Re: CSS and copy/pasting guidelines

Some developer's don't wont to get text-transform'ed result:
http://crbug.com/325231



2016年3月26日(土) 10:58 Johannes Wilm <johannes@fiduswriter.org>:

> On Thu, Mar 24, 2016 at 10:27 PM, Hallvord Reiar Michaelsen Steen <
> hsteen@mozilla.com> wrote:
>
>> On Wed, Mar 23, 2016 at 12:07 AM, Johannes Wilm
>> <johannes@fiduswriter.org> wrote:
>> >> The CSSWG discussed the effect of 'text-transform' on copy/pasted text
>> >> last year:
>> >>   https://lists.w3.org/Archives/Public/www-style/2015Apr/0282.html
>> >
>> > This sounds like this is about the plain text version of the copied
>> > contents. There it may make sense.
>> >
>> > Just to make sure: the html/richtext version should probably not do
>> anything
>> > like this, as JS editor programmers will want o manipualte clipboard
>> > contents before the actual pasting is happening, cleaning it up in ways
>> we
>> > cannot quite imagine here right now.
>> >
>> > Also, I wonder if this is something the clipboard API people have an
>> opinion
>> > about (CC Hallvord).
>>
>> I do have an opinion (FWIW), in short we should strive to not surprise
>> the user. If the user thinks s/he is copying upper-case text we should
>> paste upper-case text. So that's what I'd recommend doing for
>> plain-text. For rich text we should grab the original source code
>> including the case and preferably include a CSS fragment that does the
>> transform.
>>
>
>
> If we can do it in a way that is predictable to the web developer and also
> something that one can get around by filtering the HTML before inserting
> the paste, I think that will be fine.
>
> For richtext, what I would like not to happen is for the browser people to
> make their own interpretations of what is supposed to happen without
> letting the the web developer have a chance of intervening. Things like
> "Oh, the user is pasting some text from Ms Word that contains footnotes.
> Well, we're on the web, where footnotes don't exist, so let's just get rid
> of those. Surely the user didn't really want them anyway" will turn into a
> headache for a lot of JavaScript developers. Same applies if one just
> capitalizes text directly in the richtext-version instead of doing it
> through CSS.
>
> What the default type of richtext-pasting looks like is less important, I
> think. As long as things can be undone in JavaScript, it's fine to make
> assumptions for how to paste by default.
>
> I agree with Yoshifumi that it is not obvious for users why generated
> content is not selectable and cannot be copied, not even in the plaintext
> version.
>
>
>
>
>> >> One principle that was proposed is that CSS in general shouldn't affect
>> >> copy/paste operations except for
>> >>   * generated content
>> >>   * the 'display' property
>>
>> What about hidden? If I have this document:
>> <p style="visibility:hidden">Hello world!</p>
>> and do ctrl-a, ctrl-c - what should the plain text on the clipboard
>> be, if anything? Or opacity:0? This is perhaps best covered by people
>> spec'ing Selection-stuff though.
>> -Hallvord
>>
>
> --
> Johannes Wilm
> Fidus Writer
> http://www.fiduswriter.org
>