[wip] ...

This commit is contained in:
Klesh Wong 2020-11-30 02:04:21 +08:00
parent fdaeb69ec2
commit 312e049862

249
bin/utf8_nf Executable file
View File

@ -0,0 +1,249 @@
#!/bin/env python3
import gzip
import argparse
import re
import sys
import io
from typing import List, TextIO
def extract_nerdfont_ranges(font_patcher: TextIO):
range_re = re.compile(r"'((Src|Sym)(Start|End))':\s*([xXA-Fa-f0-9]+|None)")
name_re = re.compile(r"'Name':\s+\"(.*?)\"")
for line in font_patcher:
results = range_re.findall(line)
if len(results) != 4:
continue
props = {r[0]: int(r[3], base=16) if r[3] != "None" else None for r in results}
r = (
props["SrcStart"] or props["SymStart"],
props["SrcEnd"] or props["SymEnd"],
)
if r != (0, 0):
m = name_re.search(line)
if not m:
raise Exception("unable to find name: %s" % line)
yield r[0], r[1], m[1]
def inject_nerdfont_ranges(ranges: List[tuple], textin: TextIO, textout: TextIO):
scope = None
comment_char = None
escape_char = None
length = len(ranges)
keyword_re = re.compile(r'<(\w+)>\s+(\S+)\s*')
def charmap_line(start: int, end: int, comment: str):
if end < start:
return
textout.write("<U%04X>" % start)
if end > start:
textout.write("..<U%04X>" % end)
textout.write(" ")
escaped = "".join(map(lambda x: "%sx%02x" % (escape_char, x), chr(start).encode('utf-8')))
textout.write(escaped)
textout.write(" ")
textout.write(comment)
textout.write("\n")
charmap = {
're': re.compile(r'<U([A-Z0-9]+)>(..<U([A-Z0-9]+)>)?\s+(\S+)\s+(.*?)$'),
'cursor': 0,
'writeline': charmap_line
}
def width_line(start: int, end: int, comment: str):
if end < start:
return
textout.write("<U%04X>" % start)
if end > start:
textout.write("...<U%04X>" % end)
textout.write("\t%s\n" % comment)
width = {
're': re.compile(r'<U([A-Z0-9]+)>(...<U([A-Z0-9]+)>)?(\s+)(\d+)$'),
'cursor': 0,
'writeline': width_line,
'comment': '2',
}
def pop_scope():
ss, se, comment = ranges[scope['cursor']]
scope['writeline'](ss, se, scope.get('comment', comment))
scope['cursor'] += 1
return scope['cursor'] < length
for line in textin:
if textout.closed:
return
line = line.strip()
if line.startswith("CHARMAP"):
scope = charmap
elif line.startswith("WIDTH"):
scope = width
elif line.startswith("END "):
while scope['cursor'] < length:
pop_scope()
scope = None
elif comment_char and line.startswith(comment_char):
pass
elif scope and scope['cursor'] < length:
m = scope['re'].match(line)
if not m:
raise Exception("unkown line: " + line)
ts = int(m[1], base=16)
te = int(m[3], base=16) if m[3] else ts
ss, se, comment = ranges[scope['cursor']]
# if ts == 0x5e00:
# import ipdb; ipdb.set_trace()
if te < ss: # no intersection
pass
# elif ts > se:
# while ts > se and pop_scope():
# ss, se, comment = ranges[scope['cursor']]
elif ts >= ss and te <= se: # subset
continue
else:
if ss <= te:
scope['writeline'](ts, ss - 1, m[5])
# if se <= te:
# pop_scope()
while se <= te and pop_scope():
if ts <= se:
scope['writeline'](se + 1, te, m[5])
ss, se, comment = ranges[scope['cursor']]
continue
elif line:
m = keyword_re.match(line)
if m:
if m[1] == "comment_char":
comment_char = m[2]
elif m[1] == 'escape_char':
escape_char = m[2]
textout.write(line + "\n")
def test():
TARGET = (
"<comment_char> %\n"
"<escape_char> /\n\n"
"CHARMAP\n"
"<U0000> /x00 no\n"
"<U0001>..<U0005> /x01 left\n"
"<U0006> /x06 subset\n"
"<U0007> /x07 subset\n"
"<U0008>..<U0009> /x08 right\n"
"<U4E00>..<U4E99> /xe4/xb8/x80 superset\n"
"<U5E00> /x00 gap\n"
"END CHARMAP\n"
"WIDTH\n"
"<U0000>...<U0004>\t0\n"
"<U6F00>...<U7FFF>\t1\n"
"END WIDTH\n"
)
ranges = [
(4, 8, "test"),
(0x4e03, 0x4e0a, "test2"),
(0x4F00, 0x4F00, "test3"),
(0x6F00, 0x6F00, "tail"),
]
EXPECT = (
"<comment_char> %\n"
"<escape_char> /\n\n"
"CHARMAP\n"
"<U0000> /x00 no\n"
"<U0001>..<U0003> /x01 left\n"
"<U0004>..<U0008> /x04 test\n"
"<U0009> /x09 right\n"
"<U4E00>..<U4E02> /xe4/xb8/x80 superset\n"
"<U4E03>..<U4E0A> /xe4/xb8/x83 test2\n"
"<U4E0B>..<U4E99> /xe4/xb8/x8b superset\n"
"<U4F00> /xe4/xbc/x80 test3\n"
"<U5E00> /x00 gap\n"
"<U6F00> /xe6/xbc/x80 tail\n"
"END CHARMAP\n"
"WIDTH\n"
"<U0000>...<U0003>\t0\n"
"<U0004>...<U0008>\t2\n"
"<U4E03>...<U4E0A>\t2\n"
"<U4F00>\t2\n"
"<U6F00>\t2\n"
"<U6F01>...<U7FFF>\t1\n"
"END WIDTH\n"
)
target = io.StringIO(TARGET)
result = io.StringIO()
inject_nerdfont_ranges(ranges, target, result)
if result.getvalue() != EXPECT:
print("\033[42m origin\033[0m")
print(TARGET)
print()
print("\033[42m inject\033[0m")
for r in ranges:
print("%04X %04X %s" % r)
print()
print("\033[42m expect \033[0m", len(EXPECT))
print(EXPECT)
print()
print("\033[42m result \033[0m", len(result.getvalue()))
print(result.getvalue())
else:
print("pass")
if __name__ == "__main__":
parser = argparse.ArgumentParser(
description="patch charmap to make NerdFont icons double width"
)
parser.add_argument(
"-i", "--in-charmap",
dest="in_charmap",
default="/usr/share/i18n/charmaps/UTF-8.gz",
help="input charmap file path")
parser.add_argument(
"-o", "--out",
dest="out_charmap",
default="/usr/share/i18n/charmaps/UTF-8NF.gz",
help="output charmap file path")
parser.add_argument(
"-f", "--font-patcher",
dest="font_patcher",
required=True,
help="file path of font_patcher from NerdFont")
parser.add_argument(
"--plainout",
dest="plainout",
action="store_true",
help="write to stdout in plain-text")
parser.add_argument(
"--test",
dest="test",
action="store_true",
help="run test case")
args = parser.parse_args()
font_patcher = open(args.font_patcher, 'r',)
double_width_ranges = extract_nerdfont_ranges(font_patcher)
if args.test:
test()
exit()
in_charmap = gzip.open(args.in_charmap, 'rt', encoding="ascii")
if args.plainout:
out_charmap = sys.stdout
elif args.out_charmap.endswith('.gz'):
out_charmap = gzip.open(args.out_charmap, 'wt')
else:
out_charmap = open(args.out_charmap, 'wt')
ranges = sorted(double_width_ranges, key=lambda x: x[0])
for r in ranges:
print("%04X-%04X %s" % r)
inject_nerdfont_ranges(ranges, in_charmap, out_charmap)
# add `en_US.UTF-8NF UTF-8NF` to `/etc/locale.gen`
# run `locale-gen`
# update `/etc/locale.conf` to `LANG=en_US.UTF-8NF`
# restart