Re: Proposal: Minor 'text-transform' property revision

On Wed, Feb 19, 2014 at 4:39 PM, Jonathan Kew <>wrote:
> Some possible options:
> (a) No title-casing blacklists; specify explicitly that the transform
> applies to all words. Authors who want better control need to use other
> mechanisms.

This is what the "capitalize" value does.

> authors who care about precise control should not rely on text-transform.

This is why a UA-based approach works well.  A prescriptive word list is
not only difficult (impossible?) to author, but ostensibly saying that,
once it is implemented, you're done.  However, any implementation can only
be more comprehensive, not complete.  For this reason, establishing grammar
rules and a general approach are appropriate.

> (c) Create a new property such as
>   text-transform-titlecase-exceptions: "a" "an" "the";
> that authors can use to control the behavior. This could be used together
> with :lang(...) to provide language-appropriate behavior even in
> multilingual pages -- and the UA's default stylesheet could use it to
> provide some predefined (but customizable) behavior.

Might as well allow this to be applied for all text-transform modes, and
call it  text-transform-exceptions.  The conventions behind the value
syntax should be similar to font-family– a comma-delimited list whose
elements don't always require quotes.

Received on Wednesday, 19 February 2014 23:19:43 UTC