Merge lp://staging/~jdstrand/webapps-core/lp1296170 into lp://staging/webapps-core

Is there a specific reason why you have something like [a-z]+ and not e.g. [a-z][a-z0-9]+ ?

Also, the url patterns specified at this level (specified in --webappUrlPatterns) are actually a weak versions of wildcard patterns (with '?' as an extra pattern char). So full blown regexp should not really used (although not strictly forbidden). In this context maybe 'https?://*.facebook.com/*' would match your needs a bit more (it'll be translated to 'https?://[^\\./]*.facebook.com/*' ?

I just didn't want '*' in there and they seem to be using one letter hostnames, so I went with that. Note this:
  [a-z][a-z0-9]+

requires at least a two-character hostname and would not resolve the issue for me.

I forgot about the translation of 'https?://*.facebook.com/*' to 'https?://[^\\./]*.facebook.com/*'. I would be fine with: 'https?://*.facebook.com/*'. Should I resubmit?

> I just didn't want '*' in there and they seem to be using one letter
> hostnames, so I went with that. Note this:
> [a-z][a-z0-9]+
>
> requires at least a two-character hostname and would not resolve the issue for
> me.

right,

>
> I forgot about the translation of 'https?://*.facebook.com/*' to
> 'https?://[^\\./]*.facebook.com/*'. I would be fine with:
> 'https?://*.facebook.com/*'. Should I resubmit?

I'd say yes, just to avoid too much heterogeneous approaches when we can,

Works