[HarfBuzz] harfbuzz: Branch 'master' - 5 commits

[Jonathan Kew]

On 17/7/14 20:59, Behdad Esfahbod wrote:
> New commits:
> commit 82f4d9d53f348f41b14b877c1ac77c0372c49caa
> Author: Behdad Esfahbod <behdad at behdad.org>
> Date:   Thu Jul 17 15:57:37 2014 -0400
>
>      [arabic] Add note re disabled 'cswh'
>
> diff --git a/src/hb-ot-shape-complex-arabic.cc 
> b/src/hb-ot-shape-complex-arabic.cc
> index 1b4ebc9..b90abfe 100644
> --- a/src/hb-ot-shape-complex-arabic.cc
> +++ b/src/hb-ot-shape-complex-arabic.cc
> @@ -184,6 +184,15 @@ collect_features_arabic (hb_ot_shape_planner_t *plan)
>     map->add_global_bool_feature (HB_TAG('c','a','l','t'));
>     map->add_gsub_pause (NULL);
>
> +  /* The spec includes 'cswh'.  Earlier versions of Windows
> +   * used to enable this by default, but testing suggests
> +   * that Windows 8 and later do not enable it by default,
> +   * and spec now says 'Off by default'.
> +   * We disabled this in ae23c24c32.
> +   * Note that IranNastaliq uses this feature extensively
> +   * to fixup broken glyph sequences.  Oh well...
> +   * Test case: U+0643,U+0640,U+0631. */
> +  map->add_global_bool_feature (HB_TAG('c','s','w','h'));
>     map->add_global_bool_feature (HB_TAG('m','s','e','t'));
>   }

This looks like it re-enables 'cswh' as an on-by-default feature for 
Arabic. Was that your intention? I believe that'll produce undesired 
behavior with fonts (was Lotus one? I forget) that have 'cswh' features, 
but do not expect them applied by default.

JK