MeowOS-For-R3P2/patch_v2/reader_app/test/seg_smoke.c

/* test/seg_smoke.c — os.meow.reader 分段流式缓存宿主单测（纯 C，无 SDK/Freetype）。
 *
 * 覆盖：
 *   S1 UTF-8 大文件按 ~512KB 切段：段数>1、每段为合法 UTF-8（码位对齐，不切在字符中间）、
 *      拼接后 == 原文件字节。
 *   S2 GBK(GB18030) 大文件切段：每段 raw 在完整双字节边界、逐段 rd_to_utf8 拼接 == 整文件
 *      一次转换（边界不把双字节对拆开）。
 *   S3 幂等：同 (size,mtime,enc) 再次 build 段数不变、不重写。
 *   S4 每段原始字节长与 rdseg_load 读回一致。
 */
#include "seg.h"
#include "rdenc.h"
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

static int g_pass = 0, g_fail = 0;
#define CHECK(cond, msg) do { if (cond) { g_pass++; } else { g_fail++; printf("FAIL %s:%d  %s\n", __FILE__, __LINE__, msg); } } while (0)

/* 生成一个 utf8 大文件：CJK 3 字节 + ASCII + 换行，填充到 > 512KB */
static void write_utf8_big(const char *path)
{
    FILE *f = fopen(path, "wb");
    const uint8_t cjk[] = { 0xE4, 0xBD, 0xA0, 0xE5, 0xA5, 0xBD };            /* 你好 */
    size_t total = 0;
    while (total < RDSEG_SIZE + 100000) {
        for (int i = 0; i < 16; i++) { fwrite(cjk, 1, sizeof cjk, f); total += sizeof cjk; }
        const char *line = "这是一段用于验证分段对齐的中文正文，English words 123.\n";
        size_t ll = strlen(line); fwrite(line, 1, ll, f); total += ll;
    }
    fclose(f);
}

/* 生成一个 GBK 大文件：双字节对（你好世界）重复，填到 > 512KB */
static void write_gbk_big(const char *path)
{
    FILE *f = fopen(path, "wb");
    const uint8_t gbk[] = { 0xC4, 0xE3, 0xBA, 0xC3, 0xCA, 0xC0, 0xBD, 0xE7 };  /* 你好世界 */
    size_t total = 0;
    while (total < RDSEG_SIZE + 100000) {
        for (int i = 0; i < 16; i++) { fwrite(gbk, 1, sizeof gbk, f); total += sizeof gbk; }
        fputc('\n', f); total += 1;
    }
    fclose(f);
}

static void test_align_utf8(const char *seg_dir, const char *path)
{
    /* 每段合法 UTF-8 */
    int n = rdseg_n(seg_dir);
    CHECK(n >= 2, "utf8 segs >= 2");
    for (int i = 0; i < n; i++) {
        size_t rl = 0;
        CHECK(rdseg_rawlen(seg_dir, i, &rl) == 0 && rl > 0, "utf8 rawlen>0");
        uint8_t *b = (uint8_t *)malloc(rl ? rl : 1);
        size_t got = 0;
        CHECK(rdseg_load(seg_dir, i, b, rl, &got) == 0 && got == rl, "utf8 load==rawlen");
        CHECK(rd_utf8_validate(b, got) == 1, "utf8 seg valid (no mid-char split)");
        free(b);
    }
    /* 拼接后 == 原文件字节 */
    size_t fz = 0;
    FILE *src = fopen(path, "rb"); fseek(src, 0, SEEK_END); fz = (size_t)ftell(src); fseek(src, 0, SEEK_SET);
    uint8_t *full = (uint8_t *)malloc(fz ? fz : 1);
    (void)!fread(full, 1, fz, src); fclose(src);
    size_t o = 0, ok = 1;
    for (int i = 0; i < n; i++) {
        size_t rl = 0; rdseg_rawlen(seg_dir, i, &rl);
        uint8_t *b = (uint8_t *)malloc(rl ? rl : 1);
        size_t got = 0; rdseg_load(seg_dir, i, b, rl, &got);
        if (o + got > fz || memcmp(full + o, b, got) != 0) ok = 0;
        o += got; free(b);
    }
    CHECK(ok && o == fz, "utf8 concat == original bytes");
    free(full);
}

static void test_gbk_roundtrip(const char *seg_dir, const char *path)
{
    int n = rdseg_n(seg_dir);
    CHECK(n >= 2, "gbk segs >= 2");
    /* 整文件一次转换（金标准） */
    FILE *src = fopen(path, "rb"); fseek(src, 0, SEEK_END); long fz = ftell(src); fseek(src, 0, SEEK_SET);
    uint8_t *raw = (uint8_t *)malloc((size_t)fz ? (size_t)fz : 1);
    (void)!fread(raw, 1, (size_t)fz, src); fclose(src);
    uint8_t *full = (uint8_t *)malloc((size_t)fz * 2 + 64);
    size_t fl = 0;
    int rc = rd_gb18030_to_utf8(raw, (size_t)fz, full, (size_t)fz * 2 + 64, &fl);
    CHECK(rc == RD_ENC_OK, "gbk full convert ok");
    free(raw);
    /* 逐段转换拼接 */
    uint8_t *concat = (uint8_t *)malloc((size_t)fz * 2 + 64);
    size_t o = 0, ok = 1;
    for (int i = 0; i < n; i++) {
        size_t rl = 0; rdseg_rawlen(seg_dir, i, &rl);
        uint8_t *b = (uint8_t *)malloc(rl ? rl : 1);
        size_t got = 0; rdseg_load(seg_dir, i, b, rl, &got);
        uint8_t *u8 = (uint8_t *)malloc(got * 2 + 64);
        size_t ul = 0;
        if (rd_gb18030_to_utf8(b, got, u8, got * 2 + 64, &ul) != RD_ENC_OK) ok = 0;
        if (o + ul > (size_t)fz * 2 + 64) ok = 0;
        if (ok) { memcpy(concat + o, u8, ul); o += ul; }
        free(u8); free(b);
    }
    CHECK(ok && o == fl && memcmp(concat, full, fl) == 0, "gbk per-seg convert concat == full convert");
    free(concat); free(full);
}

int main(void)
{
    rdseg_set_base("/tmp/meow_seg_base");
    char dir[600];
    const char *u8 = "/tmp/seg_u8.txt", *gbk = "/tmp/seg_gbk.txt";
    write_utf8_big(u8); write_gbk_big(gbk);

    /* UTF-8 */
    CHECK(rdseg_dir_for(u8, dir, sizeof dir) == 0, "utf8 dir_for");
    int n = rdseg_build(u8, RD_ENC_UTF8, dir);
    CHECK(n >= 2, "utf8 build nsegs>=2");
    test_align_utf8(dir, u8);
    /* 幂等：同键再建段数不变 */
    CHECK(rdseg_build(u8, RD_ENC_UTF8, dir) == n, "utf8 build idempotent");

    /* GBK */
    CHECK(rdseg_dir_for(gbk, dir, sizeof dir) == 0, "gbk dir_for");
    n = rdseg_build(gbk, RD_ENC_GB18030, dir);
    CHECK(n >= 2, "gbk build nsegs>=2");
    test_gbk_roundtrip(dir, gbk);

    /* 大文件不切（清空再建，确保得到多个段也覆盖单段路径） */
    {
        char small[600]; rdseg_set_base("/tmp/meow_seg_base2");
        const char *tiny = "/tmp/seg_tiny.txt";
        FILE *f = fopen(tiny, "wb"); fwrite("hello 你好 world\n", 1, strlen("hello 你好 world\n"), f); fclose(f);
        CHECK(rdseg_dir_for(tiny, small, sizeof small) == 0, "tiny dir");
        int tn = rdseg_build(tiny, RD_ENC_UTF8, small);
        CHECK(tn == 1, "tiny build nsegs==1");
        size_t rl = 0; CHECK(rdseg_rawlen(small, 0, &rl) == 0 && rl == strlen("hello 你好 world\n"), "tiny rawlen");
    }

    remove(u8); remove(gbk); remove("/tmp/seg_tiny.txt");
    printf("seg_smoke: %d pass, %d fail\n", g_pass, g_fail);
    return g_fail ? 1 : 0;
}