-
Notifications
You must be signed in to change notification settings - Fork 134
/
charwidths.jl
157 lines (136 loc) · 5.95 KB
/
charwidths.jl
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
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
# Following work by @jiahao, we compute character widths using a combination of
# * advance widths from GNU Unifont (advance width 512 = 1 en)
# * UAX 11: East Asian Width
# * a few exceptions as needed
# Adapted from http://nbviewer.ipython.org/gist/jiahao/07e8b08bf6d8671e9734
#
# Requires Julia (obviously) and FontForge.
#############################################################################
# Julia 0.3/0.4 compatibility (taken from Compat package)
if VERSION < v"0.4.0-dev+1419"
const UInt16 = Uint16
end
#############################################################################
# Widths from GNU Unifont
universion="7.0.06"
for fontfile in ["unifont-$universion", "unifont_upper-$universion"]
isfile("$fontfile.ttf") || download("http://unifoundry.com/pub/unifont-$universion/font-builds/$fontfile.ttf", "$fontfile.ttf")
isfile("$fontfile.sfd") || run(`fontforge -lang=ff -c "Open(\"$fontfile.ttf\");Save(\"$fontfile.sfd\");Quit(0);"`)
end
#Read sfdfile for character widths
function parsesfd(filename::String, CharWidths::Dict{Int,Int}=Dict{Int,Int}())
state=:seekchar
lineno = 0
for line in readlines(open(filename))
lineno += 1
if state==:seekchar #StartChar: nonmarkingreturn
if contains(line, "StartChar: ")
codepoint = nothing
width = nothing
state = :readdata
end
elseif state==:readdata #Encoding: 65538 -1 2, Width: 1024
contains(line, "Encoding:") && (codepoint = int(split(line)[3]))
contains(line, "Width:") && (width = int(split(line)[2]))
if codepoint!=nothing && width!=nothing && codepoint >= 0
CharWidths[codepoint]=width
state = :seekchar
end
end
end
CharWidths
end
CharWidths=parsesfd("unifont-$universion.sfd")
CharWidths=parsesfd("unifont_upper-$universion.sfd", CharWidths)
# convert from advance width (512 units to the en) to character width
for (c,v) in CharWidths
CharWidths[c] = div(v, 512)
end
#############################################################################
# Widths from UAX #11: East Asian Width
isfile("EastAsianWidth.txt") || download("http://www.unicode.org/Public/UNIDATA/EastAsianWidth.txt", "EastAsianWidth.txt")
for line in readlines(open("EastAsianWidth.txt"))
#Strip comments
line[1] == '#' && continue
precomment = split(line, '#')[1]
#Parse code point range and width code
tokens = split(precomment, ';')
length(tokens) >= 2 || continue
charrange = tokens[1]
width = strip(tokens[2])
#Parse code point range into Julia UnitRange
rangetokens = split(charrange, "..")
charstart = uint32("0x"*rangetokens[1])
charend = uint32("0x"*rangetokens[length(rangetokens)>1 ? 2 : 1])
#Assign widths
for c in charstart:charend
width=="N" && continue #Ignore neutral characters
CharWidths[c]=(width=="W" || width=="F") ? 2 : #Wide or full
(width=="Na"|| width=="H" || width=="A") ? 1 : #Narrow or half or ambiguous (default to narrow in non-East-Asian contexts, which we can assume to be the default)
error("Unknown East Asian width code: $width for code point: $c")
end
end
#############################################################################
# A few exceptions to the above cases, found by manual comparison
# to other wcwidth functions.
# Use ../libutf8proc for category codes, rather than the one in Julia,
# to minimize bootstrapping complexity when a new version of Unicode comes out.
function catcode(c)
uint(c) > 0x10FFFF && return 0x0000 # see utf8proc_get_property docs
return unsafe_load(ccall((:utf8proc_get_property,"../libutf8proc"), Ptr{UInt16}, (Int32,), c))
end
# use Base.UTF8proc module to get category codes constants, since
# we aren't goint to change these in utf8proc.
import Base.UTF8proc
# make sure format control character (category Cf) have width 0,
# except for the Arabic characters 0x06xx (see unicode std 6.2, sec. 8.2)
for c in keys(CharWidths)
if catcode(c)==UTF8proc.UTF8PROC_CATEGORY_CF &&
c ∉ [0x0601,0x0602,0x0603,0x06dd]
CharWidths[c]=0
end
end
#By definition, should have zero width (on the same line)
#0x002028 '
' category: Zl name: LINE SEPARATOR/
#0x002029 '
' category: Zp name: PARAGRAPH SEPARATOR/
CharWidths[0x2028]=0
CharWidths[0x2029]=0
#By definition, should be narrow = width of 1 en space
#0x00202f ' ' category: Zs name: NARROW NO-BREAK SPACE/
CharWidths[0x202f]=1
#By definition, should be wide = width of 1 em space
#0x002001 ' ' category: Zs name: EM QUAD/
#0x002003 ' ' category: Zs name: EM SPACE/
CharWidths[0x2001]=2
CharWidths[0x2003]=2
#############################################################################
# Non-printable control characters will be assigned a width of zero
# (wcwidth returns -1 for such characters)
isprintable(c::Union(Char,Integer)) = c <= 0x10ffff && is_valid_char(c) && isprintable_category(catcode(c))
isprintable_category(category) =
!( category==UTF8proc.UTF8PROC_CATEGORY_CN # Unassigned
|| category==UTF8proc.UTF8PROC_CATEGORY_CS # Surrogate
|| category==UTF8proc.UTF8PROC_CATEGORY_CC # Control
|| category==0 # Invalid
)
# Question: should we just use Julia's isprint algorithm here? It is different,
# though it is also based on the character category.
#############################################################################
# Output (to a file or pipe) for processing by data_generator.rb
# ... don't bother to output zero widths since that will be the default.
firstc = 0x000000
lastv = 0
uhex(c) = uppercase(hex(c,4))
for c in 0x0000:0x110000
v = isprintable(c) ? get(CharWidths, c, 0) : 0
if v != lastv || c == 0x110000
v < 4 || error("invalid charwidth $v for $c")
if firstc+1 < c
println(uhex(firstc), "..", uhex(c-1), "; ", lastv)
else
println(uhex(firstc), "; ", lastv)
end
firstc = c
lastv = v
end
end