[gs-cvs] rev 6916 - trunk/gs/lib

[leonardo at ghostscript.com]

Author: leonardo
Date: 2006-07-13 05:25:38 -0700 (Thu, 13 Jul 2006)
New Revision: 6916

Modified:
   trunk/gs/lib/gs_ttf.ps
Log:
Fix (True Type font loader) : A wrong encoding created when no 'post' presents.

DETAILS :

Bug 688789 "incorrect encoding displayed using ArialUnicodeMS".

The old code in .charkeys computes (with dropping minor details) :
    CharStrings = glyphencoding^-1
    Encoding = cmap*glyphencoding

It is correct when 'post' presents due to glyphencoding = post,  
and post maps glyph indices to glyph names.

In the test case (with no 'post') .charkeys must compute (with dropping same details) :
    CharStrings = glyphencoding^-1 * cmap
    Encoding = glyphencoding
because glyphencoding is stubbed with an encoding.

I'm not sure why the old code worked so long with no problem.
Likely every font manufacturers loved to define 'post'.
Now we got one which doesn't.

BTW, I guess an useful optimization would be to replace 'numloca'
with a better estimation. In the test case it appears too big,
so we process a huge trash and then drop it.
The better estimation, the smaller time consumption.

Another useful optimization is marked with
"% a stub for a while. Must skip the entry."
in the new code. Dalaying it for better times.

Inserted a debug printing for CharStrings.
Using it one can se how big trash it processes.

The old code truncates Encoding if it is longer than 'loca'.
It appears wrong with the new CharString composition logics
- creates a tool short Encoding for fonmts in bug_687457.pdf .
The truncation has been removed from .charkeys .
It should work fine because recently we improved
C code about acessing glyphs outside 'loca' -
see the revision 6874.

The test case to be added to comparefiles.

DIFFERENCES :

None.


Modified: trunk/gs/lib/gs_ttf.ps
===================================================================
--- trunk/gs/lib/gs_ttf.ps	2006-07-13 05:38:11 UTC (rev 6915)
+++ trunk/gs/lib/gs_ttf.ps	2006-07-13 12:25:38 UTC (rev 6916)
@@ -1142,7 +1142,17 @@
   } if
 		% Hack: if there is no usable post table but the cmap uses
 		% the Microsoft Unicode encoding, use ISOLatin1Encoding.
+     % if 'post' presents, .charkeys computes (with dropping minor details) :
+     %     CharStrings = glyphencoding^-1
+     %     Encoding = cmap*glyphencoding
+     % because 'post' maps glyph indices to glyph names.
+     % Otherwise .charkeys must compute (with dropping same details) :
+     %     CharStrings = glyphencoding^-1 * cmap
+     %     Encoding = glyphencoding
+     % because glyphencoding is stubbed with an encoding, 
+     % which maps char codes to glyph names.
   glyphencoding length 0 eq {
+    /have_post false def
     cmapsub 0 4 getinterval <00030001> eq {
     PDFDEBUG { (No post but have cmap 3.1, so use ISOLatin1Encoding) = } if
       /glyphencoding ISOLatin1Encoding dup length array copy def
@@ -1150,9 +1160,11 @@
       PDFDEBUG { (No encoding info, use .GS_extended_SymbolEncoding) = } if
       /glyphencoding /.GS_extended_SymbolEncoding findencoding dup length array copy def
     } ifelse
-  } if
+  } {
+    /have_post true def
+  } ifelse
 		% If necessary, fabricate additional glyphencoding entries
-		% to cover all of loca, or truncate glyphencoding.
+		% to cover all of loca
   glyphencoding length numloca lt {
     /glyphencoding numloca array
       glyphencoding length dup 1 sub 0 1 3 2 roll {
@@ -1164,9 +1176,8 @@
         1 index exch dup .nname put
       } for
     def
-  } {
-    /glyphencoding glyphencoding 0 numloca getinterval def
-  } ifelse
+  } if
+  /cmapa cmaptab cmaparray def
 		% Some badly designed Chinese fonts have a post table
 		% in which all glyphs other than 0 are named .null.
 		% Use CharStrings to keep track of the reverse map from
@@ -1175,7 +1186,16 @@
   /CharStrings glyphencoding dup length 1 add dict	% +1 for .notdef
     0 1 3 index length 1 sub {
 		% Stack: glyphencoding dict index
-      2 index 1 index get 2 index 1 index known
+      2 index 1 index get 
+      have_post not { 
+        exch dup cmapa length lt {
+          cmapa exch get 
+        } { 
+          pop 0 % a stub for a while. Must skip the entry.
+        } ifelse
+        exch
+      } if
+      2 index 1 index known
       1 index /.notdef ne and {	%		ignore extra .notdef glyphs
 		% The same name maps to more than one glyph.
 		% Change the name.
@@ -1185,6 +1205,9 @@
     } for exch pop
 		% If there is no .notdef entry, map it to glyph 0.
   dup /.notdef known not { dup /.notdef 0 put } if
+  TTFDEBUG { 
+    (CharStrings:)= dup { exch =string cvs print ( ) print == } forall
+  } if
   //postalias {
     2 index 1 index known {
       pop pop
@@ -1198,10 +1221,16 @@
     } ifelse  
   } forall
   readonly
-  /Encoding
-    [ cmaptab cmaparray dup length 256 gt { 0 256 getinterval } if
-    { glyphencoding exch get } forall
-    counttomark 256 exch sub { /.notdef } repeat ]
+  /Encoding [
+      have_post {
+        cmapa dup length 256 gt { 0 256 getinterval } if
+        { glyphencoding exch get } forall
+      } {
+        glyphencoding  dup length 256 gt { 0 256 getinterval } if
+        aload pop
+      } ifelse
+      counttomark 256 exch sub { /.notdef } repeat 
+    ]
   TTFDEBUG { (Encoding: ) print dup === flush } if
 } bind def
 
@@ -1220,6 +1249,7 @@
 
 % <file> .loadttfont <type42font>
 /.loadttfont {
+  TTFDEBUG { (.loadttfont) = } if
   //false 0 .loadttfonttables
   .makesfnts
   .getpost
@@ -1514,6 +1544,7 @@
 
 % <file> <is_symbolic> <Encoding|null> .loadpdfttfont <type42font>
 /.loadpdfttfont {
+  TTFDEBUG { (.loadpdfttfont Encoding=) print dup == } if
   /prebuilt_encoding gput     % for .pdfcharkeys
   /is_symbolic gput
   //false 0 .loadttfonttables