d重构代码与优化表

L1: foreach (...) {
    ...
    switch (auto ret = 重构代码(...)) {
        case 1: break L1;
        case -1: continue;
    }//这样来实现`断/下`.
    ...
}

---

我写了个CP949转换表,别人建议用shared static this()来初化不变关联数组.成功.
但dmd在有优化的发布版编译失败.
dmd -release -O失败.

有个ICE的编译器漏洞.可不用phobos,再报告漏洞.

这是内存问题.因为dmd对每个值等在静态数组发出指令.在优化前,内存就不足了.

用'dmd -O'花了一分钟,一定有问题.也许可用dmd -O -lowmem.
或许可分开,再合并为AA:

ushort[] keys = /* ... */;
ushort[] values = /* ... */;

这样-O更好.37秒,稍微好点.代码:

import std.stdio;
import std.algorithm;
import std.range;

void main() {
  auto f = File("deleteme.d", "w");
//生成文件.
  enum lineFormat = "%-( %-( 0x%04X,%|%)\n%)";

  auto keys = cp949_table.keys.sort;
  auto values = keys.map!(key => cp949_table[key]);

  f.writefln!("ushort[] keys = [\n" ~ lineFormat ~ "\n];")(keys.chunks(8));
  f.writefln!("ushort[] values = [\n" ~ lineFormat ~ "\n];")(values.chunks(8));
}

再复制过来:

pure auto make_cp949_table() {
  ushort[] keys = [
    0x8141, 0x8142, 0x8143, 0x8144, 0x8145, 0x8146, 0x8147, 0x8148,
    // ...
    0xFDF7, 0xFDF8, 0xFDF9, 0xFDFA, 0xFDFB, 0xFDFC, 0xFDFD, 0xFDFE,
   ];

  ushort[] values = [
    0xAC02, 0xAC03, 0xAC05, 0xAC06, 0xAC0B, 0xAC0C, 0xAC0D, 0xAC0E,
    // ...
    0x7199, 0x71B9, 0x71BA, 0x72A7, 0x79A7, 0x7A00, 0x7FB2, 0x8A70,
   ];

  /*
    如下编译失败,生成段错误:
    import std.array : assocArray;
    return assocArray(keys, values);
  */

  import std.range : zip;

  ushort[ushort] result;
  foreach (t; zip(keys, values)) {
    result[t[0]] = t[1];
  }

  return result;
}

shared static this() {
  cp949_table = make_cp949_table();
}

dmd对这些表的优化,仍然很差.

对查找表,可用树表(trie),生成只读全局数组.就很快了.