Adjust pulli and language specific nuktas

When the Tamil glyphs were no longer scaled,
the spacing of the nuktas and pulli was overlooked,
causing the nuktas to touch the base character.

While touching turned out to be desired for one language,
it was not desired for the other languages. So now,
most of the languages have spacing like before the
changing of the scaling of the glyphs.

To simplify using the language specific nuktas,
specific fonts were created for each language.

FONTLOG.txt

@@ -29,6 +29,8 @@ ChangeLog
 
 Unreleased
 - Slightly expand and improve the character set
+- Move and adjust form of some of the nuktas
+- Provide language specific fonts
 04 Jan 2019 (Dev NLCI) NLCI Tamil fonts version 0.400
 - Keep the new nukta codepoint, but provide all nukta forms
   (including a new double ring) on the old codepoint as well.

source/master.gdl

@@ -63,7 +63,7 @@ bkurm {
 
 jkurm {
     languages = ("xuj");
-    nuka = ring;
+    nuka = dotdbl;
 }
 
 irula {

tests/CTTnukta.htxt

@@ -1,7 +1,11 @@
 RenderingUnknown
-single nukta
+Single nukta (future)
 \u0b95\u0bc1\u0bae\u0bcd\u0baa\U0001133B\u0bc6\U0001133B\u0b9f\U0001133B\u0bc6\U0001133B\u0b95\U0001133B\u0bae\u0bcd\u0baa\u0bc6\U0001133B\u0ba4\u0bc6\U0001133B
 \u0b95\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\u0bc6\U0001133B\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\U0001133B\u0bc6\U0001133B\u0ba9\u0bbe\u0b9f\U0001133B\u0bc6\U0001133B \u0ba4\u0bc6\U0001133B\u0baa\U0001133B\u0bc6\U0001133B
-double nukta
+\u0ba4\U0001133B\u0bcd\u0b95\U0001133B\u0bbe\u0b95\u0bbf\U0001133B\u0b9e\u0bc1\U0001133B\u0b95\u0bc1\U0001133B\u0bcd\u0b95\u0bc1\U0001133B
+\u0b9f\u0bbf\U0001133B \u0b9f\u0bc1\U0001133B \u0b9f\u0bc2\U0001133B \u0ba4\u0bc1\U0001133B \u0ba4\u0bc2\U0001133B\u0bae\u0bcd\u0baa\u0bc1\u0bcd\u0bc7\u0bb2\u0bb5\u0bbf
+Double and single nuktas
 \u0b95\u0bc1\u0bae\u0bcd\u0baa\U0001133C\u0bc6\U0001133C\u0b9f\U0001133C\u0bc6\U0001133C\u0b95\U0001133C\u0bae\u0bcd\u0baa\u0bc6\U0001133C\u0ba4\u0bc6\U0001133C
 \u0b95\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\u0bc6\U0001133C\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\U0001133C\u0bc6\U0001133C\u0ba9\u0bbe\u0b9f\U0001133C\u0bc6\U0001133C \u0ba4\u0bc6\U0001133C\u0baa\U0001133C\u0bc6\U0001133C
+\u0ba4\U0001133C\u0bcd\u0b95\U0001133C\u0bbe\u0b95\u0bbf\U0001133C\u0b9e\u0bc1\U0001133C\u0b95\u0bc1\U0001133C\u0bcd\u0b95\u0bc1\U0001133C
+\u0b9f\u0bbf\U0001133C \u0b9f\u0bc1\U0001133C \u0b9f\u0bc2\U0001133C \u0ba4\u0bc1\U0001133C \u0ba4\u0bc2\U0001133C\u0bae\u0bcd\u0baa\u0bc1\u0bcd\u0bc7\u0bb2\u0bb5\u0bbf

tests/IRUnukta.htxt

@@ -1,7 +1,11 @@
 RenderingUnknown
-single nukta
+Single nukta (future)
 \u0b95\u0bc1\u0bae\u0bcd\u0baa\U0001133B\u0bc6\U0001133B\u0b9f\U0001133B\u0bc6\U0001133B\u0b95\U0001133B\u0bae\u0bcd\u0baa\u0bc6\U0001133B\u0ba4\u0bc6\U0001133B
 \u0b95\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\u0bc6\U0001133B\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\U0001133B\u0bc6\U0001133B\u0ba9\u0bbe\u0b9f\U0001133B\u0bc6\U0001133B \u0ba4\u0bc6\U0001133B\u0baa\U0001133B\u0bc6\U0001133B
-double nukta
+\u0ba4\U0001133B\u0bcd\u0b95\U0001133B\u0bbe\u0b95\u0bbf\U0001133B\u0b9e\u0bc1\U0001133B\u0b95\u0bc1\U0001133B\u0bcd\u0b95\u0bc1\U0001133B
+\u0b9f\u0bbf\U0001133B \u0b9f\u0bc1\U0001133B \u0b9f\u0bc2\U0001133B \u0ba4\u0bc1\U0001133B \u0ba4\u0bc2\U0001133B\u0bae\u0bcd\u0baa\u0bc1\u0bcd\u0bc7\u0bb2\u0bb5\u0bbf
+Double and single nuktas
 \u0b95\u0bc1\u0bae\u0bcd\u0baa\U0001133C\u0bc6\U0001133C\u0b9f\U0001133C\u0bc6\U0001133C\u0b95\U0001133C\u0bae\u0bcd\u0baa\u0bc6\U0001133C\u0ba4\u0bc6\U0001133C
 \u0b95\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\u0bc6\U0001133C\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\U0001133C\u0bc6\U0001133C\u0ba9\u0bbe\u0b9f\U0001133C\u0bc6\U0001133C \u0ba4\u0bc6\U0001133C\u0baa\U0001133C\u0bc6\U0001133C
+\u0ba4\U0001133C\u0bcd\u0b95\U0001133C\u0bbe\u0b95\u0bbf\U0001133C\u0b9e\u0bc1\U0001133C\u0b95\u0bc1\U0001133C\u0bcd\u0b95\u0bc1\U0001133C
+\u0b9f\u0bbf\U0001133C \u0b9f\u0bc1\U0001133C \u0b9f\u0bc2\U0001133C \u0ba4\u0bc1\U0001133C \u0ba4\u0bc2\U0001133C\u0bae\u0bcd\u0baa\u0bc1\u0bcd\u0bc7\u0bb2\u0bb5\u0bbf

tests/XUBnukta.htxt

@@ -1,7 +1,11 @@
 RenderingUnknown
-single nukta
+Single nukta (future)
 \u0b95\u0bc1\u0bae\u0bcd\u0baa\U0001133B\u0bc6\U0001133B\u0b9f\U0001133B\u0bc6\U0001133B\u0b95\U0001133B\u0bae\u0bcd\u0baa\u0bc6\U0001133B\u0ba4\u0bc6\U0001133B
 \u0b95\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\u0bc6\U0001133B\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\U0001133B\u0bc6\U0001133B\u0ba9\u0bbe\u0b9f\U0001133B\u0bc6\U0001133B \u0ba4\u0bc6\U0001133B\u0baa\U0001133B\u0bc6\U0001133B
-double nukta
+\u0ba4\U0001133B\u0bcd\u0b95\U0001133B\u0bbe\u0b95\u0bbf\U0001133B\u0b9e\u0bc1\U0001133B\u0b95\u0bc1\U0001133B\u0bcd\u0b95\u0bc1\U0001133B
+\u0b9f\u0bbf\U0001133B \u0b9f\u0bc1\U0001133B \u0b9f\u0bc2\U0001133B \u0ba4\u0bc1\U0001133B \u0ba4\u0bc2\U0001133B\u0bae\u0bcd\u0baa\u0bc1\u0bcd\u0bc7\u0bb2\u0bb5\u0bbf
+Double and single nuktas
 \u0b95\u0bc1\u0bae\u0bcd\u0baa\U0001133C\u0bc6\U0001133C\u0b9f\U0001133C\u0bc6\U0001133C\u0b95\U0001133C\u0bae\u0bcd\u0baa\u0bc6\U0001133C\u0ba4\u0bc6\U0001133C
 \u0b95\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\u0bc6\U0001133C\u0bc1\u0baf\u0bcd\u0bb0\u0bcd\u0b95\U0001133C\u0bc6\U0001133C\u0ba9\u0bbe\u0b9f\U0001133C\u0bc6\U0001133C \u0ba4\u0bc6\U0001133C\u0baa\U0001133C\u0bc6\U0001133C
+\u0ba4\U0001133C\u0bcd\u0b95\U0001133C\u0bbe\u0b95\u0bbf\U0001133C\u0b9e\u0bc1\U0001133C\u0b95\u0bc1\U0001133C\u0bcd\u0b95\u0bc1\U0001133C
+\u0b9f\u0bbf\U0001133C \u0b9f\u0bc1\U0001133C \u0b9f\u0bc2\U0001133C \u0ba4\u0bc1\U0001133C \u0ba4\u0bc2\U0001133C\u0bae\u0bcd\u0baa\u0bc1\u0bcd\u0bc7\u0bb2\u0bb5\u0bbf

tools/anchors.py

@@ -7,10 +7,19 @@
 # Open UFO
 ufo = sys.argv[1]
 font = OpenFont(ufo)
-print(f'Add anchors for {ufo}')
+print(f'Move anchors for {ufo}')
 
 # Modify UFO
 
+## Move single ring closer to base character
+ring = font['u1133C']
+for anchor in ring.anchors:
+    if anchor.name == '_N':
+        if font.info.familyName == 'ThiruValluvar':
+            anchor.y = -500
+        else:
+            anchor.y = -244
+
 ## Move one nukta to the left
 tti = font['u0B9F_u0BBF']
 for anchor in tti.anchors:

tools/ffaddapstotaml

@@ -18,9 +18,9 @@ vgw = (1 + dotmargin) * 0.5 * (vgr - vgl)
 # ThiruValluvar has a UPM of 2048, the others have a UPM of 1000.
 # Therefore, the location of the nukta AP is different.
 if f.familyname == 'ThiruValluvar':
-    ypos = -500.
+    ypos = -357
 else:
-    ypos = -244.
+    ypos = -174
 
 addAnchorClass(f, 'N')
 for n in ('u1133C.double', 'u1133C.ring', 'u1133C', 'u1133C.ringdbl') :

tools/ffaddvaptotaml

@@ -10,9 +10,9 @@ addAnchorClass(f, 'V')
 # ThiruValluvar has a UPM of 2048, the others have a UPM of 1000.
 # Therefore, the distance from the virama glyph to the AP is different.
 if f.familyname == 'ThiruValluvar':
-    offset = 200
+    offset = 143
 else:
-    offset = 98
+    offset = 70
 
 # Vaigai has two stray contours on the virama glyph.
 # As a result, the bounding box values for this glyph are not useful.

wscript

@@ -28,6 +28,13 @@ DESC_NAME='NLCI-' + script
 DEBPKG='fonts-nlci-' + script
 getufoinfo('source/ThiruValluvar-Regular.ufo')
 
+langinfo = {
+    'xub' : 'Betta Kurumba',
+    'xuj' : 'Jennu Kurumba',
+    'iru' : 'Irula',
+    'ctt' : 'Chetti'
+}
+
 # set test parameters
 TESTSTRING=u'\u0c15'
 #ftmlTest('tools/FTMLcreateList.xsl')
@@ -111,3 +118,18 @@ for f in faces:
             package = p,
             fret = fret(params = '')
         )
+
+        for langcode in langinfo.keys():
+            langname = langinfo[langcode]
+            langfontfilename = tag + f + langname.replace(' ', '') + snf
+            font(target = process(langfontfilename + '.ttf',
+                    cmd('ttfdeflang -d ' + langcode + ' ${DEP} ${TGT}'),
+                    name(tag + ' ' + f + ' ' + langname, lang='en-US', subfamily=(sn))
+                    ),
+                source = fontfilename + '.ttf',
+                opentype = internal(),
+                graphite = internal(),
+                script = 'taml',
+                package = p,
+                fret = fret(params = '')
+            )