[HarfBuzz] Quick show of hands needed

[Behdad Esfahbod]

On 07/20/11 23:34, Pravin Satpute wrote:
> On Thursday 21 July 2011 08:51 AM, Behdad Esfahbod wrote:
>> I'm wondering: should harfbuzz try compatibility composition/decomposition
>> (NFKC/NFKD) if a font doesn't support a character?
>>
>> It's easy to do so, but I'm sure how desirable that is.
> 
> 1) In pango we have faced problem of backspace due to decomposition.
> i.e. U+0958  was getting decomposed into U+0915 and U+093C, so for entered
> character was required 2 backspaces to delete.

That's independent of how the shaping engine works.


> 2) CCMP feature of OT fonts also provide this options to font developer. (i
> have used this lot in Arabic fonts since U+0600 charts contains )

Right.  There are three different things:

  - 'ccmp' is supposed to do canonical composition/decomposition.  We already
implement that.

  - The shaping engine can use Unicode canonical composition/decomposition to
workaround incomplete 'ccmp' features.  I'm working on doing that right now.

  - The Unicode *compatibility* decomposition can be used to do fallback
shaping of characters that are not resolved by the above paths.  Compatibility
decomposition by definition means that some meaning is lost in the rendering.
 This is the case I'm curious about.

behdad


> 
> Regards,
> Pravin S