Please address the comments from this code review:

## Overall Comments
- The change in `MessageConverser.arraysToString` from `if (MsgTypeEnum.text.equals(item.getType()))` to `if (!MsgTypeEnum.text.equals(item.getType()))` inverts the previous behavior and looks unintended, as it now treats non-text as the default branch; please verify and restore the original condition if this wasn’t deliberate.
- In `MessageConverser.arrayToPlainText`, consider using `MsgTypeEnum.text.toString()` (or a shared constant) instead of the hard-coded "text" literal for `getStringData` to keep the enum-based key usage consistent and reduce risk of breakage if the enum name changes.

## Individual Comments

### Comment 1
<location> `src/main/java/com/mikuac/shiro/common/utils/MessageConverser.java:21-24` </location>
<code_context>
     public static String arraysToString(List<ArrayMsg> array) {
         StringBuilder builder = new StringBuilder();
         for (ArrayMsg item : array) {
-            if (MsgTypeEnum.text.equals(item.getType())) {
+            if (!MsgTypeEnum.text.equals(item.getType())) {
                 builder.append(item.toCQCode());
</code_context>

<issue_to_address>
**suggestion:** Avoid hardcoded "text" literal for key and reuse the enum value for consistency.

This method mixes `item.getStringData("text")` with `item.getStringData(MsgTypeEnum.text.toString())` used elsewhere in the class. If the enum name or underlying key changes, this will behave differently. Use `MsgTypeEnum.text.toString()` here as well to keep behavior consistent and robust against refactors.

Suggested implementation:

```java
    public static String arraysToString(List<ArrayMsg> array) {
        StringBuilder builder = new StringBuilder();
        for (ArrayMsg item : array) {
            if (!MsgTypeEnum.text.equals(item.getType())) {
                builder.append(item.toCQCode());
            } else {
                builder.append(item.getStringData(MsgTypeEnum.text.toString()));
            }
        }
        return builder.toString();
    }

```

If there are any other occurrences of `item.getStringData("text")` elsewhere in `MessageConverser.java` (or related classes), they should also be updated to `item.getStringData(MsgTypeEnum.text.toString())` to keep the behavior consistent and robust against future enum refactors.
</issue_to_address>

### Comment 2
<location> `src/test/java/com/mikuac/shiro/common/utils/MessageConverserTest.java:183-188` </location>
<code_context>
         assertEquals(originalResult, optimizedResult);
     }

+    @Test
+    void testPlainMsgTest() {
+        val msg = "[CQ:at,qq=1122334455]测试消息1[CQ:face,id=1]测试消息2[CQ:video,file=https://test.com/1.mp4][CQ:image,file=test1.image,url=https://test.com/1.jpg]\n[CQ:image,file=test2.image,url=https://test.com/2.jpg]";
+        val expected = "测试消息1测试消息2\n";
+        val plainText = MessageConverser.stringToPlainText(msg);
+        assertEquals(expected, plainText);
+    }
+
</code_context>

<issue_to_address>
**suggestion (testing):** Add more edge-case coverage for `stringToPlainText` and `arrayToPlainText`.

The new `testPlainMsgTest` covers one complex scenario, but the utilities would benefit from a few more targeted tests:

- A case with **only CQ codes and no text** (plain text should be an empty string).
- A case with **only text** and no CQ codes to confirm it’s a no-op.
- A case with **leading/trailing CQ codes around text** with different separators (spaces/newlines) to confirm formatting.
- A direct unit test for `arrayToPlainText(List<ArrayMsg>)` using manually constructed `ArrayMsg` instances, so issues aren’t hidden by `stringToArray`.

These will help validate `plainText` across the main message patterns you expect to handle.

Suggested implementation:

```java
    @Test
    void testPlainMsgTest() {
        val msg = "[CQ:at,qq=1122334455]测试消息1[CQ:face,id=1]测试消息2[CQ:video,file=https://test.com/1.mp4][CQ:image,file=test1.image,url=https://test.com/1.jpg]\n[CQ:image,file=test2.image,url=https://test.com/2.jpg]";
        val expected = "测试消息1测试消息2\n";
        val plainText = MessageConverser.stringToPlainText(msg);
        assertEquals(expected, plainText);
    }

    /**
     * 仅包含 CQ 码、没有任何纯文本，结果应为空字符串
     */
    @Test
    void testStringToPlainTextOnlyCQCodes() {
        val msg = "[CQ:at,qq=1122334455][CQ:face,id=1][CQ:image,file=test.image,url=https://test.com/1.jpg]";
        val expected = "";
        val plainText = MessageConverser.stringToPlainText(msg);
        assertEquals(expected, plainText);
    }

    /**
     * 仅包含纯文本，stringToPlainText 应为 no-op
     */
    @Test
    void testStringToPlainTextOnlyText() {
        val msg = "这是一个纯文本消息 without any CQ codes.\n第二行文本。";
        val expected = msg;
        val plainText = MessageConverser.stringToPlainText(msg);
        assertEquals(expected, plainText);
    }

    /**
     * 文本前后存在 CQ 码与不同分隔符，确保格式（空格/换行）被保留
     */
    @Test
    void testStringToPlainTextLeadingAndTrailingCQCodes() {
        val msg = "[CQ:at,qq=1122334455] 前置文本 [CQ:face,id=1]\n后置文本 [CQ:image,file=test.image,url=https://test.com/1.jpg]";
        // 预期只去掉 CQ 码，本身的空格和换行应当保留
        val expected = " 前置文本 \n后置文本 ";
        val plainText = MessageConverser.stringToPlainText(msg);
        assertEquals(expected, plainText);
    }

    /**
     * 直接使用手动构造的 ArrayMsg 列表测试 arrayToPlainText，
     * 避免 stringToArray 中间步骤掩盖潜在问题
     */
    @Test
    void testArrayToPlainTextDirectly() {
        val arrayMsgs = new java.util.ArrayList<ArrayMsg>();

        // 非文本类型（如 at），不应该出现在 plain text 中
        {
            val atMsg = new ArrayMsg();
            atMsg.setType("at");
            atMsg.setText(null);
            arrayMsgs.add(atMsg);
        }

        // 第一段纯文本
        {
            val textMsg1 = new ArrayMsg();
            textMsg1.setType("text");
            textMsg1.setText("测试消息1");
            arrayMsgs.add(textMsg1);
        }

        // 表情类 CQ 码，假设不参与 plain text 输出
        {
            val faceMsg = new ArrayMsg();
            faceMsg.setType("face");
            faceMsg.setText(null);
            arrayMsgs.add(faceMsg);
        }

        // 第二段纯文本（带换行）
        {
            val textMsg2 = new ArrayMsg();
            textMsg2.setType("text");
            textMsg2.setText("测试消息2\n");
            arrayMsgs.add(textMsg2);
        }

        // 图片/视频等富媒体消息
        {
            val imageMsg = new ArrayMsg();
            imageMsg.setType("image");
            imageMsg.setText(null);
            arrayMsgs.add(imageMsg);
        }

        val expected = "测试消息1测试消息2\n";
        val plainText = MessageConverser.arrayToPlainText(arrayMsgs);
        assertEquals(expected, plainText);
    }

    @Test
    void rawToArrayMsg2Test() {

```

1. 确认 `ArrayMsg` 类的包名和 API：
   - 如果当前测试文件尚未导入 `ArrayMsg`，需要在文件顶部添加正确的 `import`，如 `import com.mikuac.shiro.dto.event.message.ArrayMsg;`（请根据实际项目结构调整）。
   - 如果 `ArrayMsg` 不是通过无参构造 + `setType`/`setText` 方式使用，而是有构造函数或 builder，请将 `new ArrayMsg(); setType(...); setText(...);` 改成项目中实际使用的构建方式。
2. 如果 `arrayToPlainText` 的语义与这里的假设不同（例如会保留某些非 text 类型的内容，或对换行有特殊处理），请同步调整各测试中的 `expected` 字符串，以匹配实际行为。
</issue_to_address>