normally however I kept getting:
"An error occurred while collecting items to be installed No repository found containing: org.apache.oro/osgi.bundle/2.0.8.v200903061218"

I couldn't see any dependencies or where this was getting pulled in. Finally, I went to the Available Software tab and checked Manage Sites, here I found that even though I was only requesting to install from http://eclipse-cs.sf.net/update/ there were a number of other sites checked. I removed all checks except the checkstyle one and the install worked great.

If your not using checkStyle it's a great tool for forcing coding standards check it out.

Upgrading Dell E1505 hard disk

Well this was more than a simple task with many lessons painfully learned. Hopefully this will help someone out there to not have to learn them also. So the goal was pretty simple I had a 60Gb hard drive pretty full and wanted to upgrade to a larger one. The Dell was three years old and I now know was running Media Center 2. There are plenty of blogs and sites out there that now reference the issue with just cloning the hard drive, unfortunately I didn't see them till I had truncated my 320 Gb drive to 60Gb. If you have done this already the tool to fix it is here just make sure you are running it on a 32bit processor and yes you will need to plug it into a desktop since it can't work over usb. (At least at this writing).

Okay so here is how I finally was successful, may be other ways but this is what I did.
Assuming you purchased your new drive, you have an external drive casing so you can connect your disk via usb and that you have the Dell media CD to reinstall the OS. If this is not the case go out to Dell.com I believe you can order them.

1. Pull out old drive and put in new one, this is just a few screws, but Dell.com support can help you if you are not familiar.
2. Put in CD and boot to install OS.
3. Partition new drive:
I made two partitions, c: = partition 250Gb d:= 65GB. format ntfs quick
4. Install OS.
5. Once installed I didn't bother with the drivers, just the base OS.
I put the old drive in the external casing and connected via usb. Fire up cloning software. (Acronis Home Image, Clonezilla, Partition Magic, etc; I was using seagate drives so I used disk wizard from seagate which is a limited version of Acronis Home Image)
Create a partition image of the old c:\ drive and place it in the new d:\drive space. ( Here we are assuming that your old drive partition image will fit in the d:\drive. If not you would probably need a third drive to do it this way. Also note your old drive will have several detected partitions, the largest was your old c:\drive it will have a different drive label now obviously)
6. Using your cloning software Restore from image the newly created partition image to your new hard drive c:\ partition.
7. Reboot and done.

Other notes: if you have space to defragement the drive before hand it is recommended on some blogs.

Hope this helps.

xslt function upper-case() in MSXML transformations

Today I wanted to use xslt to convert some incoming element to upper case so I can then use it in a look up routine. (mentioned in previous blog) Some simple Googling will most likely point to the XPath Version="2.0" function fn:upper-case(string).

However, using Microsoft's MSXML this function is not supported by their transformation engine. However, the fn:translate(...) function is supported so here is a simple way to convert all letters to uppercase.


Here is a sample conversion template:

<?xml version='1.0'?> <xsl:stylesheet version="2.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"> <?xml version='1.0'?> <xsl:stylesheet version="2.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"> <xsl:output method="xml" version="1.0" encoding="iso-8859-1" indent="yes" /> <xsl:variable name="lcletters">abcdefghijklmnopqrstuvwxyz</xsl:variable> <xsl:variable name="ucletters">ABCDEFGHIJKLMNOPQRSTUVWXYZ</xsl:variable> <xsl:template match="convert"> <converted> <toupper><xsl:value-of select="translate(lower,$lcletters,$ucletters)" /></toupper> <tolower><xsl:value-of select="translate(upper,$ucletters,$lcletters)" /></tolower> <touppermixed><xsl:value-of select="translate(mixed,$lcletters,$ucletters)" /></touppermixed> </converted> </xsl:template> </xsl:stylesheet>

Here is a sample xml file:

<?xml version='1.0'?> <convert> <lower>convert to upper</lower> <upper>CONVERT TO LOWER</upper> <mixed>Convert This To All Upper</mixed> </convert>

Here is a sample asp vbscript page:

<%@ Language=VBScript %> <%Option Explicit%> <% Dim xslt Dim xslProc Dim xslDoc Dim oXmlDoc Set oXmlDoc = Server.CreateObject("Microsoft.XMLDOM") oXmlDoc.load(Server.MapPath("convert.xml")) set xslDoc = Server.CreateObject("MICROSOFT.FreeThreadedXMLDOM") xslDoc.async = false xslDoc.load(Server.MapPath("conversion.xsl")) if xslDoc.parseError.errorCode = 0 Then set xslt = Server.CreateObject("Msxml2.XSLTemplate") xslt.stylesheet = xslDoc set xslProc = xslt.createProcessor() xslProc.input = oXmlDoc xslProc.transform() response.write "Conversion =" & Server.htmlencode(xslProc.output) else response.write "Error::" & xslDoc.parseError.reason end if %>

Give it a shot you should end up with something like:

Conversion =<?xml version="1.0" encoding="UTF-16"?> <converted> <toupper>CONVERT TO UPPER</toupper> <tolower>convert to lower</tolower> <touppermixed>CONVERT THIS TO ALL UPPER</touppermixed> </converted>

Lookup Tables in XSLT

Wow what a find today, I needed to do some data scrubbing while doing a transformation to guarantee some xml elements matched up to the enumerations in the target schema. I thought the easiest way would be to do a data lookup with the incoming value and just map it to the output necessary. This article is very well written and great example to follow.

http://www.ibm.com/developerworks/xml/library/x-xsltip.html

Transforming xml String in ASP/vbScript

When using the Microsoft XMLDOM in ASP there are plenty of tutorials like the ones at w3schools.com that describe loading and transforming an xml file on the server. In this case we were calling to a web service that returned xml in a string and wanted to send that string through an xslt transformation.

If you use the xmlDom.load(variableName) function like many of the examples you receive a error like:

The filename, directory name, or volume label syntax is incorrect

To correct this you need to use the xmlDom.loadXml(variableName)

Here is a sample ASP vbscript to show it working, notice the loadXml on line 29.



<%@ Language=VBScript %>

<%Option Explicit%>

<%
Dim xslt
Dim xslProc
Dim xslDoc
Dim xmlDoc
Dim sXslt
Dim sResponse
Dim sXmlString
Dim sErrMsg

sXslt = "myXslt.xsl"
sXmlString = "<?xml version=" & Chr(34) & "1.0" & Chr(34) & "?>" & vbCrLf
sXmlString = sXmlString & "<catalog>"& vbCrLf
sXmlString = sXmlString & "<cd><title>The Alter and The Door</title><artist>Casting Crowns</artist></cd>" & vbCrLf
sXmlString = sXmlString & "<cd><title>The Day Of</title><artist>David Wells</artist></cd></catalog>"

set xslDoc = Server.CreateObject("MICROSOFT.FreeThreadedXMLDOM")
xslDoc.async = false
xslDoc.load(Server.MapPath(sXslt))

if xslDoc.parseError.errorCode = 0 Then
set xmlDoc = Server.CreateObject("MICROSOFT.FreeThreadedXMLDOM")
xmlDoc.async = false
xmlDoc.loadXml(sXmlString)
if xmlDoc.parseError.errorCode = 0 Then
set xslt = Server.CreateObject("Msxml2.XSLTemplate")
xslt.stylesheet = xslDoc
set xslProc = xslt.createProcessor()
xslProc.addParameter "userid" , "1"
xslProc.addParameter "amount" , "100"
xslProc.addParameter "paymentDescriptor" , "Sample"
xslProc.addParameter "feeDescriptor" , "fee Sample"
xslProc.input = xmlDoc
xslProc.transform()
sResponse = xslProc.output
else
response.write "Internal Error " & xml2Doc.parseError.reason
end if
else
response.write "Internal Error : " & xslDoc.parseError.reason
end if

response.write "xslt = " & Server.HTMLEncode(xslDoc.xml) & "<hr>"
response.write "input xml String = " & Server.HTMLEncode(sXmlString) & "<hr>"
response.write "output xml = " & Server.HTMLEncode(sResponse)

%>


The xslt in this simplified example is:


<?xml version='1.0'?>
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:param name="userid"/>
<xsl:param name="amount" />
<xsl:param name="paymentDescriptor"/>
<xsl:param name="feeDescriptor"/>
<xsl:output method="xml" version="1.0" encoding="iso-8859-1" indent="yes" />
<xsl:template match="/">
<transaction>
<userid><xsl:value-of select="$userid"/></userid>
<amount><xsl:value-of select="$amount"/></amount>
<payment_descriptor><xsl:value-of select="$paymentDescriptor"/></payment_descriptor>
<catalog>
<xsl:for-each select="catalog/cd">
<cd>
<title><xsl:value-of select="title"/></title>
<artist><xsl:value-of select="artist"/></artist>
<fee_descriptor><xsl:value-of select="$feeDescriptor"/></fee_descriptor>
</cd>
</xsl:for-each>
</catalog>
<status_code><xsl:value-of select="transaction/status_code"/></status_code>
<status_text><xsl:value-of select="transaction/status_text"/></status_text>
</transaction>
</xsl:template>
</xsl:stylesheet>


CVSNT to Linux CVS

Converting a cvsnt repository to native linux cvs was a simple copy of the repository except for the fact that cvsnt handles binary files differently.

Found this blog entry http://firefang.net/english/converting-cvsnt-repository-format-to-cvs-repository-format which described the difference in handling binary files. It resulted in running the commandline script below in the cvsroot directory after making sure all files/folders where writable by whoever is running the command. This is a one time operation, the back up scripts can easily be removed once the repository is verified.

Careful with the quotes this is to be a single line command. I had to work with the one from the blog above a bit, this worked notice the single quotes around expand:

find . -not -name "*,v.bak" -exec grep -q "kopt.*b" {} \; -not -exec grep -q "expand. *@b@;" {} \; -print0 | xargs -0 -i sh -c "echo conv {} ; cp {} {}.bak ; cat {}.bak | sed 2a\ 'expand @b@;' > {} "

ASP Respose.Redirect to JSP jsession

Well this starts my first ever blog, I've searched and used many such articles and find that it is high time to start giving back. Hope it's helpful to someone.

Today I found my self trying to link a legacy ASP web application to a J2EE web application. The idea was simple, take a request from the ASP page, formulate an XML request to the J2EE app for authentication, receive back the url from the J2EE app where to forward the user to complete the task. Pretty straight forward and simple ASP/VBScript until it came time to redirect the user.

Searching for how to Redirect in ASP one finds many articles on Response.Redirect ("url")

However, in my case the url looked something like:
http://my.servername.com:9999;jsessionid=skskdlrickdrfjfkwesf!20090624?myid=123456

When you push this through Response.Redirect it encodes the ";","="and "!" before the querystring thus you get something like:
http://my.servername.com:9999%3Bjsessionid%3Dskskdlrickdrfjfkwesf%2120090624?myid=123456

This doesn't play well on the J2EE server and hence you don't get to where you want to be.

Solution:

Take matters into your own hands with the HTTP Headers:

Response.Status = "302 Object moved"
Response.AddHeader "Location", Url
Response.End


The hint that was very helpful to me came from http://www.somacon.com/p145.php

Hope this helps.