cesu8str 0.2.4

Convert to and from CESU-8 or Modified UTF-8 encodings, only when necessary.
Documentation 
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
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268

use std::collections::VecDeque;
use std::fs::File;
use std::io::{self, Read};
use std::str::Utf8Error;

macro_rules! cow {
    ($e: expr) => {
        std::borrow::Cow::from($e)
    };
}

#[derive(Default, Clone, Copy, Debug)]
struct Counters {
    len4: usize,
    len3: usize,
    len2: usize,
    ascii: usize,
    nul: usize,
}
impl Counters {
    fn count(chunk: &[u8]) -> Counters {
        let mut counts = Counters::default();
        for b in chunk {
            // if we aren't on a continuation byte
            if b & 0b11_00_0000 != 0b10_00_0000 {
                match *b {
					// 0 => counts.nul += 1,
					b if b               == 0b0000_0000 => counts.nul += 1,
					b if b & 0b1000_0000 == 0b0000_0000 => counts.ascii += 1,
					b if b & 0b1110_0000 == 0b1100_0000 => counts.len2 += 1,
					b if b & 0b1111_0000 == 0b1110_0000 => counts.len3 += 1,
					b if b & 0b1111_1000 == 0b1111_0000 => counts.len4 += 1,
					b => unreachable!("found non-continuation byte that is not nul/ascii/len(2,3,4)-utf8 char: {:X?}", b),
				}
            }
        }
        if chunk.len() != counts.bytes() {
            match std::str::from_utf8(chunk) {
				Ok(_) => panic!("failed to accurately count byte types for valid UTF-8 file (total = {}, calculated = {})", chunk.len(), counts.bytes()),
				Err(_) => {
					eprintln!("invalid UTF-8 passed, converting string to lossy UTF-8. Count may be off (current chunk byte count off by {})", chunk.len() as isize - counts.bytes() as isize);
				},
			}
        }

        // breaks if a chunk doesn't end in valid UTF-8
        counts
    }
    fn chars(&self) -> usize {
        self.len4 + self.len3 + self.len2 + self.ascii + self.nul
    }
    fn bytes(&self) -> usize {
        self.len4 * 4 + self.len3 * 3 + self.len2 * 2 + self.ascii + self.nul
    }
    fn as_json(&self) -> String {
        format!(
            "{{ \"nul\": {}, \"ascii\": {}, \"len2\": {}, \"len3\": {}, \"len4\": {} }}",
            self.nul, self.ascii, self.len2, self.len3, self.len4
        )
    }
}
impl std::ops::Add<Counters> for Counters {
    type Output = Counters;

    #[inline]
    fn add(self, rhs: Counters) -> Counters {
        Counters {
            len4: self.len4 + rhs.len4,
            len3: self.len3 + rhs.len3,
            len2: self.len2 + rhs.len2,
            ascii: self.ascii + rhs.ascii,
            nul: self.nul + rhs.nul,
        }
    }
}
impl std::ops::AddAssign<Counters> for Counters {
    #[inline]
    fn add_assign(&mut self, rhs: Counters) {
        *self = *self + rhs;
    }
}
impl<'b> std::ops::AddAssign<&'b [u8]> for Counters {
    fn add_assign(&mut self, rhs: &'b [u8]) {
        *self = *self + Counters::count(rhs)
    }
}

fn main() {
    let exit_code = real_main();
    std::process::exit(exit_code);
}

fn real_main() -> i32 {
    // skip bin name
    let mut opts: VecDeque<_> = std::env::args().collect();
    let exe = opts
        .pop_front()
        .map(|s| cow!(s))
        .unwrap_or_else(|| cow!("count_bytes"));

    if opts.iter().any(|s| s == "--help") {
        eprintln!("USAGE: {exe:?} [--json] [file] [KB_chunk_size=4]");
        eprintln!();
        eprintln!("Reads a file and emits the number of occurences differing lengths of UTF-8 codepoints on stdout.");
        eprintln!();
        eprintln!("If `file` is not supplied, stdin will be read.");
        eprintln!();
        eprintln!("Exit Codes:");
        eprintln!("0\tSuccess");
        eprintln!("1\tCommand-line error");
        eprintln!("2\tIO Error");
        return 1;
    }

    let output_json = if let Some(i) = opts.iter().position(|s| s == "--json") {
        let _ = opts.remove(i).unwrap();
        true
    } else {
        false
    };

    let fname = opts.pop_front();

    let chunk_size: Option<usize> = match opts
        .pop_front()
        .map(|s| s.parse::<usize>().map_err(|e| (s, e)))
        .transpose()
    {
        Ok(Some(0)) => {
            eprintln!("supplied zero-length chunk size. using default chunk size.");
            None
        }
        Ok(o) => o,
        Err((s, e)) => {
            eprintln!("USAGE: {exe} [--json] [file] [KB_chunk_size=4]");
            eprintln!("error parsing argument `chunk_size` (provided: {s:?}) as positive integer: {e}",);
            return 1;
        }
    };

    let chunk = chunk_size.unwrap_or(4) * 1024; // 4 KiB by default

    let count: io::Result<Result<Counters, Utf8Error>> = match fname.as_deref() {
        None | Some("-") => {
            // eprintln!("using stdin");
            let stdin = io::stdin();
            let mut stdin = stdin.lock();
            readloop(&mut stdin, chunk)
        }
        Some(fname) => File::open(fname).and_then(|mut f| readloop(&mut f, chunk)),
    };

    match count {
        Ok(Ok(n)) if output_json => {
            println!("{}", n.as_json());
            0
        }
        Ok(Ok(n)) => {
            println!("len4: {}", n.len4);
            println!("len3: {}", n.len3);
            println!("len2: {}", n.len2);
            println!("ascii: {}", n.ascii);
            println!("nul: {}", n.nul);
            println!("total (chars/bytes): {}/{}", n.chars(), n.bytes());
            0
        }
        Ok(Err(e)) => {
            eprintln!("error reading input: {e}");
            2
        }
        Err(e) => {
            eprintln!("error reading input: {e}");
            2
        }
    }
}

fn _count_chunk_bytes(chunk: &[u8]) -> usize {
    chunk
        .iter()
        .filter(|&b| b & 0b1111_1000 == 0b1111_0000)
        .count()
}

/// Get the index of an ending, whole UTF-8 character. Assumes input is valid UTF-8.
///
/// If the chunk contains whole characters, this will be `chunk.len()`
fn assumed_utf8_chunk(total_i: usize, chunk: &[u8]) -> usize {
    if chunk.is_empty() {
        return 0;
    }

    // find first non-continuation byte, starting from end
    // return either it's index, or the end of it

    for (i, &b) in chunk.iter().enumerate().rev() {
        // return for ascii, continue on continuation bytes
        let utf_len = match b {
			b if b & 0b1000_0000 == 0b0000_0000 => return i+1, // ascii/nul
			b if b & 0b1110_0000 == 0b1100_0000 => 2, // len2
			b if b & 0b1111_0000 == 0b1110_0000 => 3, // len3
			b if b & 0b1111_1000 == 0b1111_0000 => 4, // len4
			b if b & 0b1100_0000 == 0b1000_0000 => { continue; }, // continuation
			b => unreachable!("found non-continuation byte that is not nul/ascii/len(2,3,4)-utf8 char: {:X?} (at {:x})", b, total_i + i),
		};

        // if there is enough space for the char, return after it - otherwise, return beginning of it
        if i + utf_len > chunk.len() {
            return i;
        } else {
            // note: specifically allowing (i+utf_len == chunk.len) as this should be an exclusive end index
            return i + utf_len;
        }
    }

    // zero bytes or all continuation bytes
    if chunk.is_empty() {
        0
    } else {
        // since this is a utility helper/checker program, I don't mind simply panicking on this case
        panic!("chunk only contains continuation bytes?");
    }
}

fn readloop(input: &mut dyn Read, chunk: usize) -> io::Result<Result<Counters, Utf8Error>> {
    let bufsize: usize = chunk;
    let mut summative = Counters::default();

    // 1 megabyte buffer
    let mut buf = vec![0; bufsize];
    let mut buffered = 0;
    let mut total_i = 0;

    loop {
        // eprintln!("reading, buffered = {}", buffered);
        match input.read(&mut buf[buffered..])? {
            0 => {
                break;
            }
            n => {
                // get valid section
                let valid_up_to = assumed_utf8_chunk(total_i, &buf[..buffered + n]);
                let valid = &buf[..valid_up_to];

                // if valid_up_to != buffered+n {
                // 	eprintln!("note: splitting chunk input[0x{:X}..0x{:X}] as input[0x{:X}..0x{:X}] (last bytes: {:X?})", total_i, total_i+n, total_i, total_i+valid_up_to-buffered, &buf[(buffered+n).saturating_sub(5)..buffered+n]);
                // }

                // count up bytes
                summative += Counters::count(valid);

                // put unused bytes at the beginning
                buf.copy_within(valid_up_to..buffered + n, 0);
                buffered = buffered + n - valid_up_to;
                total_i += n;
            }
        }
    }

    if buffered == 0 {
        Ok(Ok(summative))
    } else {
        let uerr = std::str::from_utf8(&buf[..buffered]).expect_err(
            "if there is buffered data left, there should be a UTF-8 error of some sort",
        );
        Ok(Err(uerr))
    }
}