Hi

PowerShell has a lot of the things one might find i SQL; select, where, group, sum. But one thing it lacks i "JOIN". So I've created 2 new CmdLets, New-BOList and Join-BOList to try to fill that void.

New-BOList is to used place a arbitrary list inside a BOList-object and give it a name.

Join-BOList takes two BOList-objects and joins it's list's based on a script-block which is added as a parameter. The output will be a list of PSObjects, each with 2 properties named after the names of the 2 BOLists.

Join-BOList can be used with the parameters -left, -right, -inner and -outer in the same way SQL works.

"BO" in BOList stands for Björn Österman and is more an effect of poor imagination than an act of hybris. =). I would love some feedback about a better name for it.

You can download BOListCommands from here

Disclaimer: This software is provided "AS IS" with no warranties of any kind. The entire risk arising out of the use or performance of the software is with you.

Installation:

1. Download and install the MSI-packet
2. Add the CmdLet to the Powershell environment by running:
      Add-PSSnapIn BOList

You need to do the last step everytime you start PowerShell if you don't add it to your PowerShell profile.

So, how can we use it?

One example is comparing 2 directories. The setup is as shown in the picture below:

The 2 directories differ both in the names of the files, and in one case the size.

First we use "dir" on each directory to get the list of files and put it inside a BOList-object

The next step is to join the two lists, but first I want to explain why we need the BOList-object at all.

Actually, I could have created Join-BOList as Join-List and skipped the the BOList-objects, but I really wanted to give the lists a name, this is because it's going to be used several times, and without the names i would have been forced to use generic names like "a" and "b", or "first" and "second".

Now, this is how we join the lists: 

Let's start by explaining the command.

The 2 lists are specified as the 2 first arguments. And as the "-where"-argument there's a scriptblock. The scriptblock is executed once for each combination of one row from each list and needs to return a bool indicating whether the 2 rows should be joined. Inside the scriptblock there will be 2 variables with the same names as the names we specified for the BOLists.

In the example above, join-bolist returns the files that match each other (by name at least). But it would also be interresting to show that are NOT in both directories.

There are a couple of switches we can use.

• -inner (default)
• -left
• -right
• -outer

These work the same way as in SQL. Here's some examples:

Here we see all the files missing from the compared directories and also the files existing in both directories. If we want to concentrate on the missing files we can filter them out like this:

Now we see the files missing from each directory.

But if we remeber correctly, there are 2 files with the same name but different sizes. Lets change the scriptblock to take that into account:

Ah, and there's the file "Pic 064"!

I'll try to get you another example of how you can use it in my next post