Commit 465a53243f256cefad250b4140b3c35a29fcfeca
2005-03-10T18:31:49
* src/tools/glnames.py: adding comment explaining the compression being used for the Adobe Glyph List.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95
diff --git a/ChangeLog b/ChangeLog
index 707c1ec..9aa9709 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+2005-03-10 David Turner <david@freetype.org>
+
+ * src/tools/glnames.py: adding comment explaining the compression
+ being used for the Adobe Glyph List.
+
2005-03-10 Werner Lemberg <wl@gnu.org>
* src/truetype/ttpload.c (tt_face_load_cvt, tt_face_load_fpgm):
diff --git a/src/tools/glnames.py b/src/tools/glnames.py
index 82d959e..8b9b374 100644
--- a/src/tools/glnames.py
+++ b/src/tools/glnames.py
@@ -4758,6 +4758,77 @@ class StringTable:
write( line + "\n };\n\n\n" )
+#
+# here's an explanation about the way we now store the Adobe Glyph List.
+# First of all, we store the list as a tree. Consider for example that
+# you want to store the following name mapping:
+#
+# A => 1
+# Aacute => 6
+# Abalon => 2
+# Abstract => 4
+#
+# it's possible to store them in a tree, as in:
+#
+# A => 1
+# |
+# +-acute => 6
+# |
+# +-b
+# |
+# +-alone => 2
+# |
+# +-stract => 4
+#
+# we see that each node in the tree has:
+#
+# - one or more 'letters'
+# - an optional value
+# - zero or more child nodes
+#
+# you can build such a tree with:
+#
+# root = StringNode( "",0 )
+# for word in map.values():
+# root.add(word,map[word])
+#
+# this will create a large tree where each node has only one letter
+# then call:
+#
+# root = root.optimize()
+#
+# which will optimize the tree by mergin the letters of successive
+# nodes whenever possible
+#
+# now, each node of the tree is stored as follows in the table:
+#
+# - first the node's letters, according to the following scheme:
+#
+#
+# name bitsize description
+# -----------------------------------------
+# notlast 1 set to 1 if this is not the last letter
+# in the word
+# ascii 7 the letter's ASCII value
+#
+# - then, the children count and optional value:
+#
+# name bitsize description
+# -----------------------------------------
+# hasvalue 1 set to 1 if a 16-bit Unicode value follows
+# num_children 7 number of childrens. can be 0 only if
+# 'hasvalue' is set to 1
+# if (hasvalue)
+# value 16 optional Unicode value
+#
+# - followed by the list of 16-bit absolute offsets to the children.
+# Children must be sorted in increasing order of their first letter.
+#
+# All 16-bit quantities are stored in big-endian. If you don't know why,
+# you've never debugged this kind of code ;-)
+#
+# Finally, the root node has first letter = 0, and no value.
+#
class StringNode:
def __init__( self, letter, value ):
self.letter = letter