src/tools/unicode-table-generator/src/cascading_map.rs

   1 use crate::fmt_list;
   2 use crate::raw_emitter::RawEmitter;
   3 use std::collections::HashMap;
   4 use std::fmt::Write as _;
   5 use std::ops::Range;
   6
   7 impl RawEmitter {
   8     pub fn emit_cascading_map(&mut self, ranges: &[Range<u32>]) -> bool {
   9         let mut map: [u8; 256] = [
  10             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  11             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  12             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  13             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  14             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  15             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  16             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  17             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  18             0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
  19         ];
  20
  21         let points = ranges
  22             .iter()
  23             .flat_map(|r| (r.start..r.end).into_iter().collect::<Vec<u32>>())
  24             .collect::<Vec<u32>>();
  25
  26         println!("there are {} points", points.len());
  27
  28         // how many distinct ranges need to be counted?
  29         let mut codepoints_by_high_bytes = HashMap::<usize, Vec<u32>>::new();
  30         for point in points {
  31             // assert that there is no whitespace over the 0x3000 range.
  32             assert!(point <= 0x3000, "the highest unicode whitespace value has changed");
  33             let high_bytes = point as usize >> 8;
  34             let codepoints = codepoints_by_high_bytes.entry(high_bytes).or_insert_with(Vec::new);
  35             codepoints.push(point);
  36         }
  37
  38         let mut bit_for_high_byte = 1u8;
  39         let mut arms = Vec::<String>::new();
  40
  41         let mut high_bytes: Vec<usize> =
  42             codepoints_by_high_bytes.keys().map(|k| k.clone()).collect();
  43         high_bytes.sort();
  44         for high_byte in high_bytes {
  45             let codepoints = codepoints_by_high_bytes.get_mut(&high_byte).unwrap();
  46             if codepoints.len() == 1 {
  47                 let ch = codepoints.pop().unwrap();
  48                 arms.push(format!("{} => c as u32 == {:#04x}", high_byte, ch));
  49                 continue;
  50             }
  51             // more than 1 codepoint in this arm
  52             for codepoint in codepoints {
  53                 map[(*codepoint & 0xff) as usize] |= bit_for_high_byte;
  54             }
  55             arms.push(format!(
  56                 "{} => WHITESPACE_MAP[c as usize & 0xff] & {} != 0",
  57                 high_byte, bit_for_high_byte
  58             ));
  59             bit_for_high_byte <<= 1;
  60         }
  61
  62         writeln!(&mut self.file, "static WHITESPACE_MAP: [u8; 256] = [{}];", fmt_list(map.iter()))
  63             .unwrap();
  64         self.bytes_used += 256;
  65
  66         writeln!(&mut self.file, "#[inline]").unwrap();
  67         writeln!(&mut self.file, "pub fn lookup(c: char) -> bool {{").unwrap();
  68         writeln!(&mut self.file, "    match c as u32 >> 8 {{").unwrap();
  69         for arm in arms {
  70             writeln!(&mut self.file, "        {},", arm).unwrap();
  71         }
  72         writeln!(&mut self.file, "        _ => false,").unwrap();
  73         writeln!(&mut self.file, "    }}").unwrap();
  74         writeln!(&mut self.file, "}}").unwrap();
  75
  76         true
  77     }
  78 }