# Exception naming conventions: rely on module name?

**URL:** <https://www.ruby-forum.com/t/exception-naming-conventions-rely-on-module-name/80662>\
**Category:** Ruby\
**Created:** [December 29, 2006, 6:50pm UTC](https://www.ruby-forum.com/t/exception-naming-conventions-rely-on-module-name/80662 "2006-12-29T18:50:20Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![efine](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/a698b9/32.png) [@efine](https://www.ruby-forum.com/u/efine)\
**Post date:** [December 29, 2006, 6:50pm UTC](https://www.ruby-forum.com/t/exception-naming-conventions-rely-on-module-name/80662/1 "2006-12-29T18:50:20Z")

</div>

I am wondering about design alternatives regarding exception naming in  
Ruby. Is there a naming convention for exceptions that are declared  
inside a module?

For example, let’s say we have a module named MP3, and we want to have a  
base exception for MP3-related errors. There are two basic choices I am  
considering here:

module MP3  
class MP3Error \< StandardError; end # Alternative 1  
class Error \< StandardError; end # Alternative 2  
end

Outside the module, Error would be disambiguated by MP3 (MP3::Error).  
However, I am uneasy about this, because of the following situation.  
Imagine someone has some existing code that mixes in module Other:

module Other  
class Error \< StandardError; end  
end

class Foo  
include Other  
def Foo.oops  
raise Error, “Who am I?”  
rescue Error =\> e  
puts “#{e} #{e.class.name}”  
end  
end

Foo.oops # Who am I? Other::Error

Now this hapless person mixes in MP3.

class Foo  
include Other  
include MP3  
def Foo.oops  
raise Error, “Who am I?”  
rescue Error =\> e  
puts “#{e} #{e.class.name}”  
end  
end

Foo.oops # Who am I? MP3::Error

- Is this a real concern?
- Do you think it would be better to have exception names that are less  
likely to clash at the cost of some duplication (e.g. MP3::MP3Error  
instead of MP3::Error)?

---

<div class="post-metadata">

**Author:** ![efine](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/a698b9/32.png) [@efine](https://www.ruby-forum.com/u/efine)\
**Post date:** [December 29, 2006, 10:39pm UTC](https://www.ruby-forum.com/t/exception-naming-conventions-rely-on-module-name/80662/2 "2006-12-29T22:39:57Z")

</div>

On Dec 29, 2006, at 09:50, Edwin F. wrote:

> class Error \< StandardError; end # Alternative 2  
> end

I typically use #2

> def Foo.oops  
> raise Error, “Who am I?”  
> rescue Error =\> e  
> puts “#{e} #{e.class.name}”  
> end  
> end
> 
> Foo.oops # Who am I? Other::Error

Always fully qualify.

> end
> 
> Foo.oops # Who am I? MP3::Error
> 
> - Is this a real concern?

No. Fully qualify and it won’t be a problem.

> - Do you think it would be better to have exception names that are  
> less likely to clash at the cost of some duplication (e.g.  
> MP3::MP3Error instead of MP3::Error)?

MP3::MP3Error is redundant. There is no problem if you raise  
MP3::Error instead. If a consumer of your library decides to omit  
Error that’s their bug, not yours.

–  
Eric H. - removed\_email\_address@domain.invalid - [http://blog.segment7.net](http://blog.segment7.net)

I LIT YOUR GEM ON FIRE!

---

<div class="post-metadata">

**Author:** ![efine](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/a698b9/32.png) [@efine](https://www.ruby-forum.com/u/efine)\
**Post date:** [December 30, 2006, 5:00am UTC](https://www.ruby-forum.com/t/exception-naming-conventions-rely-on-module-name/80662/3 "2006-12-30T05:00:58Z")

</div>

Eric H. wrote:

> MP3::MP3Error is redundant. There is no problem if you raise  
> MP3::Error instead. If a consumer of your library decides to omit  
> Error that’s their bug, not yours.
> 
> –  
> Eric H. - removed\_email\_address@domain.invalid - [http://blog.segment7.net](http://blog.segment7.net)
> 
> I LIT YOUR GEM ON FIRE!

Thank you for your advice. I will follow it.  
Ed