Regex constraints (in schema) not working

[scrybbler]

There are a few regex patterns in my schema, some more ambitious than others. I tested my regex at http://regexpal.com/ so it's DEFINITELY working (ha ha... anything tested on the internet is guaranteed to work!)

But anyhow, the Postal Code pattern is a lot easier to test and parse than the Hours pattern. The postal code pattern violations occur on lines 6 and 8 of column F. Hours violations at 3,4, 16, and 17 column J.

Here are my csv and schema respectively: https://drive.google.com/file/d/0B3CDKorSOdGNNm1CbXFrUmlFY1U/view?usp=sharing https://drive.google.com/file/d/0B3CDKorSOdGNN3JlZ2pDLWNKaXc/view?usp=sharing

The validator reports no errors or warnings on these data columns.

[ldodds]

See comment on #183, there are some issues in the schema which means its not being applied.

I also tried the homepage regex at http://regexpal.com/, it also gives me an error (the asterisk is highlighted in red) using that service. I'd recommend Rubular though as its using ruby the same as csvlint.

[scrybbler]

Looks like the Postal Code pattern is working with a valid schema! Here's the schema: https://drive.google.com/file/d/0B3CDKorSOdGNd0ZMWDVZVlIwUTQ/view?usp=sharing

But the Hours pattern is not. Specifically, the problem cases that I inserted on lines 3, 4, 16, and 17 (column J) are not being flagged.

Rubular seems to be ok with my regex, which I admit may be dodgy.

Also, there doesn't seem to be a way to build a schema that does pattern-matching on Hours that ignores empty Hours fields. In other words, if you want to pattern-match on a field in your CSV, any record with null in that field will be flagged. I tried building one with a special case pattern that looks for nulls (https://drive.google.com/file/d/0B3CDKorSOdGNNU1sTjdqYkpmYlE/view?usp=sharing), but it then throws no errors on any of the Hours fields. Rubular is ok with this syntax but I admit I have no idea if CSVLint supports the eol/start-of-line business.