238771 - Want NT3.51 fallback code for EnumFontFamiliesEx

Status: RESOLVED FIXED

(Core Graveyard :: GFX: Win32, enhancement)

Description

[[e-mail]]

Since NT3.51 doesn't support EnumFontFamiliesEx I want to have some fallback
code that calls EnumFontFamilies instead. These functions are quite similar but
the Ex version also returns the unicode subrange bitfield. One of the calls to
EnumFontFamiliesEx does use the bitfield for truetype fonts so on NT3.51 I make
it fall back to !Ex with a flag not to test for truetype fonts. The other call
to EnumFontFamiliesEx does not use the bitfield, so a fallback was not strictly
necessary although I coded it that way for consistency.

Comment 1

[[e-mail]]

Attachment: Proposed patch

Comment 2

[[e-mail]]

Oh, I should point out that these calls are needed a) for bold to work and
b) for the font lists in e.g. font preferences or editor/message compose.

Comment 3

[timeless]

Comment on attachment 144833 [details] [diff] [review]
Proposed patch

>      * EnumFontFamiliesExW is only on NT/2000
should that be changed to NT4+ as it now applies to XP and S2k3 and we know it
doesn't apply to NT3.51

Comment 4

[Ere Maijala (slow)]

Comment on attachment 144833 [details] [diff] [review]
Proposed patch

>+    if (!gGlobalFonts->Count())
...
>+  if (!weightInfo.mFontCount)

I would prefer | == 0| for integers and it would be more consistent with the
current code. Could you change these and the comment as timeless said?
Otherwise, r=ere

Comment 5

[[e-mail]]

Attachment: Addressed comments (!)

Comment 6

[Ere Maijala (slow)]

To me it looks like the same patch as the first one.

Comment 7

[[e-mail]]

Attachment: Fix dyslexia

Sigh... it does help if I cvs diff to 238771.diff before I attach it :-[

Comment 8

[rbs]

jshin, this patch will undo some changes you did earlier. Are you okay? Or will
it regress Korean fonts?

Comment 9

[Jungshik Shin]

(In reply to comment #8)
> jshin, this patch will undo some changes you did earlier. Are you okay? Or will
> it regress Korean fonts?

Actually, the part of the change I made was to check the Unicode subrange for
Devanagari, Tamil and other scripts (that are not covered by TTF OS2 table), I
guess. Anyway, as far as I can tell (and as Neil wrote to me off-line), it
doesn't seem to do any harm because it's just adding a fallback for NT 3.51. 

In short, I'm Ok with the patch.

Comment 10

[rbs]

Comment on attachment 147445 [details] [diff] [review]
Fix dyslexia

sr=rbs

Comment 11

[[e-mail]]

Fix checked in.