Comments on: Helpful syntax errors

By: Diego Scataglini

Diego Scataglini — Tue, 24 Aug 2010 02:53:22 +0000

Wow just realized how old is this post

By: Diego Scataglini

Diego Scataglini — Tue, 24 Aug 2010 02:50:39 +0000

That’s awesome, great job.
I wonder though if it’d be possible and then if it’d be useful if it also printed out the name of the unclosed element like so:

missing ‘end’ for ‘class’ ‘Spaghetti’ started on line 1.

missing ‘end’ for ‘class Spaghetti’ started on line 1.

Although easier to read/understand, would that be more helpful or is it superfluous information?

By: Jobs In Manchester

Jobs In Manchester — Mon, 02 Aug 2010 10:00:54 +0000

Seems like only a minor update but it will help the debugging process dramatically. It’s always useful to find the offending piece of code as you’ve hugely reduced the area you’re looking for an error in. I’ve spent hours debugging code ust to find on missing end or an open bracket i forgot to close! So frustrating!
Reece

By: Yugui

Yugui — Sat, 05 Dec 2009 01:27:36 +0000

To Evan and ujihisa,

The patch has already been applied to ruby 1.9.

with -W option, you will get the following messages for the example code.

-:6: warning: mismatched indentations at ‘end’ with ‘while’ at 4
-:7: warning: mismatched indentations at ‘end’ with ‘def’ at 3
-:8: warning: mismatched indentations at ‘end’ with ‘class’ at 2
-:9: syntax error, unexpected $end, expecting keyword_end

By: ujihisa

ujihisa — Thu, 03 Dec 2009 20:36:35 +0000

Hi Evan,

> In the future, it might be possible to try and use indentation to try and narrow down where the missing ‘end’ should be. But for now, every little bit helps.

Logic Junction once made exactly what you want. I saw it at RubyKaigi2007.
http://jp.rubyist.net/RubyKaigi2007/Log0609-LT06.html (ja)
Unfortunately now the patch is unavailable…

By: Wilson Bilkovich

Wilson Bilkovich — Tue, 01 Dec 2009 22:33:10 +0000

Charles:
That change is from commit 7537128aa905b1721a58c719d25e44604a5ed6fe

By: Charles Oliver Nutter

Charles Oliver Nutter — Tue, 01 Dec 2009 08:07:59 +0000

What commit was this in? The JRuby parser is basically the same structure as MRI’s, so it would be nice to add some of these smarts to our parser too.

By: Paul Brannan

Paul Brannan — Mon, 30 Nov 2009 22:00:01 +0000

I like this. For some reason I was annoyed when the 1.8 series started issuing errors like this, because they are so cryptic. Even though ruby 1.6′s “parse error” contained less information, it was easier on the eyes.

Now Rubinius gives us the best of both worlds. Nice.

By: Jamie Orchard-Hays

Jamie Orchard-Hays — Mon, 30 Nov 2009 16:17:43 +0000

That is a nice improvement, Evan. Good one.

By: August Lilleaas

August Lilleaas — Mon, 30 Nov 2009 07:44:26 +0000

Let’s hope this change makes it’s way to MRI as well