# Diagnosis Image URL Prefix Fix (诊断图片URL前缀修复)

## Issue
Image URLs in the diagnosis detail (tongue_images and report_files) were stored as relative paths without the domain prefix, causing issues when accessing them from external sources or different domains.

## Solution
Added logic to automatically prepend the current domain to image URLs that don't already have an `http://` or `https://` prefix.

## Changes Made

### File: `server/app/adminapi/logic/tcm/DiagnosisLogic.php`

#### 1. Updated tongue_images Processing
Added domain prefix logic after parsing the images array:

```php
if (!empty($diagnosis['tongue_images'])) {
    // 先尝试 JSON 解析（兼容旧数据）
    $files = json_decode($diagnosis['tongue_images'], true);
    if (is_array($files)) {
        $diagnosis['tongue_images'] = $files;
    } else {
        // JSON 解析失败则按逗号分割
        $diagnosis['tongue_images'] = array_filter(explode(',', $diagnosis['tongue_images'])) ?: [];
    }
    
    // 为没有 http 前缀的图片添加域名
    $domain = request()->domain();
    $diagnosis['tongue_images'] = array_map(function($url) use ($domain) {
        if (empty($url)) {
            return $url;
        }
        // 如果已经包含 http:// 或 https://，则跳过
        if (stripos($url, 'http://') === 0 || stripos($url, 'https://') === 0) {
            return $url;
        }
        // 添加域名前缀
        return $domain . $url;
    }, $diagnosis['tongue_images']);
} else {
    $diagnosis['tongue_images'] = [];
}
```

#### 2. Updated report_files Processing
Applied the same logic to report files:

```php
if (!empty($diagnosis['report_files'])) {
    // 先尝试 JSON 解析（兼容旧数据）
    $files = json_decode($diagnosis['report_files'], true);
    if (is_array($files)) {
        $diagnosis['report_files'] = $files;
    } else {
        // JSON 解析失败则按逗号分割
        $diagnosis['report_files'] = array_filter(explode(',', $diagnosis['report_files'])) ?: [];
    }
    
    // 为没有 http 前缀的文件添加域名
    $domain = request()->domain();
    $diagnosis['report_files'] = array_map(function($url) use ($domain) {
        if (empty($url)) {
            return $url;
        }
        // 如果已经包含 http:// 或 https://，则跳过
        if (stripos($url, 'http://') === 0 || stripos($url, 'https://') === 0) {
            return $url;
        }
        // 添加域名前缀
        return $domain . $url;
    }, $diagnosis['report_files']);
} else {
    $diagnosis['report_files'] = [];
}
```

## Logic Flow

### URL Processing Logic:
1. Parse the images/files array (JSON or comma-separated)
2. Get current domain using `request()->domain()`
3. For each URL in the array:
   - Check if URL is empty → skip
   - Check if URL starts with `http://` or `https://` → skip (already has protocol)
   - Otherwise → prepend domain to the URL

### Examples:

#### Relative Path (Needs Domain):
```php
Input:  "/uploads/images/tongue/2024/01/image.jpg"
Domain: "https://admin.zhenyangtang.com.cn"
Output: "https://admin.zhenyangtang.com.cn/uploads/images/tongue/2024/01/image.jpg"
```

#### Absolute URL (Skip):
```php
Input:  "https://cdn.example.com/images/tongue.jpg"
Output: "https://cdn.example.com/images/tongue.jpg"
```

#### HTTP URL (Skip):
```php
Input:  "http://example.com/image.jpg"
Output: "http://example.com/image.jpg"
```

#### Empty URL (Skip):
```php
Input:  ""
Output: ""
```

## Benefits

### 1. Cross-Domain Compatibility
Images can now be accessed from different domains or external applications without path issues.

### 2. API Response Consistency
API responses always return complete, accessible URLs regardless of how they were stored.

### 3. Backward Compatibility
- Existing relative paths are automatically converted
- Existing absolute URLs are preserved
- No database migration required

### 4. Flexible Storage
- Database can store either relative or absolute paths
- System handles both formats transparently
- Future-proof for CDN integration

## Use Cases

### 1. Mobile App Access
Mobile apps can directly use the returned URLs without needing to construct them.

### 2. Third-Party Integration
External systems can access images using the complete URLs from API responses.

### 3. CDN Migration
When migrating to CDN, images with absolute CDN URLs will work alongside local relative paths.

### 4. Multi-Domain Setup
System works correctly across different domains (dev, staging, production).

## Testing Steps

### 1. Test Relative Path URLs
1. Create a diagnosis with tongue images stored as relative paths:
   ```
   /uploads/images/tongue/2024/01/image.jpg
   ```
2. Fetch diagnosis detail via API
3. Verify response contains:
   ```json
   {
     "tongue_images": [
       "https://admin.zhenyangtang.com.cn/uploads/images/tongue/2024/01/image.jpg"
     ]
   }
   ```

### 2. Test Absolute URLs
1. Create a diagnosis with absolute URL:
   ```
   https://cdn.example.com/images/tongue.jpg
   ```
2. Fetch diagnosis detail
3. Verify URL is unchanged:
   ```json
   {
     "tongue_images": [
       "https://cdn.example.com/images/tongue.jpg"
     ]
   }
   ```

### 3. Test Mixed URLs
1. Create a diagnosis with both relative and absolute URLs:
   ```
   [
     "/uploads/local.jpg",
     "https://cdn.example.com/remote.jpg"
   ]
   ```
2. Fetch diagnosis detail
3. Verify correct processing:
   ```json
   {
     "tongue_images": [
       "https://admin.zhenyangtang.com.cn/uploads/local.jpg",
       "https://cdn.example.com/remote.jpg"
     ]
   }
   ```

### 4. Test Empty/Null Values
1. Create a diagnosis with empty images
2. Fetch diagnosis detail
3. Verify empty array is returned:
   ```json
   {
     "tongue_images": []
   }
   ```

### 5. Test Report Files
Repeat all tests above for `report_files` field.

## Technical Details

### Domain Detection
Uses `request()->domain()` which returns the current request domain including protocol:
- Development: `http://localhost:8000`
- Production: `https://admin.zhenyangtang.com.cn`

### Case-Insensitive Protocol Check
Uses `stripos()` for case-insensitive checking:
- Matches: `http://`, `HTTP://`, `Http://`
- Matches: `https://`, `HTTPS://`, `Https://`

### Array Processing
Uses `array_map()` for efficient processing of all URLs in the array.

## Related Files

### Backend:
- `server/app/adminapi/logic/tcm/DiagnosisLogic.php`
  - Updated `tongue_images` processing
  - Updated `report_files` processing

### Database:
- `tcm_diagnosis` table
  - `tongue_images` column (stores JSON or comma-separated paths)
  - `report_files` column (stores JSON or comma-separated paths)

## Notes

### Storage Format
The database continues to store paths as-is (relative or absolute). The conversion happens only when reading data for API responses.

### Performance
The `array_map()` operation is efficient and adds minimal overhead. For typical diagnosis records with 1-5 images, the performance impact is negligible.

### Future Enhancements
If needed, this logic could be extracted into a helper function for reuse across other file/image fields in the system.

## Summary

✅ Automatically adds domain prefix to relative image URLs  
✅ Preserves absolute URLs (http/https)  
✅ Handles both tongue_images and report_files  
✅ Backward compatible with existing data  
✅ Case-insensitive protocol detection  
✅ Handles empty/null values gracefully  
✅ No database migration required  
✅ Works across different domains  

Image URLs in diagnosis details are now always complete and accessible, improving API usability and cross-domain compatibility!