Simplify conversion to OTF via ufo2fdk.

The issues with dropped glyphs during OTF conversion can be fixed with a custom glyph file, which addresses an issue with ufo2fdk.

1 files changed, 5 insertions, 50 deletions

diff --git a/scripts/lib/fontbuild/features.py b/scripts/lib/fontbuild/features.py
index 60f5155..fb1f1c9 100755
--- a/scripts/lib/fontbuild/features.py
+++ b/scripts/lib/fontbuild/features.py
@@ -8,17 +8,15 @@ import re
 # http://www.adobe.com/devnet/opentype/afdko/topic_feature_file_syntax.html

 _glyphNameChars = r"[A-Za-z_][\w.]"

 _glyphName = r"%s{,30}" % _glyphNameChars

-_className = re.compile(r"@%s{,29}" % _glyphNameChars)

+_className = r"@%s{,29}" % _glyphNameChars

 _classValToken = (

-    r"(?:%s|%s(?:\s*-\s*%s)?)" % (_className.pattern, _glyphName, _glyphName))

-_classVal = re.compile(

-    r"\[\s*(%s(?:\s+%s)*)\s*\]" % (_classValToken, _classValToken))

-_classDef = re.compile(

-    r"(%s)\s*=\s*%s\s*;" % (_className.pattern, _classVal.pattern))

+    r"(?:%s|%s(?:\s*-\s*%s)?)" % (_className, _glyphName, _glyphName))

+_classVal = r"\[\s*(%s(?:\s+%s)*)\s*\]" % (_classValToken, _classValToken)

+_classDef = re.compile(r"(%s)\s*=\s*%s\s*;" % (_className, _classVal))

 _featureDef = re.compile(

     r"(feature\s+(?P<tag>[A-Za-z]{4})\s+\{.*?\}\s+(?P=tag)\s*;)",

     re.DOTALL)

-_subRuleToken = r"(?:%s|%s)'?" % (_glyphName, _className.pattern)

+_subRuleToken = r"(?:%s|%s)'?" % (_glyphName, _className)

 _subRuleTokenList =  (

     r"\[?\s*(%s(?:\s+%s)*)\s*\]?" % (_subRuleToken, _subRuleToken))

 _subRule = re.compile(

@@ -102,49 +100,6 @@ def _isValidRef(referencer, ref, font):
     return True

 

 

-def _matchLine(line):

-    """Try to match a line against some feature file language patterns."""

-

-    match = _classDef.match(line)

-    if match:

-        return match, "classDef"

-    match = _subRule.match(line)

-    if match:

-        return match, "subRule"

-    return None, ""

-

-

-def replaceFeatureFileReferences(font, replace):

-    """Replace references according to a given mapping of old names to new."""

-

-    lines = font.features.text.splitlines()

-    for i, line in enumerate(lines):

-        match, flag = _matchLine(line)

-        if not match:

-            continue

-

-        # check for reference in class definitions

-        if flag == "classDef":

-            name, value = match.groups()

-            value = " ".join([replace.get(n, n) for n in value.split()])

-            lines[i]= "%s = [%s];" % (name, value)

-

-        # check in substitution rules

-        elif flag == "subRule":

-            indentation, subbed, sub = match.groups()

-            subbed = " ".join([replace.get(n, n) for n in subbed.split()])

-            sub = " ".join([replace.get(n, n) for n in sub.split()])

-

-            # put brackets around tokens if they were there before

-            if re.match(r"\s*sub(stitute)?\s+\[.+\]\s+by", line):

-                subbed = "[%s]" % subbed

-            if re.match(r"\s*sub(stitute)?.+by\s+\[.+\]\s*;", line):

-                sub = "[%s]" % sub

-            lines[i] = "%ssub %s by %s;" % (indentation, subbed, sub)

-

-    font.features.text = "\n".join(lines)

-

-

 def generateFeatureFile(font):

     """Populate a font's feature file text from its classes and features."""