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
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
|
defmodule String.Tokenizer do
@moduledoc false
data_path = Path.join(__DIR__, "UnicodeData.txt")
{letter_uptitlecase, start, continue, _} =
Enum.reduce(File.stream!(data_path), {[], [], [], nil}, fn line, acc ->
{letter_uptitlecase, start, continue, first} = acc
[codepoint, line] = :binary.split(line, ";")
[name, line] = :binary.split(line, ";")
[category, _] = :binary.split(line, ";")
{codepoints, first} =
case name do
"<" <> _ when is_integer(first) ->
last = String.to_integer(codepoint, 16)
{Enum.to_list(last..first), nil}
"<" <> _ ->
first = String.to_integer(codepoint, 16)
{[first], first + 1}
_ ->
{[String.to_integer(codepoint, 16)], nil}
end
cond do
category in ~w(Lu Lt) ->
{codepoints ++ letter_uptitlecase, start, continue, first}
category in ~w(Ll Lm Lo Nl) ->
{letter_uptitlecase, codepoints ++ start, continue, first}
category in ~w(Mn Mc Nd Pc) ->
{letter_uptitlecase, start, codepoints ++ continue, first}
true ->
{letter_uptitlecase, start, continue, first}
end
end)
prop_path = Path.join(__DIR__, "PropList.txt")
{start, continue, patterns} =
Enum.reduce(File.stream!(prop_path), {start, continue, []}, fn line, acc ->
[codepoints | category] = :binary.split(line, ";")
pos =
case category do
[" Other_ID_Start" <> _] -> 0
[" Other_ID_Continue" <> _] -> 1
[" Pattern_White_Space" <> _] -> 2
[" Pattern_Syntax" <> _] -> 2
_ -> -1
end
if pos >= 0 do
entries =
case :binary.split(codepoints, "..") do
[<<codepoint::4-binary, _::binary>>] ->
[String.to_integer(codepoint, 16)]
[first, <<last::4-binary, _::binary>>] ->
Enum.to_list(String.to_integer(last, 16)..String.to_integer(first, 16))
end
put_elem(acc, pos, entries ++ elem(acc, pos))
else
acc
end
end)
id_upper = letter_uptitlecase -- patterns
id_start = start -- patterns
id_continue = continue -- patterns
unicode_upper = Enum.filter(id_upper, &(&1 > 127))
unicode_start = Enum.filter(id_start, &(&1 > 127))
unicode_continue = Enum.filter(id_continue, &(&1 > 127))
rangify = fn [head | tail] ->
{first, last, acc} =
Enum.reduce(tail, {head, head, []}, fn
number, {first, last, acc} when number == first - 1 ->
{number, last, acc}
number, {first, last, acc} ->
{number, number, [{first, last} | acc]}
end)
[{first, last} | acc]
end
@compile {:inline, ascii_upper?: 1, ascii_start?: 1, ascii_continue?: 1}
defp ascii_upper?(entry), do: entry in ?A..?Z
defp ascii_start?(?_), do: true
defp ascii_start?(entry), do: entry in ?a..?z
defp ascii_continue?(entry), do: entry in ?0..?9
range = rangify.(unicode_upper)
for {first, last} <- range do
if first == last do
defp unicode_upper?(unquote(first)), do: true
else
defp unicode_upper?(entry) when entry in unquote(first)..unquote(last), do: true
end
end
defp unicode_upper?(_), do: false
range = rangify.(unicode_start)
for {first, last} <- range do
if first == last do
defp unicode_start?(unquote(first)), do: true
else
defp unicode_start?(entry) when entry in unquote(first)..unquote(last), do: true
end
end
defp unicode_start?(_), do: false
unless {13312, 19893} in range do
raise "CHECK: CJK Ideograph not in range"
end
for {first, last} <- rangify.(unicode_continue) do
if first == last do
defp unicode_continue?(unquote(first)), do: true
else
defp unicode_continue?(entry) when entry in unquote(first)..unquote(last), do: true
end
end
defp unicode_continue?(_), do: false
# Pattern is used as a performance check since most
# atoms and variables end with an atom character.
for {first, last} <- rangify.(patterns), last <= 127 do
if first == last do
defp ascii_pattern?(unquote(first)), do: true
else
defp ascii_pattern?(entry) when entry in unquote(first)..unquote(last), do: true
end
end
defp ascii_pattern?(_), do: false
def tokenize([head | tail]) do
cond do
ascii_upper?(head) ->
validate(continue(tail, [head], 1, true, []), :alias)
ascii_start?(head) ->
validate(continue(tail, [head], 1, true, []), :identifier)
unicode_upper?(head) ->
validate(continue(tail, [head], 1, false, []), :atom)
unicode_start?(head) ->
validate(continue(tail, [head], 1, false, []), :identifier)
true ->
{:error, :empty}
end
end
def tokenize([]) do
{:error, :empty}
end
defp continue([?! | tail], acc, length, ascii_letters?, special) do
{[?! | acc], tail, length + 1, ascii_letters?, [?! | special]}
end
defp continue([?? | tail], acc, length, ascii_letters?, special) do
{[?? | acc], tail, length + 1, ascii_letters?, [?? | special]}
end
defp continue([?@ | tail], acc, length, ascii_letters?, special) do
continue(tail, [?@ | acc], length + 1, ascii_letters?, [?@ | List.delete(special, ?@)])
end
defp continue([head | tail] = list, acc, length, ascii_letters?, special) do
cond do
ascii_start?(head) or ascii_upper?(head) or ascii_continue?(head) ->
continue(tail, [head | acc], length + 1, ascii_letters?, special)
not ascii_pattern?(head) and
(unicode_start?(head) or unicode_upper?(head) or unicode_continue?(head)) ->
continue(tail, [head | acc], length + 1, false, special)
true ->
{acc, list, length, ascii_letters?, special}
end
end
defp continue([], acc, length, ascii_letters?, special) do
{acc, [], length, ascii_letters?, special}
end
defp validate({acc, rest, length, ascii_letters?, special}, kind) do
acc = :lists.reverse(acc)
if ascii_letters? or :unicode.characters_to_nfc_list(acc) == acc do
{kind, acc, rest, length, ascii_letters?, special}
else
{:error, {:not_nfc, acc}}
end
end
end
|
