math_cms/app/Services/SourceFileParserService.php

<?php

namespace App\Services;

use App\Models\MarkdownImport;
use App\Models\SourceFile;
use Illuminate\Support\Arr;
use Illuminate\Support\Facades\DB;
use Illuminate\Support\Str;

class SourceFileParserService
{
    public function storeFromMarkdown(
        string $filename,
        string $rawMarkdown,
        ?MarkdownImport $import = null,
        array $fileMetadata = [],
        ?string $storagePath = null
    ): SourceFile {
        $normalized = $this->normalizeFilename($filename);
        $extension = pathinfo($filename, PATHINFO_EXTENSION) ?: null;
        $extracted = $this->extractMetadataFromFilename($filename);
        $rawHash = sha1($rawMarkdown);

        // 去重：同名且内容一致时直接复用，避免重复 source_file
        $existing = SourceFile::query()
            ->where('normalized_filename', $normalized)
            ->orWhere('original_filename', $filename)
            ->get()
            ->first(function (SourceFile $file) use ($rawHash) {
                $oldHash = sha1((string) $file->raw_markdown);
                return $oldHash === $rawHash;
            });

        if ($existing) {
            // 也把 hash 写入已存在记录的元数据，方便后续识别
            $meta = $existing->extracted_metadata ?? [];
            if (empty($meta['raw_hash'])) {
                $meta['raw_hash'] = $rawHash;
                $existing->update(['extracted_metadata' => $meta]);
            }

            if ($import) {
                $import->update([
                    'source_name' => $existing->normalized_filename,
                    'file_name' => $existing->original_filename,
                ]);
            }

            return $existing;
        }

        return DB::transaction(function () use ($filename, $normalized, $extension, $storagePath, $rawMarkdown, $fileMetadata, $extracted, $import) {
            $sourceFile = SourceFile::create([
                'uuid' => (string) Str::uuid(),
                'original_filename' => $filename,
                'normalized_filename' => $normalized,
                'extension' => $extension,
                'storage_path' => $storagePath,
                'raw_markdown' => $rawMarkdown,
                'file_metadata' => $fileMetadata,
                'extracted_metadata' => array_merge($extracted, [
                    'raw_hash' => sha1($rawMarkdown),
                ]),
            ]);

            if ($import) {
                $import->update([
                    'source_name' => $normalized,
                    'file_name' => $filename,
                ]);
            }

            return $sourceFile;
        });
    }

    /**
     * 从文件名提取教材信息（版别、年级、学期、章节）。
     */
    public function extractMetadataFromFilename(string $filename): array
    {
        $info = [];
        $basename = str_replace([' ', '　'], '', pathinfo($filename, PATHINFO_FILENAME));

        // 统一下划线为分隔符，便于模式匹配
        $normalized = str_replace(['-', '（', '）', '(', ')'], '_', $basename);

        $editionPatterns = [
            '人教' => 'PEP',
            '苏教' => 'SJ',
            '浙教' => 'ZJ',
            '北师' => 'BS',
            '沪教' => 'HJ',
            '北师大' => 'BS',
            '北师大版' => 'BS',
        ];

        foreach ($editionPatterns as $key => $code) {
            if (Str::contains($normalized, $key)) {
                $info['edition'] = $code;
                break;
            }
        }

        // 年级：支持中文数字与阿拉伯数字
        if (preg_match('/高[一二三]|高[123]/u', $normalized, $m)) {
            $info['grade'] = $m[0];
        } elseif (preg_match('/初[一二三]|初[123]|初中/u', $normalized, $m)) {
            $info['grade'] = $m[0];
        } elseif (preg_match('/([1-9])[_年]?\s*年级?/u', $normalized, $m)) {
            $num = (int) ($m[1] ?? 0);
            if ($num > 0) {
                $info['grade'] = $num . '年级';
            }
        } elseif (preg_match('/[一二三四五六七八九]年级/u', $normalized, $m)) {
            $info['grade'] = $m[0];
        }

        // 学期/册次：上=1, 下=2，或文件名带 _1/_2；全册=0（便于前端/枚举）
        if (preg_match('/上册|下册|第[12]学期|第[12]册/u', $normalized, $m)) {
            $info['term'] = $m[0];
        } elseif (preg_match('/[_-](1|2)(?:\\.md)?$/', $normalized, $m)) {
            $info['term'] = $m[1] === '1' ? '上册' : '下册';
        } elseif (Str::contains($normalized, ['全册', '全集'])) {
            $info['term'] = '0'; // 全册
        }

        if (preg_match('/第[一二三四五六七八九十0-9]+章/u', $normalized, $m)) {
            $info['chapter'] = $m[0];
        }

        if (preg_match('/20[0-9]{2}/', $basename, $m)) {
            $info['year'] = $m[0];
        }

        if (preg_match('/培优|专项|期中|期末|模拟|基础卷|提升卷|练习卷/u', $normalized, $m)) {
            $info['source_type'] = $m[0];
        }

        if (Str::contains($normalized, '数学')) {
            $info['subject'] = '数学';
        }

        return $info;
    }

    protected function normalizeFilename(string $filename): string
    {
        $basename = pathinfo($filename, PATHINFO_FILENAME);
        $normalized = Str::of($basename)
            ->replace(['（', '）', '(', ')', ' '], ['_', '_', '_', '_', '_'])
            ->slug('_')
            ->toString();

        return (string) $normalized;
    }
}