Posts

I need to resume my blog to talk about some projects I’ve been working on and I’ve migrated over to using Hugo as the blogging platform from my old MiniBlog based solution. The main reasons are ease of writing since the content typically has some code in it, and with Hugo I can write in Markdown. Went relatively easily since I didn’t have that many posts to fix but did find some that had gone missing 😄.

I’ve released NCheck V4, largely tidy up and removal of legacy code and changed the framework to support net472, net48 and netstandard2.0.

The dictionary/list checking has been reworked as well so we can get per-element reporting e.g. if we look at the list test fixture we can see which elements are not in alignment,

We now get error messages like this…

SampleList.Children
Count: Expected:<3>. Actual:<2>
[1].Id: Expected:<2>. Actual:<4>

Useful when you have more than a couple of items in the list.

NCheck V3, has just been released, not much change but it does now support net35, net452, net462 and netstandard1.5 and above.

My main project is using OData a lot and one of the requirements was to ensure that certain properties should not be updatable via PATCH - this is to ensure the integrity of the object e.g. audit fields should not be changeable post-hoc.

There’s a couple of possibilities, depending on your use case…

1. You want to exclude the changes if they are supplied
2. You want to throw an error if non-editable fields are updated.

The way we approached this was to start with an attribute

I’ve released a little project Meerkat.NameParser which assists in splitting up personal names into their component parts.

This code has been around in one form or another since around 1988 when I wrote the first version in Informix 4GL as part of a project for a client. Subsequent versions were written in Clipper, Visual Basic and this one in C#.

It’s not perfect by any means but on projects I’ve used it on we were getting error rates of ~2 per 100K.