ny-exam-platform/lib/Roster.php

<?php
/**
 * 学号名单管理
 * 存储：data/roster.json  -> {"students":[{"sid":"1","name":"张三","cls":"计算机2401"}], "updated_at":"..."}
 * cls 为班级（可空），用于一份考试涉及多个班级时区分。
 */
declare(strict_types=1);

class Roster
{
    private static function file(): string
    {
        return NY_DATA . '/roster.json';
    }

    /** 数据库是否可作为名单后端 */
    private static function useDb(): bool
    {
        return class_exists('Db') && Db::pdo() !== null;
    }

    /**
     * 首次启用 SQLite 时，把已有的 data/roster.json 迁进数据库（老数据不丢）。
     * 只在「DB 名单为空 且 文件有数据」时执行一次，之后由 DB 全权负责。
     */
    private static function migrateFileToDbOnce(): void
    {
        if (!self::useDb()) {
            return;
        }
        try {
            if (Db::rosterCount() > 0) {
                return;                       // DB 已有数据，不覆盖
            }
            $d = ny_read_json(self::file(), null);
            if (!is_array($d) || empty($d['students']) || !is_array($d['students'])) {
                return;
            }
            $clean = [];
            foreach ($d['students'] as $s) {
                if (!is_array($s)) {
                    continue;
                }
                $clean[] = [
                    'sid'   => trim((string)($s['sid'] ?? '')),
                    'name'  => trim((string)($s['name'] ?? '')),
                    'cls'   => trim((string)($s['cls'] ?? '')),
                    'grp'   => trim((string)($s['grp'] ?? '')),
                    'login' => trim((string)($s['login'] ?? '')),
                ];
            }
            if ($clean !== []) {
                Db::rosterReplace($clean);
            }
        } catch (Throwable $e) {
            // 迁移失败不影响主流程：仍可退回文件读取
        }
    }

    /** @return array<int,array{sid:string,name:string,cls:string}> */
    public static function all(): array
    {
        if (self::useDb()) {
            self::migrateFileToDbOnce();
            return Db::rosterAll();
        }
        $d = ny_read_json(self::file(), null);
        if (!is_array($d) || !isset($d['students']) || !is_array($d['students'])) {
            return [];
        }
        $out = [];
        foreach ($d['students'] as $s) {
            if (!is_array($s)) {
                continue;
            }
            $out[] = [
                'sid'   => trim((string)($s['sid'] ?? '')),
                'name'  => trim((string)($s['name'] ?? '')),
                'cls'   => trim((string)($s['cls'] ?? '')),
                'grp'   => trim((string)($s['grp'] ?? '')),
                'login' => trim((string)($s['login'] ?? '')),
            ];
        }
        return $out;
    }

    public static function meta(): array
    {
        $d = ny_read_json(self::file(), null);
        return [
            'count'      => self::count(),
            'updated_at' => self::useDb() ? date('Y-m-d H:i:s') : (is_array($d) ? (string)($d['updated_at'] ?? '') : ''),
        ];
    }

    public static function save(array $students): bool
    {
        $clean = [];
        foreach ($students as $s) {
            $sid  = trim((string)($s['sid'] ?? ''));
            $name = trim((string)($s['name'] ?? ''));
            if ($sid === '' && $name === '') {
                continue;
            }
            $clean[] = [
                'sid'   => $sid,
                'name'  => $name,
                'cls'   => trim((string)($s['cls'] ?? '')),
                'grp'   => trim((string)($s['grp'] ?? '')),
                'login' => trim((string)($s['login'] ?? '')),
            ];
        }
        // 按学号去重（同 sid 取最后一条）：roster 表 sid 为主键，重复 sid 会触发约束冲突导致整批回滚。
        // 去重避免「保存成功但名单实际没变」的静默丢失。
        $bySid = [];
        $noSid = [];
        foreach ($clean as $s) {
            if ((string)($s['sid'] ?? '') !== '') {
                $bySid[$s['sid']] = $s;
            } else {
                $noSid[] = $s;
            }
        }
        $clean = array_values($bySid);
        foreach ($noSid as $s) {
            $clean[] = $s;
        }
        if (self::useDb()) {
            return Db::rosterReplace($clean);
        }
        return ny_write_json(self::file(), [
            'students'   => $clean,
            'updated_at' => date('Y-m-d H:i:s'),
        ]);
    }

    public static function clear(): bool
    {
        if (self::useDb()) {
            Db::rosterReplace([]);
            return true;
        }
        return ny_write_json(self::file(), ['students' => [], 'updated_at' => date('Y-m-d H:i:s')]);
    }

    public static function count(): int
    {
        if (self::useDb()) {
            return Db::rosterCount();
        }
        return count(self::all());
    }

    /**
     * 名单指纹：名单内容一变就变，用于数据缓存 key。
     * 只取「人数 + 文件修改时间 + 内容的轻量哈希」，避免每次全量序列化。
     */
    public static function fingerprint(): string
    {
        if (self::useDb()) {
            // 人数 + 内容哈希：名单内容变了指纹才变（只靠人数会在"换人"时误命中缓存）
            $n = self::count();
            $hash = md5(json_encode(self::all(), JSON_UNESCAPED_UNICODE));
            return 'db:' . $n . ':' . $hash;
        }
        $f = self::file();
        $mt = is_file($f) ? (string)filemtime($f) : '0';
        $sz = is_file($f) ? (string)filesize($f) : '0';
        $d  = ny_read_json($f, null);
        $hash = is_array($d) ? md5(json_encode($d['students'] ?? [], JSON_UNESCAPED_UNICODE)) : '0';
        return $mt . ':' . $sz . ':' . $hash;
    }


    /**
     * 解析上传文件 -> 学生数组
     * @return array{count:int,students:array}
     */
    /**
     * 解析上传文件 -> 学生数组
     * @return array{count:int,skipped:int,duplicates:int,headers:array,students:array}
     */
    public static function importFile(string $file, string $ext = ''): array
    {
        if ($ext === '') {
            $ext = strtolower(pathinfo($file, PATHINFO_EXTENSION));
        }
        $rows = [];
        if ($ext === 'xlsx' || $ext === 'xlsm') {
            $rows = XlsxReader::read($file);
        } elseif ($ext === 'csv' || $ext === 'txt') {
            $rows = self::parseCsv($file);
        } else {
            // 尝试按 xlsx 解析，失败再按 csv
            try {
                $rows = XlsxReader::read($file);
            } catch (Throwable $e) {
                $rows = self::parseCsv($file);
            }
        }

        if ($rows === []) {
            throw new RuntimeException('文件为空或无法解析（请确认是 xlsx/csv 且包含数据行）');
        }
        $meta = [];
        $students = self::rowsToStudents($rows, $meta);
        return [
            'count'      => count($students),
            'skipped'    => $meta['skipped'],
            'duplicates' => $meta['duplicates'],
            'headers'    => $meta['headers'],
            'students'  => $students,
        ];
    }

    public static function parseCsv(string $file): array
    {
        $raw = (string)@file_get_contents($file);
        $raw = preg_replace('/^\xEF\xBB\xBF/', '', $raw) ?? $raw;
        $rows = [];
        if (($fh = fopen('php://memory', 'r+')) === false) {
            return [];
        }
        fwrite($fh, $raw);
        rewind($fh);
        while (($line = fgetcsv($fh, 0, ',')) !== false) {
            if ($line === [null] || (count($line) === 1 && trim((string)$line[0]) === '')) {
                continue;
            }
            $rows[] = array_map(static fn($v) => self::conv((string)$v), $line);
        }
        fclose($fh);
        if ($rows === []) {
            // 可能是 tab 分隔
            foreach (preg_split('/\r?\n/', $raw) ?: [] as $line) {
                if (trim($line) === '') {
                    continue;
                }
                $rows[] = array_map([self::class, 'conv'], explode("\t", $line));
            }
        }
        return $rows;
    }

    private static function conv(string $v): string
    {
        $v = trim($v);
        $enc = mb_detect_encoding($v, ['UTF-8', 'GB18030', 'GBK', 'BIG5'], true);
        if ($enc !== false && strtoupper($enc) !== 'UTF-8') {
            $v = mb_convert_encoding($v, 'UTF-8', $enc);
        }
        return $v;
    }

    /**
     * 识别 学号/姓名/班级/组别/登陆账号 列
     * 登陆账号仅用于随机抽取页导出白名单，界面任何位置都不显示
     * @param array $rows
     * @param array $meta 输出：['skipped'=>int,'duplicates'=>int,'headers'=>array]
     * @return array<int,array{sid:string,name:string,cls:string,grp:string,login:string}>
     */
    public static function rowsToStudents(array $rows, array &$meta = []): array
    {
        $meta = ['skipped' => 0, 'duplicates' => 0, 'headers' => []];
        $sidCol   = -1;
        $nameCol  = -1;
        $clsCol   = -1;
        $grpCol   = -1;
        $loginCol = -1;
        $headerIdx = -1;

        // 找表头：同一行需同时命中「学号类」与「姓名类」关键词（容错全角空格/间隔）
        foreach ($rows as $i => $row) {
            $joined = implode('|', array_map('strval', $row));
            if (preg_match('/学\s*号|编\s*号|序\s*号|学\s*员\s*号|号\s*码|no\.?|id/i', $joined)
                && preg_match('/姓\s*名|名\s*字|学\s*生\s*姓\s*名|姓\s*氏|成\s*员/i', $joined)) {
                $headerIdx = $i;
                break;
            }
        }
        $start = 0;
        if ($headerIdx >= 0) {
            $start = $headerIdx + 1;
            foreach ($rows[$headerIdx] as $j => $cell) {
                $c = (string)$cell;
                if ($sidCol < 0 && preg_match('/学\s*号|编\s*号|序\s*号|学\s*员\s*号|号\s*码|^\s*no\.?\s*$|^\s*id\s*$/i', $c)) {
                    $sidCol = $j;
                }
                if ($nameCol < 0 && preg_match('/姓\s*名|名\s*字|学\s*生\s*姓\s*名|姓\s*氏/i', $c)) {
                    $nameCol = $j;
                }
                if ($clsCol < 0 && preg_match('/班\s*级|班级|class/i', $c)) {
                    $clsCol = $j;
                }
                if ($grpCol < 0 && preg_match('/组\s*别|分\s*组|组\s*次|^\s*组\s*$|group/i', $c)) {
                    $grpCol = $j;
                }
                if ($loginCol < 0 && preg_match('/登\s*[陆录]\s*账\s*号|登\s*[陆录]\s*账号|账\s*号|手\s*机\s*号|手\s*机/i', $c)) {
                    $loginCol = $j;
                }
            }
            $meta['headers'] = [
                'sid' => $sidCol, 'name' => $nameCol, 'cls' => $clsCol, 'grp' => $grpCol, 'login' => $loginCol,
            ];
        }
        if ($sidCol < 0) {
            $sidCol = 0;
        }
        if ($nameCol < 0) {
            $nameCol = $sidCol === 0 ? 1 : 0;
        }

        $out = [];
        $seen = [];
        for ($i = $start; $i < count($rows); $i++) {
            $row = $rows[$i];
            $sid  = trim((string)($row[$sidCol] ?? ''));
            $name = trim((string)($row[$nameCol] ?? ''));
            if ($sid === '' && $name === '') {
                $meta['skipped']++;
                continue;
            }
            // 学号/登陆账号去掉 Excel 数字尾巴（如 2024012001.0 / 13912687810.0）
            $sid   = preg_replace('/\.0+$/', '', $sid) ?? $sid;
            $cls   = $clsCol >= 0 ? trim((string)($row[$clsCol] ?? '')) : '';
            $grp   = $grpCol >= 0 ? trim((string)($row[$grpCol] ?? '')) : '';
            $login = $loginCol >= 0 ? trim((string)($row[$loginCol] ?? '')) : '';
            $login = preg_replace('/\.0+$/', '', $login) ?? $login;
            // 同一条学号多次出现：保留首次，其余记重复跳过（避免看板/统计被重复计入）
            if ($sid !== '' && isset($seen[$sid])) {
                $meta['duplicates']++;
                continue;
            }
            $seen[$sid] = true;
            $out[] = ['sid' => $sid, 'name' => $name, 'cls' => $cls, 'grp' => $grp, 'login' => $login];
        }
        return $out;
    }
}