summaryrefslogtreecommitdiff
path: root/tex/context/base/font-otl.lua
diff options
context:
space:
mode:
authorContext Git Mirror Bot <phg42.2a@gmail.com>2015-07-01 23:15:05 +0200
committerContext Git Mirror Bot <phg42.2a@gmail.com>2015-07-01 23:15:05 +0200
commit865735f6a380cd9a5b5eb1b64d4c82b78e7c0ed1 (patch)
tree7684538cdd4ba6a2d0ee57005bfdc0d406555bd5 /tex/context/base/font-otl.lua
parentce4c6e98782de9bab14ed3963c403edbac974879 (diff)
downloadcontext-865735f6a380cd9a5b5eb1b64d4c82b78e7c0ed1.tar.gz
2015-07-01 21:42:00
Diffstat (limited to 'tex/context/base/font-otl.lua')
-rw-r--r--tex/context/base/font-otl.lua25
1 files changed, 25 insertions, 0 deletions
diff --git a/tex/context/base/font-otl.lua b/tex/context/base/font-otl.lua
new file mode 100644
index 000000000..58cce837c
--- /dev/null
+++ b/tex/context/base/font-otl.lua
@@ -0,0 +1,25 @@
+if not modules then modules = { } end modules ['font-otl'] = {
+ version = 1.001,
+ comment = "companion to font-ini.mkiv",
+ author = "Hans Hagen, PRAGMA-ADE, Hasselt NL",
+ copyright = "PRAGMA ADE / ConTeXt Development Team",
+ license = "see context related readme files",
+}
+
+-- After some experimenting with an alternative loader (one that is needed for
+-- getting outlines in mp) I decided not to be compatible with the old (built-in)
+-- one. The approach used in font-otn is as follows: we load the font in a compact
+-- format but still very compatible with the ff data structures. From there we
+-- create hashes to access the data efficiently. The implementation of feature
+-- processing is mostly based on looking at the data as organized in the glyphs and
+-- lookups as well as the specification. Keeping the lookup data in the glyphs is
+-- very instructive and handy for tracing. On the other hand hashing is what brings
+-- speed. So, the in the new approach (the old one will stay around too) we no
+-- longer keep data in the glyphs which saves us (what in retrospect looks a bit
+-- like) a reconstruction step. It also means that the data format of the cached
+-- files changes. What method is used depends on that format. There is no fundamental
+-- change in processing, and not even in data organation. Most has to do with
+-- loading and storage.
+
+-- This file is mostly used for experiments (on my machine) before they make it into
+-- the core.