developer bits

Powershell Script for Setting SharePoint's Developer Dashboard

2011-02-22T18:44:00.000-08:00

Here's a script that can be used to enable, disable or allow toggle of the Developer Dashboard in SharePoint 2010. Note that if using the V3 interface, you can customize the master page to add the dashboard control so that it will render:
<SharePoint:DeveloperDashboard runat="server" />


$svc=[Microsoft.SharePoint.Administration.SPWebService]::ContentService

$ddsetting=$svc.DeveloperDashboardSettings

$choiceOnDemand = New-Object System.Management.Automation.Host.ChoiceDescription `
        "On &Demand","Dashboard can be toggled via icon near Welcome Menu"
$choiceOn = New-Object System.Management.Automation.Host.ChoiceDescription `
        "&On","Dashboard is on for all pages"
$choiceOff = New-Object System.Management.Automation.Host.ChoiceDescription `
        "O&ff","Dashboard is off"

$choices = [System.Management.Automation.Host.ChoiceDescription[]]($choiceOnDemand, $choiceOn, $choiceOff)
$caption = "Specify Developer Dashboard Setting for Farm"
$message = "Current the dashboard is: $($ddsetting.DisplayLevel)"
$result = $Host.UI.PromptForChoice($caption,$message,$choices,0)

switch ($result) {
    0 { Write-Host 'Dashboard is now On Demand, toggle via icon near Welcome Menu'
        $ddsetting.DisplayLevel=[Microsoft.SharePoint.Administration.SPDeveloperDashboardLevel]::OnDemand 
        break
      }
     
    1 { Write-Host 'Dashboard is now On'  
        $ddsetting.DisplayLevel=[Microsoft.SharePoint.Administration.SPDeveloperDashboardLevel]::On 
        break
      }
    2 { Write-Host "Dashboard is now Off"
        $ddsetting.DisplayLevel=[Microsoft.SharePoint.Administration.SPDeveloperDashboardLevel]::Off 
        break
      }
      
    default { throw "Developer Error"}
}

if ($result -ne 2) {
  $ddsetting.TraceEnabled = $true
  Write-Host @"
  Trace was also enabled.
  Note: v3 interface requires this on the master page for the Dashboard to render:
     <SharePoint:DeveloperDashboard runat="server" />
"@
} else {
    $ddsetting.TraceEnabled = $false;
    Write-Host "Trace was also disabled."
}

$ddsetting.RequiredPermissions = 'EmptyMask'

$ddsetting.Update()

It also turns on trace, so it can be handy to find hard to debug problems.
In my case, it was helpful to find out why a delegate control wasn't rendering:


Tag(8e1n) Failed to create a user control from virtual path '/_controltemplates/CustomSearch.ascx': 
'Microsoft.SharePoint.WebControls.SearchArea, Microsoft.SharePoint, Version=12.0.0.0, Culture=neutral, PublicKeyToken=71e9bce111e9429c' 
is not allowed here because it does not extend class 'System.Web.UI.UserControl'.

Based on approach from Praveen Battula's blog

SyntaxHighleghter Setup

2011-02-22T16:42:00.000-08:00

This setup will use the latest version of hosted SyntaxHighlighter:


<link href='http://alexgorbatchev.com/pub/sh/current/styles/shThemeDefault.css' rel='stylesheet' type='text/css'/>
<link href='http://alexgorbatchev.com/pub/sh/current/styles/shCore.css' rel='stylesheet' type='text/css'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shCore.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shAutoloader.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushCss.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushJScript.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushXml.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushCSharp.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushJava.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushPhp.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushPlain.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushPowerShell.js' type='text/javascript'/>
<script src='http://alexgorbatchev.com/pub/sh/current/scripts/shBrushPython.js' type='text/javascript'/>

<script language='javascript'>
SyntaxHighlighter.config.bloggerMode = true;
SyntaxHighlighter.config.clipboardSwf = 'http://alexgorbatchev.com/pub/sh/current/scripts/clipboard.swf';
SyntaxHighlighter.all();
</script>

Just put this above the closing head tag.

Then put your code in a pre tag and apply the appropriate brush, like so:

<pre class=="brush: xml">

Remember to encode you code using something like Postable or String Functions

Editing Files in a WSP

2011-02-09T06:20:00.000-08:00

Sometimes you don't have the Visual Studio project handy for a wsp--but you want to change something (settings, etc). I can remember trying this a while back and finding it tedious--having to resort to makecab and the like. The need came up again so I figured it was worth a search to see if anything had changed, and behold, there is an easier way now. Thanks to Grumpy Wookie for posting about a nice archiving utility, IZArc, and how to use it to edit wsp's. I found a slightly simpler process. The steps are basically:

Rename the WSP to CAB
Extract the wsp to a folder
Edit the extracted files as needed
Select all the files in extracted folder (Ctrl-A) and create a zip archive (this avoids the problem Grumpy Wookie ran into with IZArc not handling subfolders)
Open the zip in IZArc and do Tools > Convert Archive and select Cabinet (.cab) as the Output Type to create it as a CAB
Edit the created cab file to change it back to a wsp

You can download IZArc from CNet.

I wouldn't normally blog about something this trivial, but I figure the more blog posts on this, the better.

Debugging Gotcha With .Net's System.Diagnostics.Debugger.Break()

2010-04-21T10:51:00.000-07:00

In developing a SharePoint feature deployed at web application scope, I was trying to attach the debugger to the w3wp.exe process so I could see what was going on in my code. I used the iisapp command to try and pick the logical process... no dice. So then I connected to ALL of the w3wp process (there were 5 for them). It still wouldn't hit my breakpoint. So then I threw a System.Diagnostics.Debugger.Break(); statement into my feature's code. When the feature activated, as expected I was given the option of selecting the debugger I wanted to use. So I chose my current VS2008 session. And here's the kicker--the next thing that happens is a dialog pops up telling you that "There is no source available at the current location". Now I had already dealt with all kinds of problems with missing symbols getting to this point, and I thought this was just the next level in the rathole I had been down on that issue. I was about to give up, but then I notided that the Call Stack pane was actually showing that the process was stopped down inside the System.Diagnostics assembly (which makes sense). All I had to do was a Debug > "Step Out" and I was back in my code, ready to debug. File under "little things that kill".

Face Detection "Hello World" in IronPython using Emgu CV and OpenCV

2009-11-11T08:02:00.000-08:00

While following the instructions for setting up IronPython to use Emgu CV:
http://www.emgu.com/wiki/index.php/Setting_up_Emgu_CV_and_IronPython
I had a few problems. I started by trying to use latest versions (2.0) of these packages but kept getting dll load errors like:
"SystemError: The type initializer for 'Emgu.CV.CvInvoke' threw an exception."
I put my best google-foo on it and tried things like upgrading .Net to 3.5 but no dice. So I backtracked to Emgu CV 1.5 and OpenCV 1.1 and was able to getting it working:


import sys
import clr  

sys.path.append(r"C:\OpenCV11\bin")
sys.path.append(r"C:\apps\Emgu1.5")

clr.AddReferenceToFile("Emgu.CV.dll")
clr.AddReferenceToFile("Emgu.Util.dll")
clr.AddReferenceToFile("Emgu.CV.UI.dll")
clr.AddReference("System.Drawing")

from Emgu.CV import Image, HaarCascade
from Emgu.CV.UI import ImageViewer
from Emgu.CV.Structure import Gray
from System import Byte
from System.Drawing import Size
from Emgu.CV.CvEnum import HAAR_DETECTION_TYPE

detector = HaarCascade(r"C:\OpenCV11\data\haarcascades\haarcascade_frontalface_alt2.xml")

img = Image[Gray, Byte](r"C:\OpenCV11\samples\c\lena.jpg")

# http://www.cognotics.com/opencv/servo_2007_series/part_2/page_2.html
#http://www.emgu.com/wiki/files/2.0.0.0/html/11c784fc-7d30-a921-07ec-ecdb7d217bbe.htm
scaleIncreaseRate = 1.1
minNeighbors = 3
minSearchScale = Size(img.Width/8, img.Height/8) # As a ratio of the image size:
#Or a fixed pixel size: minSearchScale = Size(100,100)
iChannel = 0 # only 1 channel

objectsDetected = img.DetectHaarCascade(
   detector,
   scaleIncreaseRate,
   minNeighbors,
   HAAR_DETECTION_TYPE.DO_CANNY_PRUNING,
   minSearchScale
   )[iChannel]

for obj in objectsDetected:
   print "Object Rectange: %s" % obj.rect
   img.Draw(obj.rect, Gray(255), 1)

ImageViewer.Show(img)

A couple notes compared to the Emgu wiki version--I was able to get everything working without having to disturb the IronPython install (I didn't copy any files into the IronPython start directory). I simply linked used sys.path to add the locations where the dll's live. Also, I did explicit import for everything. It's a bit more work to figure out what needs to be imported, but it avoids namespace pollution issues. Also, the img.Draw line needed a few tweaks. I had to drop the [float] bit from the original. And with the original version, I was also getting an error on this same line, saying:

TypeError: expected MCvBox2D, got MCvAvgComp

I was able to work around this by passing in the ".rect" rectangle attribute to the Draw method (I'm guessing the api has evolved since the 1.4 Emgu version used for the original). Lastly, I was getting two objects detected--so I used parameters to the detector to only get the single face in the image (basically by telling it to use a larger minimum detection size, minSearchScale). Anyway, now it works the way I was hoping.

And here's some bits about the environment, and the results it prints out:


C:\temp>"C:\Program Files\IronPython 2.6\ipy"
IronPython 2.6 (2.6.10920.0) on .NET 2.0.50727.3603
Type "help", "copyright", "credits" or "license" for more information
>>> execfile(r"C:\work\EmguTest.py")
Object Rectange: {X=222,Y=206,Width=163,Height=163}

And lastly, I'd be remiss if I didn't include lovely Lena, with face detected (in B&W, sorry about that Lena).

UPDATE:
When I tried to port the above configuration to my target server, I started getting that Emgu.CV.CvInvoke exception again. I upgraded the server to .Net 3.5 but that didn't get it. Then I applied the Microsoft Visual C++ 2005 SP1 Redistributable Package (see link) to the server and then it ran without error. So to summarize, I was able to get this running using Emgu CV 1.5, OpenCv 1.1, and the Visual C++ 2005 SP1 package installation. It runs on IronPython 2.0.1 and 2.6. I'm not certain if a .Net 3.5 had any effect:

Microsoft Visual C++ 2005 SP1 Redistributable Package Download

Annoying cache problem in FF3.0.x

2009-08-17T10:25:00.000-07:00

"If it ain't broke, fix it till it is"
Of course nobody would do that intentionally, but what if something looks broken. Well then you might start throwing the kitchen sink at it and actually break it--that's my recent experience doing some SharePoint branding. I'm making changes to the CSS to overwrite the background image for the brand. It works fine in IE but it keeps showing the image I'm trying to overwrite in FF (3.0.13 in this case). So I waste a couple hours between trying variations and replicating the problem with a simple test case--having no idea it was a caching issue. But when I couldn't replicate the problem, I used a completely different computer and realized it looked fine there on the same version of FF. Argh! All that time down the drain on something that wasn't broken. Mind you, I did suspect caching at one point early on and cleared out all the cache and tried restarting FF. Now onto the research phase... After spending some quality time with google on this subject, it seems FF3 did have some caching issues. You can read about it here:

http://www.west-wind.com/Weblog/posts/469125.aspx

And here:

http://forums.mozillazine.org/viewtopic.php?f=25&t=673135&sid=7859918dca9a1722635e233ef6f401db

The bottom line--I tried everything I could think to do with FF to get it to stop caching (about:config browser.cache settings, clearing cache umpteen different ways, physically deleting cache files, appending url parameters onto the files in question) but couldn't break the back of the beast. The one thing that finally solved it was creating a new profile for FF and using that (as suggested by jscher2000 in that second link). Here's how:

http://kb.mozillazine.org/Creating_a_new_Firefox_profile_on_Windows

As for the root cause, I'm not sure. The various posting on this show it's pretty complicated--possibly involving interactions between thinks like the type of request (GET, POST), the transport (http versus https), the cache header settings, if you have other tabs open, and installed addins (especially FireBug). The nice thing about the additional profile is I can keep it plugin free--which may be why it solved the problem. Hopefully this won't be a problem anymore in FF 3.5.

Interestingly, using about:cache and FireBug, I was able to determine that the CSS file that didn't seem to be having any effect was in fact being downloaded. But when I looked at the computed CSS in FireBug, it wasn't showing the change. That doesn't necessarily mean it's FireBug causing it, but it could be the Heisenberg Uncertainty Principal is at play.

UPDATE
A coworker suggested I try Ctrl-Shift-R which is supposed to completely bypass cache. So I fired up the handy Tamper Data FireFox plugin since it shows the "Load Flags" of each requested URL (presumably FireFox's decision about what action it needs to take regarding the request URL--pull from cache, download, etc). A quick test of this on a blogger.com page showed that it did indeed eliminate most cached elements with a status of LOAD_BYPASS_CACHE. However, a few URLs, in this case images of type .png and .gif, still had a LOAD_ONLY_IF_MODIFIED status without any further followup showing up in Tamper Data. So while Ctrl-Shift-R does resolve most caching, it is still doesn't eliminate it completely.

SharePoint Feature Scope Changes Ignored

2009-04-17T10:24:00.000-07:00

The Problem

I ran into a situation where I had deployed a broadly scoped feature and then narrowed the feature scope down, but the feature was acting like it still had the original scope settings. The feature was a delegate control to deploy JQuery using the AdditionalPageHead delegate control, just like Jan Tielens described here. When I had the feature scoped to the site collection (Scope="Site"), it was auto-deploying across the site collection as expected. But then when I tightened the scope down to specific websites (Scope="Web") and added a Scope="false" to make it require manual activation, it was behaving as if the it was still scoped at the site collection and was still auto-deploying. This continued even when I completely deleted the solution.

The Fix

After completely deleting the solution, I updated the solution and feature with new GUIDs. Thereafter, it behaved as expected. It seemed to behave like some remenant of the solution wasn't fully removed by from the solution store even though the solution was deleted. Here's a related forum posting on the issue. If it was something I did wrong, then I doubt changing the feature and solution GUIDs would have fixed it--that's my story anyway.

I'll also mention that I tried developing this with VSeWSS 1.3 but was frustrated by it always creating an assembly (even for the "empty" project template) which couldn't be removed without breaking things. I fell back to STSDEV which is remains my tool of choice.

A Basic Example of JQuery's UI.Dialog Dynamically Loading an URL with an IFRAME

2009-03-27T14:20:00.000-07:00

Given the momentum that JQuery has had lately with .Net, I wanted to try out loading a form into a "lightbox" as an alternative to writing a webpart for SharePoint. So a first step is just to get the "lightbox" approach working. I used the JQuery UI Build Your Download page to build a barebones download for version 1.7.1. This download gives a nice bundle with demo pages--however I didn't see any pages with an example of dynamically loading an URL using an IFRAME. Searching the web turned up lots of stuff, but between different versions of the toolkit, options for using other plugins, and different approaches--I didn't quite find a nice simple example. So here goes.

Upon unpacking the download archive, you'll get a folder structure. I put my sample html files along side the html files in \development-bundle\demos\dialog, so if you're following along, that's where you should put the samples below.

In this first sample, the IFRAME is placed inside a DIV. In the second sample, the IFRAME is dynamically generated into the DIV--so it doesn't have to appear in the DIV content intially. In either case, there is some boilerplate css and jquery files that are loaded (they are pretty small in size compared to some other javascript libraries fortunately). Both samples also show the $(document).ready() function used to specify settings for the dialog. Then a links onclick method invokes the showDialog method. When the IFRAME is statically part of the div tag, it's src attribute is simply set in shoeDialog using JQuery's attr method.

<!DOCTYPE html><html><head>

    <link type="text/css" href="../../themes/base/ui.all.css" rel="stylesheet" />

    <script type="text/javascript" src="../../jquery-1.3.2.js"></script>
    <script type="text/javascript" src="../../ui/ui.core.js"></script>
    <script type="text/javascript" src="../../ui/ui.dialog.js"></script>

  <script type="text/javascript">
function showDialog(){
   $("#divId").dialog("open");
   $("#modalIframeId").attr("src","http://www.google.com");
   return false;
}

$(document).ready(function() {
   $("#divId").dialog({
           autoOpen: false,
           modal: true,
           height: 500,
           width: 500
       });
});
  </script>
</head>
<body >

<a href="" onclick="return showDialog()">Show the Dialog</a>

<div id="divId" title="Dialog Title">
    <iframe id="modalIframeId" width="100%" height="100%"
    marginWidth="0" marginHeight="0" frameBorder="0" scrolling="auto"
    title="Dialog Title">Your browser does not suppr</iframe>
</div>

</body>
</html>

For the second sample, just a few lines are different--removing the IFRAME from the DIV, and instead creating it in the showDialog method.

<!DOCTYPE html><html><head>

    <link type="text/css" href="../../themes/base/ui.all.css" rel="stylesheet" />

    <script type="text/javascript" src="../../jquery-1.3.2.js"></script>
    <script type="text/javascript" src="../../ui/ui.core.js"></script>
    <script type="text/javascript" src="../../ui/ui.dialog.js"></script>

  <script type="text/javascript">
function showDialog(){
   $("#divId").html('<iframe id="modalIframeId" width="100%" height="100%" 
       marginWidth="0" marginHeight="0" frameBorder="0" scrolling="auto" />').dialog("open");
   $("#modalIframeId").attr("src","http://www.google.com");
   return false;
}

$(document).ready(function() {
   $("#divId").dialog({
           autoOpen: false,
           modal: true,
           height: 500,
           width: 500
       });
});
  </script>
</head>
<body >

<a href="" onclick="return showDialog()">Show the Dialog</a>

<div id="divId" title="Dialog Title" />

</body>
</html>

In either case, you can specify attributes on the IFRAME to control standard IFRAME features, and you can also use different options for the ui.dialog, described here (the JQuery documentation is quite good). Currently, the samples only use scrolling when required.

No BackStage Pass for Nested Classes in C#

2009-03-11T08:00:00.000-07:00

In java land, I used to use nested classes to simplify a class that was getting complicated. If a class grew a cluster of fields and logic related to a single concept, I would create an inner class to encapsulate all that thus making the outer class simpler to understand--basic OO modularity/cohesion principles. Yes, some folks don't like inner classes and in java the compiler actually opens up the access modifiers of the fields--but I found it useful. But in C#, the nested class doesn't share any special access into the outer/enclosing class (perhaps for security reasons?). The workaround seems to be to pass in an instance of the outer class. The problem with this is that without special access to the internals of the outer class (fields, etc), you have to start modifying the access modifiers of the outer class, or start pushing all the state into the nested class via constructors or methods/properties. But that defeats the purpose--the idea was to keep the API to the outer class the same and keep the change encapsulated between the outer and nested class. The workaround I'm using is to have a special method on the outer class that has all the business logic needed to construct the nested class. Not perfect, but it seems like a decent compromise.

Calling an Oracle Stored Proc from ASP.Net

2009-02-20T09:26:00.000-08:00

Here is a basic example of calling an Oracle Stored Proc from ASP.Net using a ref cursor to return the result. Basically, I installed the Oracle Data Provider for .Net downloaded from Oracle (just google "oracle .net"). I used the 11g provider even though I'm on a 9 database (it works fine so far). After setting up a connection to my Oracle database, I dropped a List View onto a page and bound it to a table in the schema just to display some data. Asp.Net does it's magic behind the scenes to compile the necessary subclasses and wire up connections and what not for everything to work. But at this point, the code behind class is basically empty. It turns out you can dig out the connection property that was generated by Asp.Net. Here's a little helper method that reuses the "MyConnection" property to serve up fresh OracleConnections:

protected OClient.OracleConnection getOracleConnection() {
  return new OClient.OracleConnection(MyConnection.ConnectionString);
}

So using this, here's the code to retrieve a value from a ref cursor filled in by a stored proc:

using System;
using System.Data;
using Oracle.DataAccess.Client;
using Oracle.DataAccess.Types;
using OClient = Oracle.DataAccess.Client;                                    // (1)

public partial class Class1 : System.Web.UI.Page {

protected void Page_Load(object sender, EventArgs e) {
    using (OClient.OracleConnection oConn = getOracleConnection()) {         // (2)      
        using (OClient.OracleCommand oCmd = new OClient.OracleCommand()) {

            oCmd.Connection = oConn;
            oCmd.CommandText = "mypackage.sayhello";                         // (3)          
            oCmd.CommandType = System.Data.CommandType.StoredProcedure;      // (4)          

            // Use name binding (instead of positional).  Makes pl/sql parameter           
            // names part of the public api (can't change them without changing           
            // calling code), but moreclear and less error prone than positional.          
            oCmd.BindByName = true;                                          // (5)          

            OClient.OracleParameter inSessionToken = new OClient.OracleParameter(
                    "in_session_token", OClient.OracleDbType.Long, ParameterDirection.Input);

            OClient.OracleParameter outCursor_text = new OClient.OracleParameter(
                    "out_text", OClient.OracleDbType.RefCursor, ParameterDirection.Output);

            // Should be first parameter added (see below)
            oCmd.Parameters.Add(outCursor_text);                             // (6)
            oCmd.Parameters.Add(inSessionToken);

            inSessionToken.Value = 462; // just to show data getting passed in

            oConn.Open();

            Label1.Text = String.Format("MYPACKAGE.SAYHELLO: {0} at {1}",
                    oCmd.ExecuteScalar(), System.DateTime.Now);              // (7)
        }
    }
}
}

Annotations:

1) There are various class name collisions between the Oracle.DataAccess.Client package and the System.Data.OracleClient package, and you need to reference both. Here the "using" alias helps make the code more readable while disambiguaging to the correct classes. Otherwise Oracle.DataAccess.Client would be sprinked throughout the code. It's really just syntactic sugar, but it helps.

2) Here the other (more important) application of "using" is managing the database resources. Note the nested calls for the connection and the command. This just saves boilerplate resource management code. Note that "using" calls IDispose on the underlying object. In the case of ADO, this should either return the connection to the pool (if pooling) or close the connection if not-\-assuming the provider was correctly written.

3) This is where the package.procedure is specified. Of course, you may just be using a procedure directly without a package (just use the procedure name by itself).

4) This is required to let the driver know you are calling a stored proc

5) I'm using named parameters, otherwise you could use positional parameters like :1, :2 , etc.

6) As I explain in the comments, it's safest to just add the ref cursor parameter first for the ExecuteScalar to work properly.

7) Note, that it doesn't make a lot of sense to spin up a ref cursor just to get a single "cell" of data out of it--you'd probably just want to return a varchar2 from a function instead. However, it's a quick way to show that it works. When a single value is all that's needed, ExecuteScalar can be used.

Here's what the stored proc looks like:

PACKAGE BODY "MYPACKAGE" IS

PROCEDURE "SAYHELLO" (
in_session_token IN PLS_INTEGER,
out_text OUT sys_refcursor) IS

BEGIN

open out_text for select 'Oracle says "Hello" for session token ' ||
in_session_token from dual;

END "SAYHELLO";

END "MYPACKAGE";

Of course, you'd want a package declaration (not shown, but it basically is just declares the procedures signature).

You can also use the OracleRefCursor class in conjunction with the
OracleDataAdapter class to populate DataTables and DataSets. Also, note that
the ExecuteScalar method for a refcursor must be returned from the function or the must be the first out bind parameter.
Oracle article on using ref cursors with .Net

The ref cursor is then accessed as either an OracleDataReader object or a
DataSet, by use of the OracleDataAdapter class.
Oracle article on using the ways you can handle a ref cursor

If you only need read-only data, use the OracleDataReader to populate a DataSet. If you need offline or updateability, use OracleDataAdapater.

Canonical Python str method

2009-01-28T12:50:00.000-08:00

Here's a python class that shows some approaches for building up a meaningful __str__() to display the class's data attributes (like overriding toString() in java). The first way is to show a specific subset (or ordering) of attributes. The second way (currently commented out) just shows all the classes attributessorted by name. I also show how to make it work with a subclass--either leveraging the superclass __str__() method or doing it all from within the subclass. Of course, it should be easy to adjust the output to your own needs (padding, etc).

Nothing earth-shattering here, but it can save some work if it's a big class with lots of attributes. The example should work with modern CPython, including Python 3:

class X:
  def __init__(self):
    self.a = "apple"
    self.b = "banana"
    self.c = "cranberry"

  def __str__(self):
  
    # to show specific variables
    showList = ["a", "b"]
    
    # to show include all variables in sorted order
    #showList = sorted(set(self.__dict__))
    
    return ("X(%i):\n" % id(self)) + "\n".join(["  %s: %s" % (key.rjust(8), self.__dict__[key]) for key in showList])
    
class Y(X):

  def __init__(self):
    X.__init__(self)
    self.d = "date"
    self.e = "elderberry"
    
  def __str__(self):
    showList = ["a","d","e"]
    
    # if you want to use the superclass as-is and supplement from this subclass
    # return X.__str__(self) + "\n  Y:\n" + "\n".join(["  %s: %s" % (key.rjust(8), self.__dict__[key]) for key in showList])
    
    # if you want to do everything from this subclass
    return ("Y(%i):\n" % id(self)) + "\n".join(["  %s: %s" % (key.rjust(8), self.__dict__[key]) for key in showList])
    
print(Y()) 
 
This outputs:
 
Y(12172528):
         a: apple
         d: date
         e: elderberry

Okay, but maybe this can just be a helper method of some kind. The class name info can be pulled of of the self.__class__ object, and then the function can be more generic:

def str_impl(anInstance, displayAttributeList):

    classLine = "%s.%s(%i):\n" % (anInstance.__class__.__module__, anInstance.__class__.__name__, id(anInstance))
    return (classLine + "\n".join(["  %s: %s" % (key.rjust(8), anInstance.__dict__[key]) for key in displayAttributeList]))
    
class XX:

  showList = ["a", "b"]

  def __init__(self):
    self.a = "apple"
    self.b = "banana"
    self.c = "cranberry"
    
  def __str__(self):
    return str_impl(self, self.__class__.showList)
    

class YY(XX):
    showList = ["a","d","e"]
    
    def __init__(self):
        X.__init__(self)
        self.d = "date"
        self.e = "elderberry"   

This outputs:

__main__.YY(12172784):
         a: apple
         d: date
         e: elderberry

Note that __main__ is the module since the classes are defined in the "main" scope. The correct module/package would print if I had used one. Again, you wouldn't even need to bother with this for a small class--but for a class with a lot of attributes (which might be a sign it's time to break it up anyway ;-), it saves some tedium.

Connecting to an Oracle Database from ASP.Net/SharePoint

2009-01-09T08:17:00.000-08:00

So you’re writing your “Hello Oracle Database” from ASP.Net/SharePoint, and you get this error: “System.Data.OracleClient Requires Oracle Client version 8.1.7 or greater”.

Perhaps you don’t really want to install the 475 Mb Oracle Client on your system just to talk to a database. Some quick googling suggests you can just install the Oracle Instant Client, and if you don’t care about internationalization, you can even go with the Instant Client “Basic Lite” which gets you down to about 18 Mb:

http://www.oracle.com/technology/software/tech/oci/instantclient/htdocs/winsoft.html [32 bit version]

Yeah, it’s an order of magnitude larger than the jdbc jar you’d use if you were connecting from Java, but at least it’s not a half Gig. But what if you don’t even want to install it? This blog entry shows the 4 files you’ll need to from the instant client:

http://www.thewayofcoding.com/2008/02/c-net-programming-tip-connecting-to-an-oracle-database/

In my case, I’m actually developing a SharePoint web part. Therefore, I’d like to make the set of driver files universally available for anything deployed on the system. So I drop those four files:
· oci.dll
· orannzsbb10.dll
· oraocci10.dll
· oraociicus10.dll

into the system driver directory (in this case C:\Windows\System32). And now I’m good to go.
Note: I’m not sure if it’s related, but I did not have to change any permissions as seems to happen when you actually install the client (at least in some cases)—as described here:

http://moustafa-arafa.blogspot.com/2007/04/systemdataoracleclient-requires-oracle.html

I suppose it’s time I update my About Me to say I do .Net now too :-)

IE6 problem with Dojo: Could not load 'dojo.parser'

2008-03-17T14:02:00.000-07:00

I encountered this problem when testing with IE 6:

Error: Could not load 'dojo.parser'; last tried './parser.js'

After researching it a bit, it usually occurs as the result of a script that loads before the dojo stuff and which does a document.write. Apparently this confuses dojo. The solution is to use the baseUrl property of djConfig to manually specify the base path (the path to the dojo.js file):

      <script type="text/javascript" src="../dojo_tools/dojo/dojo.js"
              djConfig="parseOnLoad: true, baseUrl: '/dojo_tools/dojo/'"></script>

Note the baseUrl should have a trailing '/'.

When dojo loads, if no baseUrl is provided, dojo will try to dynamically determine the base url. However, other scripts can interfere with this (at least on IE), so manually specifying it fixes it.

Dojo xhr to update a div

2008-02-28T14:35:00.000-08:00

Using ajax to update a portlet quickly becomes a requirement when you get a bunch of portlets on a page and reloading everything is slow. So in setting out to use dojo to do this, I figured it required nothing more than setting the innerHTML of my div with the new content. It's not that simple, however, if the panel contains dojo's "dijit" widgets (in this case, I had a dijit.layout.TabContainer returned with the ajax content). BTW, dojo folks, the TabContainer worked flawlessly right out of the box and was clean and simple--nice job. In comparing the before/after, it was apparent that when the div content was rendered when the entire page loaded, dojo/dijit was wiring up the tab elements with all the required events and attributes. However, when the ajax content was set as innerHTML, this wasn't happening. It was a bit tough to figure out mostly because I wasn't sure what to google on to find the solution. I tried a few things, including making the div for the content a dijit.layout.ContentPane--but that's just part of what's needed. Here's what the content pane div looks like:

<div id="ajaxReportBody" dojoType="dijit.layout.ContentPane">
    <c:import url="reportBody.jsp" />
</div>

The other piece is to make sure you use setContent() in your xhr load function on the dijit widget (in this case, the content pane div), instead of innerHMTL:

dojo.xhrPost( { 
         ...
         load: function(response, ioArgs) {
            ...
            dijit.byId("ajaxReportBody").setContent(response);  // KEY!
            return response; 
         },
         ...

When you use setContent(), dijit will redo the plumbing in the new content for you. Also note that you must use digit.byId(), not dojo.byId().

Some references indicated that you need an executeScripts="true" on the content pane div. However, this is no longer true AFAICT (I'm using dojo 1.0.2).

BTW, best reference I found on the subject was here.

Code formatting in Blogger

2008-01-08T09:59:00.000-08:00

is a real PITA. Here's an web page utility (uses javascript) that makes it easier:
Greg Houston's Formatter for blog code
Have a better solution? Post a comment.

Hacking an WSRP portlet request using Groovy

2008-01-08T09:03:00.000-08:00

I wanted to invoke a portlet without having to run it through a portal. I have a portlet currently deployed in Oracle Portal and figured it would be "fun" to try invoking it using Groovy--I didn't see any slam-dunk toolkits for this (I didn't look too hard). WSRP uses Soap, so the first step was firing up Wireshark to sniff a real request (sanitized to protect the innocent):

POST /MyPortlet/portlets/WSRPBaseService HTTP/1.1
Host: mycomputer:8080
Connection: keep-alive, TE
TE: trailers, deflate, gzip, compress
User-Agent: RPT-HTTPClient/0.3-3
Oracle-ECID: 86139989405,1
Cookie: JSESSIONID=80f4383422b8fbea0c8532804298a55ba92d09dfb9a6
SOAPAction: "urn:oasis:names:tc:wsrp:v1:getMarkup"
Accept-Encoding: gzip, x-gzip, compress, x-compress
Content-type: text/xml; charset="utf-8"
Content-length: 2165

<?xml version="1.0" encoding="UTF-8"?>
<env:Envelope xmlns:env="http://schemas.xmlsoap.org/soap/envelope/" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:ns0="urn:oasis:names:tc:wsrp:v1:types"><env:Body><ns0:getMarkup><ns0:registrationContext><ns0:registrationHandle>C:xxx.xx.xx.xx:-5fd1fa94:11740915845:-7ffe</ns0:registrationHandle></ns0:registrationContext><ns0:portletContext><ns0:portletHandle>E:0:xxx.xx.xx.xx:-5fd1fa94:11740915845:-7ffd</ns0:portletHandle></ns0:portletContext><ns0:runtimeContext><ns0:userAuthentication>wsrp:xxxxxxxx</ns0:userAuthentication><ns0:portletInstanceKey>1179_354032_1179_1_1</ns0:portletInstanceKey><ns0:namespacePrefix>__ns1179_354032_1179_1_1_</ns0:namespacePrefix></ns0:runtimeContext><ns0:userContext><ns0:userContextKey>mylogin</ns0:userContextKey></ns0:userContext><ns0:markupParams><ns0:secureClientCommunication>false</ns0:secureClientCommunication><ns0:locales>en-us</ns0:locales><ns0:mimeTypes>text/html</ns0:mimeTypes><ns0:mode>wsrp:view</ns0:mode><ns0:windowState>wsrp:normal</ns0:windowState><ns0:clientData><ns0:userAgent>Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 1.1.4322; .NET CLR 2.0.50727) RPT-HTTPClient/0.3-3</ns0:userAgent><ns0:extensions><oracle:GenericExtension xmlns:oracle="http://xmlns.oracle.com/portal/wsrp/v1"><ns0:NamedString name="X-Oracle-Device.MaxDocSize"><ns0:value>0</ns0:value></ns0:NamedString><ns0:NamedString name="X-Oracle-Device.Class"><ns0:value>pcbrowser</ns0:value></ns0:NamedString><ns0:NamedString name="X-Oracle-Device.Secure"><ns0:value>false</ns0:value></ns0:NamedString><ns0:NamedString name="X-Oracle-Device.Orientation"><ns0:value>landscape</ns0:value></ns0:NamedString></oracle:GenericExtension></ns0:extensions></ns0:clientData><ns0:markupCharacterSets>utf-8</ns0:markupCharacterSets><ns0:validNewModes>wsrp:view</ns0:validNewModes><ns0:validNewWindowStates>wsrp:normal</ns0:validNewWindowStates><ns0:validNewWindowStates>wsrp:maximized</ns0:validNewWindowStates><ns0:validNewWindowStates>wsrp:minimized</ns0:validNewWindowStates></ns0:markupParams></ns0:getMarkup></env:Body></env:Envelope>

It would probably take a bit of trial and error to hack that for another portal/portlet, but who knows.

Putting this in groovy (stripping out the JSESSIONID):

def headerLines = """\
Host: mycomputer:8080
Connection: keep-alive, TE
TE: trailers, deflate, gzip, compress
User-Agent: RPT-HTTPClient/0.3-3
Oracle-ECID: 86139989405,1
SOAPAction: "urn:oasis:names:tc:wsrp:v1:getMarkup"
Accept-Encoding: gzip, x-gzip, compress, x-compress
Content-type: text/xml; charset="utf-8"
Content-length: 2165
"""

def postBody = """\
<?xml version="1.0" encoding="UTF-8"?>
<env:Envelope xmlns:env="http://schemas.xmlsoap.org/soap/envelope/" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:ns0="urn:oasis:names:tc:wsrp:v1:types"><env:Body><ns0:getMarkup><ns0:registrationContext><ns0:registrationHandle>C:xxx.xx.xx.xx:-5fd1fa94:11740915845:-7ffe</ns0:registrationHandle></ns0:registrationContext><ns0:portletContext><ns0:portletHandle>E:0:xxx.xx.xx.xx:-5fd1fa94:11740915845:-7ffd</ns0:portletHandle></ns0:portletContext><ns0:runtimeContext><ns0:userAuthentication>wsrp:xxxxxxxx</ns0:userAuthentication><ns0:portletInstanceKey>1179_354032_1179_1_1</ns0:portletInstanceKey><ns0:namespacePrefix>__ns1179_354032_1179_1_1_</ns0:namespacePrefix></ns0:runtimeContext><ns0:userContext><ns0:userContextKey>mylogin</ns0:userContextKey></ns0:userContext><ns0:markupParams><ns0:secureClientCommunication>false</ns0:secureClientCommunication><ns0:locales>en-us</ns0:locales><ns0:mimeTypes>text/html</ns0:mimeTypes><ns0:mode>wsrp:view</ns0:mode><ns0:windowState>wsrp:normal</ns0:windowState><ns0:clientData><ns0:userAgent>Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 1.1.4322; .NET CLR 2.0.50727) RPT-HTTPClient/0.3-3</ns0:userAgent><ns0:extensions><oracle:GenericExtension xmlns:oracle="http://xmlns.oracle.com/portal/wsrp/v1"><ns0:NamedString name="X-Oracle-Device.MaxDocSize"><ns0:value>0</ns0:value></ns0:NamedString><ns0:NamedString name="X-Oracle-Device.Class"><ns0:value>pcbrowser</ns0:value></ns0:NamedString><ns0:NamedString name="X-Oracle-Device.Secure"><ns0:value>false</ns0:value></ns0:NamedString><ns0:NamedString name="X-Oracle-Device.Orientation"><ns0:value>landscape</ns0:value></ns0:NamedString></oracle:GenericExtension></ns0:extensions></ns0:clientData><ns0:markupCharacterSets>utf-8</ns0:markupCharacterSets><ns0:validNewModes>wsrp:view</ns0:validNewModes><ns0:validNewWindowStates>wsrp:normal</ns0:validNewWindowStates><ns0:validNewWindowStates>wsrp:maximized</ns0:validNewWindowStates><ns0:validNewWindowStates>wsrp:minimized</ns0:validNewWindowStates></ns0:markupParams></ns0:getMarkup></env:Body></env:Envelope>"""

def url = new URL("http://mycomputer:8080/MyPortlet/portlets/WSRPBaseService")
def conn = url.openConnection()
conn.doOutput = true

headerArray = headerLines.split("\n")
headerArray.each {
   iColon = it.indexOf(":")
   name = it[0..<iColon]
   value = it[iColon+1..-1].trim()
   println """Header $name = $value"""
   conn.setRequestProperty(name, value)
}

conn.outputStream << postBody
conn.inputStream.eachLine { println it }

Running this produces the following output:


&lt;p class="portlet-font"&gt;Welcome, this is the view mode.&lt;/p&gt;
Current Date/Time is: 2008-01-08 12:20:36.0</ns0:markupString><ns0:locale>en</ns0:locale><ns0:requiresUrlRewriting>false</ns0:requiresUrlRewriting><ns0:cacheControl><ns0:expires>0</ns0:expires><ns0:userScope>wsrp:perUser</ns0:userScope></ns0:cacheControl><ns0:preferredTitle>ComboPortlet</ns0:preferredTitle></ns0:markupContext></ns0:getMarkupResponse></env:Body></env:Envelope>

If you have suggestions for a good library for this, or improvements to the groovy code, please comment.

Opening JBoss AS HTTP Traffic Up To More Than Localhost

2007-09-21T08:45:00.000-07:00

JBoss by default binds only to 127.0.0.1 So to open it up, use:

run -b <your machine's IP address>

To really do it right: Secure JBoss wiki page

Blather: I went to show off my latest Seam, Ajax4jsf, Richfaces suggestionBox/autocomplete widget to a coworker on JBoss AS 4.2 and had problem when I tried run against anything but localhost. Even on my own machine, but using my full hostname, the browser behaved strangely, not even giving a decent status code or anything. I used the Firefox Tamper plugin (Tamper + Firebug + Web Tools = web developer nirvana [okay, I stole this tagline from somewhere else in my web travels]) to see that the request status was "pending". Perhaps that is how JBoss goes stealth against port sniffing. Anyway, I figured it was just a security setting based on the principal of requiring developers to have to conscientiously open up security. But after reading the fine manual, I still didn't find it. Turns out this is a very popular question in the forums though. And, it was actually in the readme.html (sadly I missed it there) under "Configuration Issues" for 4.2.0 GA (however, I stopped reading at 4.2.1 GA--seems it should be under 4.2.1 since it still affects it).

For those googling:
Unable to connect connection
Internet Explorer cannot display the webpage

Turning on Hibernate Logging using Seam on JBoss 4.2

2007-09-13T09:08:00.000-07:00

At some point, you're going to need to see the sql and the bind variables that hibernate is using. When running Seam on JBoss AS 4.2 and above, there are a couple things you need to do. First, add the following lines to <jboss home>/server/default/conf/jboss-log4j.xml

<category name="org.hibernate">  
      <priority value="TRACE"/>   
      <appender-ref ref="HIBERNATE"/>  
</category>

This tells log4j to start logging hibernate at a very high detail.
Next, make sure persistence-dev.xml has the hibernate.show_sql property set to true.

<persistence-unit name="AutoPropPass5">
      <provider>org.hibernate.ejb.HibernatePersistence</provider>
      <jta-data-source>java:/AutoPropPass5Datasource</jta-data-source>
      <properties>
         <property name="hibernate.hbm2ddl.auto" value="validate"/>
         <property name="hibernate.cache.use_query_cache" value="true"/>
         <property name="hibernate.show_sql" value="true"/>   <============ HERE
         <property name="jboss.entity.manager.factory.jndi.name" value="java:/AutoPropPass5EntityManagerFactory"/>
      </properties>
   </persistence-unit>

Then, add commons-logging.jar to the project's war lib directory and tweak build.xml to include this jar in the deployment:

<copy todir="${war.dir}/WEB-INF/lib">
                        <fileset dir="${lib.dir}">
                                <include name="ajax4jsf*.jar" />
                                <include name="richfaces*.jar" />
                                <include name="oscache*.jar" />
                                <include name="commons-digester-*.jar" />
                                <include name="commons-beanutils-*.jar" />
           HERE ============>   <include name="commons-logging*.jar" />
                                <include name="jsf-facelets.jar" />
                                <include name="jboss-seam-*.jar" />
                                <exclude name="jboss-seam-gen.jar" />
                        </fileset>
                </copy>

When this is all in place, you should be able to see output like the following in <jboss home>/server/default/log/server.log:

2007-09-13 09:34:55,854 TRACE [org.hibernate.type.StringType] binding 'updi' to parameter: 1
2007-09-13 09:34:56,823 DEBUG [org.hibernate.jdbc.AbstractBatcher] about to close PreparedStatement (open PreparedStatements: 1, globally: 1)
2007-09-13 09:34:56,823 TRACE [org.hibernate.jdbc.AbstractBatcher] closing statement
2007-09-13 09:34:56,823 DEBUG [org.hibernate.jdbc.ConnectionManager] aggressively releasing JDBC connection
2007-09-13 09:34:56,823 DEBUG [org.hibernate.jdbc.ConnectionManager] releasing JDBC connection [ (open PreparedStatements: 0, globally: 0) (open ResultSets: 0, globally: 0)]
2007-09-13 09:34:56,823 DEBUG [org.hibernate.util.JDBCExceptionReporter] could not execute query [select pass0_.PASS_ID as PASS1_35_, pass0_.ADDRESS as ADDRESS35_, pass0_.state as state35_, pass0_.VERSION as VERSION35_, pass0_.reason as reason35_, pass0_.status as status35_, pass0_.DESTINATION as DESTINAT7_35_, pass0_.INS_UPD_ID as INS8_35_, pass0_.INS_UPD_TS as INS9_35_, pass0_.PASS_HOLDER_PID as PASS17_35_, pass0_.PASS_CREATOR_PID as PASS18_35_, pass0_.HOLDER_GROUP_ID as HOLDER10_35_, pass0_.HOLDER_GROUP_SUP_PID as HOLDER11_35_, pass0_.CHECKOUT_DATE as CHECKOUT12_35_, pass0_.REMOVE_DATE as REMOVE13_35_, pass0_.DUE_DATE as DUE14_35_, pass0_.ADDRESS2 as ADDRESS15_35_, pass0_.CITY as CITY35_ from ERMPMGR.PP_PASS pass0_, ERMPMGR.PP_PERSON_MV person1_ where pass0_.PASS_HOLDER_PID=person1_.PERSON_ID and (upper(person1_.FULL_NAME) like upper(?+'%'))]
java.sql.SQLException: ORA-01722: invalid number

 at oracle.jdbc.driver.DatabaseError.throwSqlException(DatabaseError.java:124)
 at oracle.jdbc.driver.T4CTTIoer.processError(T4CTTIoer.java:304)
 at oracle.jdbc.driver.T4CTTIoer.processError(T4CTTIoer.java:271)
 at oracle.jdbc.driver.T4C8Oall.receive(T4C8Oall.java:625)
         ...

Notes:

Older versions of JBoss used a different file than jboss-log4j.xml (perhaps log4j.xml) and perhaps put a log4j.properties in the META-INF dir.

TRACE is used instead of DEBUG which worked in prior versions.

I'm not positive the {hibernate.show_sql bit is required, but seam-gen puts it in there by default.

In case your wondering why I was getting the "invalid number" error, it's because Oracle uses the || symbol for string concatenation, not +. When I fixed that, it started working.

Eclipse RCP missing the Business Desktop opportunity?

2007-05-19T13:51:00.000-07:00

I was recently involved (tangentially) with a biz app proof-of-concept effort that used an eclipse RCP front-end for a J2EE middle tier. The prototype was quite successful. However, when considering adopting the RCP approach, there’s concern that if we just start knocking off a bunch of apps as plug-ins, we’ll run into some problems:

duplication (code, functionality)
managing relationships/dependencies between plug-in
biz users complaining about the update mechanism

I’m willing to bet there are others I’m not thinking of. Not to suggest that these are problems with RCP or Eclipse--I’m sure a team of seasoned plug-in developers could produce and maintain an excellent suite of custom RCP biz apps. But after doing a few such plug-ins, they’d probably realize that they could refactor out a foundational plug-in that provides common support services that were duplicated among the various app plug-in. In fact, this concept already has a name… the “Business Desktop”. I heard this term from a valtech-tv.com presentation titled "Enterprise applications with Rich iDesktop applications on Eclipse" (see link below). Essentially, a Business Desktop would be an RCP-based business application environment. The Business Desktop could provide services including:

provisioning
search
authentication/authorization
preferences
Expression Language (assuming contexts are set up)
scripting
logging
error handling
service locator

The overarching theme is something like “app plug-ins do biz logic, the business desktop does everything else”. Such an approach would greatly reduce the learning curve for RCP adoption for business apps.

There are already some pieces of this available too. The Maya project is working the provisioning piece. And I’m sure others are out there (or are available via the Eclipse api but could be wrappered for convenient plug-in use). But I don’t see anything like the Business Desktop on the horizon. Is it there but I am missing it? Is it not needed? Comments welcome.

References:

Enterprise applications with Rich iDesktop applications on Eclipse:
http://www.valtech-tv.com/
I can’t give a direct link to the "" since it is flash based. Registration required unfortunately--but there are some good technical screencasts there:

Maya project:
http://www.eclipse.org/proposals/maya/

Using seam-gen.reveng.xml to control tables used for entity generation

2007-05-04T14:12:00.000-07:00

I was having problems using seam-gen to try out reverse engineering on an Oracle Express database. Seam-gen was finding all sorts of objects somehow exposed to the user, rather than the DEMO_ tables I was trying to generate off of. Turns out that a feature (undocumented?) was added in 1.2.1.GA that let's you use a reveng.xml file to control this. After you run:

C:\temp\jboss-seam-1.2.1.GA>seam new-project

You can go to the \resources\seam-gen.reveng.xml

And add a line such as this:

<table-filter match-schema="MYSCHEMA" match-name="DEMO.*" exclude="false"/>

This should effectively restrict the tables as desired. Note you can use multiple of these entries to "carve out" the set of tables you want (although I'm not sure what the exact semantics are of using various combinations of exclude=true/false for different entries).

Some things to watch out for:

Case sensitivity
At least for Oracle, you need to use upper case (I suppose you could attemp some kind of case insensitive regex).
Owning schema in seam-gen.reveng.xml
Seems like you have to specify the owning schema for the tables, even if you are connecting with a user that has privileges for the tables.
Owning schema in build.properties
You need to specify connection properties to the owning schema, at least temporarily, or you may get WARNING: Exception while trying to get indexinfo on...

I haven't tried enough variations to prove all of the above bullets, but if you're experiencing problems, try out some of these suggestions.

Reference for table-filter

junit ant task problems in eclipse

2007-03-07T12:01:00.000-08:00

So your junit ant task won't run, failing with classpath problems. How hard can that be to figure out? It sure stumped me for a while.

You figure you need ant-junit.jar on the classpath. Hmmm, is that the project classpath (a lib dir maybe)? Or the ant home dir? But wait, it's already there... so what's the problem.

The problem is that ant-junit.jar is not self-contained, as I naively assumed. You need to supply the junit.jar file to ant, which is referenced by ant-junit.jar. Make sense when you think about it--but in the haze of red error messages whizzing by, it's easy to miss. Anyway, it's easy to fix, as documented in this blog post.

Thanks Ryan!

Remote debugging in JBoss

2007-02-14T12:47:00.000-08:00

I recently had the need to fire up remote debugging in JBoss. I'm using Eclipse, but I'm sure the directions are fairly similar for any remote debug-capable IDE. Pretty much you can follow the directions here, with one change (for windows) and one suggestion:

Change: The directions suggest that for windows, you need to use the shared memory transport flag: transport=dt_shmem. However, that didn't work for me. But the original socket flat, recommended for nix, did work.

Suggestion: Rather than copy the run file and put in these parameters, it's pretty easy to set them in the environment and then just call run (and avoids copied code). Here's mine:

set JAVA_OPTS=-Xdebug -Xrunjdwp:server=y,transport=dt_socket,address=4142,suspend=n
call run.bat

Note that you don't need double quotes around the env var, despite any spaces. However, you cannot have any spaces between JAVA_OPTS, the =, and the var definition.

Happy debugging!

Hibernate alternatives for mappedBy to a superclass property

2007-01-26T11:30:00.000-08:00

I've been working on a JPA/Hibernate prototype of an application that was previously mapped using Toplink. So this is the "meet in the middle" where there is an existing domain model that must be mapped to an existing schema. In reality, the domain model is somewhat free to change as long as the public interface stays the same. And if push comes to shove, schema changes are possible (but undesirable).

While trying to map a relationship to a superclass, I assumed that O/R relationships can be inherited in a manner analogous to OO. So I naively assumed I could:

make the superclass an @Entity
use single table inheritance
use a @ManyToOne in the superclass (and then do a @OneToMany with a mappedBy= in the other side of the relationship)
set up a discriminator column on the superclass
provide discriminator values in the subclasses

and viola... other classes could then hold references to the sublclasses.

So the key assumption here is that mappedBy could simply reference the property of the subclass even though the property is actually in the superclass (as you can do in an OO sense). But I ran into problems trying this. Such as: javax.persistence.PersistenceException: org.hibernate.AnnotationException: mappedBy reference an unknown target entity property:. And no, it's not simply a field vs. accessor visibility issue. You can try different variations (including the sin of making it a public reference) and it will have no effect.

So this explores some of the options that were tried, and what the tradeoff's are...

In the examples that follow (the naive and wrong way):

ExternalContactAssignment is the subclass of ContactAssignment (via single table inheritence)
APLEntity is the class that has the @OneToMany to a subclass (ExternalContactAssignment)
ContactAssignment is the superclass that has the @ManyToOne back-reference to APLEntity

The code looked like this:

@Entity
@Inheritance(strategy=InheritanceType.SINGLE_TABLE)
@DiscriminatorColumn(name="internal_ind", discriminatorType = DiscriminatorType.STRING)
@Table(name="CONTACT_ASSIGNMENT")
public abstract class ContactAssignment {

    // bidirectional
    @ManyToOne
    @JoinColumn(name="APL_ENTITY_SEQ_NUM", nullable=false)
    private APLEntity aplEntity;
    ...

}

@Entity
@DiscriminatorValue("N")
public class ExternalContactAssignment extends
        ContactAssignment {
      
    // nothing relevant in this class
    ...
}


@Entity
@org.hibernate.annotations.Entity
@Inheritance(strategy=InheritanceType.JOINED)
@Table(name="APL_ENTITY")
public abstract class APLEntity
    implements ExternalContactAssignable {

    @OneToMany(mappedBy="aplEntity")
    @org.hibernate.annotations.Cascade({org.hibernate.annotations.CascadeType.ALL,
            org.hibernate.annotations.CascadeType.DELETE_ORPHAN})                    
    @org.hibernate.annotations.Where(clause="internal_ind='N'")        
    private List externalContactAssignments;
    ...

}

Note: I'm stripping the code down to the bare minimum here (the real classes have lots of other stuff not relevant to the problem).

According to Emmanuel Bernard (Hibernate developer), it is semantically incorrect to assume this mapping structure to work as expected.

There are three approaches, which vary in

How much schema change you'll live with
- Can you add new tables?
- Can you add new columns?
Tradeoffs between relational integrity and OO-to-relational consistency
If you can live with superclass not being an @Entity

Here are the three approaches...

@MappedSuperclass with Column per Subclass

Approach:

ContactAssignment is mapped with an @MappedSuperclass instead of an @Entity.
ContactAssignment (superclass) maintains the references to APLEntity
Table per subclass model (new table(s) required). Discriminators cannot be used (they'll be ignored).

By using @MappedSuperclass, you lose the ability to have a relationship to the superclass. You also lose polymorphic queries when using straight JPA--although Hibernate queries will still be polymorphic.

I didn't pursue this as I'm trying to minimize schema change (an evaluation criteria for Hibernate). However, the mapping would look something like this:

@MappedSuperclass
public abstract class ContactAssignment {
 ...
   @ManyToOne
   @JoinColumn(name = "APL_ENTITY_SEQ_NUM", nullable = false)
   public APLEntity getAplEntity() {
      return aplEntity;
   }

   public void setAplEntity(APLEntity aplEntity) {
      this.aplEntity = aplEntity;
   }

}


@Entity
@Table(name="EXTERNAL_CONTACT_ASSIGNMENT")  // class-specific
public class ExternalContactAssignment extends ContactAssignment {

   // not much needed

}


@Entity
@org.hibernate.annotations.Entity
@Inheritance(strategy = InheritanceType.JOINED)
@Table(name = "APL_ENTITY")
public abstract class APLEntity {

   @OneToMany(mappedBy="aplEntity")
   @org.hibernate.annotations.Cascade({org.hibernate.annotations.CascadeType.ALL,
      org.hibernate.annotations.CascadeType.DELETE_ORPHAN})                
   @org.hibernate.annotations.Where(clause="internal_ind='N'")        
   private List externalContactAssignments;

   ...

}

Relationship Column per Subclass with Discriminator Column

Approach:

Single table model
ExternalContactAssignment (subclass) maintains the references to !APLEntity
Distinct relationship (FK) column used for each subclass
Still requires a discriminator column
Duplicative @Where and @DiscriminatorColumn's (bit of a wart).

This is Emmanuel Bernard's recommended approach, as being the most consistent between the object model and the relational model.

Consequences are that pure JPA queries are no longer polymorphic (but Hibernate queries still should be). You can no longer have a relationship to the superclass ContactAssignment. It feels less "OO" since you are forced to push the relationship down to the subclass(es)... I want OO considerations to drive this, not O/R mapping considerations (transparency!). On the relational side, things get ugly. Each subclass requires its own FK column out to the APL_ENTITY table. Although this is probably why Emmanuel says it's the most consistent, I don't think it's worth the price. XOR columns like that don't play well with referential integrity. A given row should only have one value populated no matter how many subclasses you have. It make it harder to query and index, and conceptually make the design harder to understand--and it only gets worse as you add more subclasses to the mix. It also doesn't make sense to me to have a discriminator column and still require multiple FK cols. Here it is:

@Entity
@Inheritance(strategy=InheritanceType.SINGLE_TABLE)
@DiscriminatorColumn(name="internal_ind", discriminatorType = DiscriminatorType.STRING)
@Table(name="CAU_CONTACT_ASSIGNMENT") // Note: referencing a new modified table
public abstract class ContactAssignment {

   public abstract void setAplEntity(APLEntity aplEntity);

   public abstract APLEntity getAplEntity();
}


@Entity
@DiscriminatorValue("N")
public class ExternalContactAssignment extends
        ContactAssignment {

    // bidirectional
    @ManyToOne
    @JoinColumn(name="EXT_APL_ENTITY_SEQ_NUM")
    private APLEntity aplEntity;

    @Override
    public APLEntity getAplEntity() {
        return aplEntity;
    }

    @Override
    public void setAplEntity(APLEntity aplEntity) {
        this.aplEntity = aplEntity;
    }
}


@Entity
@Inheritance(strategy=InheritanceType.JOINED)
@Table(name="APL_ENTITY")
public abstract class APLEntity
    implements ExternalContactAssignable {


    @OneToMany(mappedBy="aplEntity", cascade={CascadeType.ALL})
    @org.hibernate.annotations.Cascade(value=org.hibernate.annotations.CascadeType.DELETE_ORPHAN)
    @org.hibernate.annotations.Where(clause="internal_ind='N'")        
    private List externalContactAssignments;

}

Unidirectional Read Only Back-Reference

If I understand this correctly, this is basically ignoring Hibernate's ability to manage a bidirection relationship, and mapping a unidirectional read-only back reference from ContactAssignment to APLEntity. I believe this is what is discussed in Section 6.4.3 of Java Persistence With Hibernate. Making the back reference read-only tells Hibernate not to do a duplicative update when a ContactAssignment changes an APLEntity reference. Approach:

ContactAssignment is mapped with an @Entity.
ContactAssignment (superclass) maintains the references to !APLEntity
Single table model
Still requires a discriminator column
Duplicative @Where and @DiscriminatorColumn's (bit of a wart).

Emmanuel describes this approach as making the data design weaker. I'm not sure exactly how that is (or maybe what it means), or what tradeoff's are implied, but it's certainly is closest to what I was looking for:

no schema change required --> so no loss of relational integrity possible
scales easily with additional subclasses
let's OO considerations drive domain model design

Here's what it looks like

@Entity
@Inheritance(strategy=InheritanceType.SINGLE_TABLE)
@DiscriminatorColumn(name="internal_ind", discriminatorType = DiscriminatorType.STRING)
@Table(name="CONTACT_ASSIGNMENT")
public abstract class ContactAssignment {

    @ManyToOne
    @JoinColumn(name="APL_ENTITY_SEQ_NUM", nullable=false)
    private APLEntity aplEntity;

    ...
}

@Entity
@DiscriminatorValue("N")
public class ExternalContactAssignment extends
        ContactAssignment {

   // not much needed
}


@Entity
@org.hibernate.annotations.Entity
@Inheritance(strategy=InheritanceType.JOINED)
@Table(name="APL_ENTITY")
public abstract class APLEntity
    implements ExternalContactAssignable, Authorizable {


    @OneToMany(cascade={CascadeType.ALL})
    @JoinColumn(name="APL_ENTITY_SEQ_NUM", insertable=false, updatable=false)
    @org.hibernate.annotations.Cascade(value=org.hibernate.annotations.CascadeType.DELETE_ORPHAN)         
    @org.hibernate.annotations.Where(clause="internal_ind='N'")        
    private List externalContactAssignments;

    ...
}

Anyway, so I'll be moving ahead with this last approach as it is closest to what I want.

OC4J Rmi Client: NoClassDefFoundError: javax/ejb/EJBHome

2007-01-10T10:04:00.000-08:00

Trying to run an OC4J rmi client where the oc4jclient.jar was packaged with the project (no longer in the jdeveloper home directory) produced some vexing problems with the following error:

java.lang.NoClassDefFoundError: javax/ejb/EJBHome
at java.lang.Class.getDeclaredMethods0(Native Method)
...

despite having the ejb.jar file on the classpath. After much frustration, it turns out that oc4jclient.jar uses a Class-Path attribute in its manifest:

Class-Path: lib/ejb.jar lib/mail.jar lib/oc4j_orb.jar lib/orbbase.jar
lib/iiop_support.jar lib/jms.jar lib/jta.jar ../../lib/xmlparserv2.ja
r ../../opmn/lib/optic.jar ../../oracle/jlib/oraclepki.jar ../../jlib
/oraclepki.jar ../../oracle/jlib/ojpse.jar ../../jlib/ojpse.jar

The problem seems to be that although the application classloader does have ejb.jar on the classpath, the oc4jclient.jar gets loaded with a different classloader that can't see it, and expects the ejb.jar to be in a lib dir relative to itself. Turns out you can ignore all the other jar references, only ejb.jar is required (luckily). So, where ever you place the oc4jclient.jar, just create a lib dir in the same directory and drop the ejb.jar into it--then the problem should be solved.

Another bit of weirdness was that it wouldn't seem to work with jdk 1.5.0_06. We looked in the ext dir for anything that might be messing it up but didn't find anything obvious. It seems too bizarre to the the jdk version was also causing a problem, but simply switching to 1.5.0_02 would fix other classpath problems.

developer bits

Powershell Script for Setting SharePoint's Developer Dashboard

SyntaxHighleghter Setup

Editing Files in a WSP

Debugging Gotcha With .Net's System.Diagnostics.Debugger.Break()

Face Detection "Hello World" in IronPython using Emgu CV and OpenCV

Annoying cache problem in FF3.0.x

SharePoint Feature Scope Changes Ignored

The Problem

The Fix

A Basic Example of JQuery's UI.Dialog Dynamically Loading an URL with an IFRAME

No BackStage Pass for Nested Classes in C#

Calling an Oracle Stored Proc from ASP.Net

Canonical Python __str__ method

Connecting to an Oracle Database from ASP.Net/SharePoint

IE6 problem with Dojo: Could not load 'dojo.parser'

Dojo xhr to update a div

Code formatting in Blogger

Hacking an WSRP portlet request using Groovy

Opening JBoss AS HTTP Traffic Up To More Than Localhost

Turning on Hibernate Logging using Seam on JBoss 4.2

Eclipse RCP missing the Business Desktop opportunity?

Using seam-gen.reveng.xml to control tables used for entity generation

junit ant task problems in eclipse

Remote debugging in JBoss

Hibernate alternatives for mappedBy to a superclass property

@MappedSuperclass with Column per Subclass

Relationship Column per Subclass with Discriminator Column

Unidirectional Read Only Back-Reference

OC4J Rmi Client: NoClassDefFoundError: javax/ejb/EJBHome

Canonical Python str method