c/old/old_std_wc.c · 4589 字节 · 下载原文

/* old_std_wc.c —— 仿 Unix 的 wc（命令行的，`getchar` 状态机 + **EOF 收尾**）
 * old_std_wc.c -- a Unix-like wc (command line, a `getchar` state machine + **EOF termination**)
 *
 * 类别：std
 * Category: std
 * 兼容面：**逐字符读 `getchar()`**、`while ((c = getchar()) != EOF)` 收尾、
 * Compatibility: **reading `getchar()` character by character**, terminating on `while ((c = getchar()) != EOF)`,
 *         `isspace` 状态机、`%d` 输出
 *         an `isspace` state machine, and `%d` output
 * 出处：自写，仿 Unix 第 7 版那种一次只做一件小事的工具。
 * Origin: self-written, imitating the do-one-small-thing tools of Unix Version 7.
 *
 * ## 这个文件改过一次，原因值得记下来
 * ## This file was changed once, and the reason is worth writing down
 *
 * 第一版用的是哨兵字符 `.` 收尾，注释里写着「`EOF` 在本平台等不到」——
 * The first version terminated on the sentinel character `.`, with a comment saying "`EOF` cannot be reached on this platform" --
 * 那是**当时的事实**，绕开它是不得已。根因后来查明并修掉了：不是宿主不给 EOF，
 * that was **the truth at the time**, and working around it was unavoidable. The root cause was later found and fixed: it was not that the host does not provide EOF,
 * 而是 `io.c` 的 `getchar` 从来没把 R0 置 1（`asm("SYSCALL #5, ${1}")` 里的
 * but that `getchar` in `io.c` never set R0 to 1 (the `${1}` inside `asm("SYSCALL #5, ${1}")`
 * `${1}` 不是变量、什么都不生成），于是它一直走**非阻塞**分支，
 * is not a variable and generates nothing), so it always took the **non-blocking** branch,
 * 输入一空就返回 0 —— "读到的字符是 0"看起来就像"没有 EOF 这个信号"。
 * returning 0 as soon as the input was empty -- "the character read is 0" looked exactly like "there is no EOF signal".
 *
 * 现在它走 `SYSCALL #14`（与 #5 同，只差"耗尽时给 EOF(-1) 而不是空行"），
 * Now it goes through `SYSCALL #14` (the same as #5, differing only in "give EOF(-1) when exhausted instead of an empty line"),
 * 所以**这一版用的是老程序原本那句 `!= EOF`** —— 这正是本文件存在的意义：
 * so **this version uses the very `!= EOF` line the original old program had** -- and that is exactly why this file exists:
 * 它是「EOF 收尾能不能用」这条的判据。哨兵收尾的写法在别处另有示例。
 * it is the criterion for the "can EOF termination be used" item. Sentinel termination is demonstrated elsewhere.
 *
 * ## 判据怎么读
 * ## How to read the criterion
 *
 * 喂 `--stdin 'hello world\nfoo bar\n'` 应得「字符 23，词 4，行 2」左右——
 * Feeding `--stdin 'hello world\nfoo bar\n'` should give roughly "23 chars, 4 words, 2 lines" --
 * 具体数字不必与真实 `wc` 对齐（本平台的行尾换行由输入的 `\n` 决定），
 * the exact numbers need not match the real `wc` (on this platform the trailing newline is decided by the input's `\n`),
 * **要看的是"它自己收尾了、不是被超时掐掉的"**（被掐的话宿主会打
 * **what matters is that "it terminated by itself, rather than being cut off by a timeout"** (if it were cut off the host would print
 * `VM execution cancelled`，而且退出码非 0）。
 * `VM execution cancelled`, and the exit code would be non-zero).
 */
#include <stdio.h>
#include <ctype.h>
#include <vml_ui.h>     /* 只为 ui_scr_get_language() —— 老程序不该猜系统语言 */
                            /* only for ui_scr_get_language() -- an old program must not guess the system language */

int main(void)
{
    int c;
    int chars = 0, words = 0, lines = 0;
    int in_word = 0;
    int lang;   /* 界面语言：开局查一次 */
                /* UI language: queried once at start */

    lang = ui_scr_get_language();

    printf(lang == 0 ? "输入文本，Ctrl+D / Ctrl+Z 结束（脚本化运行时 = 输入用完）\n"
                     : "Type some text, end with Ctrl+D / Ctrl+Z (in a scripted run = input exhausted)\n");

    while ((c = getchar()) != EOF) {
        chars++;
        if (c == '\n') lines++;

        if (isspace(c)) {
            in_word = 0;
        } else if (!in_word) {
            in_word = 1;
            words++;
        }
    }

    printf(lang == 0 ? "字符 %d，词 %d，行 %d\n" : "%d chars, %d words, %d lines\n",
           chars, words, lines);
    return 0;
}