Outlook EML文件附件解析问题：javax.mail获取文件名返回Null

我之前也碰到过一模一样的情况——用javax.mail解析EML附件时，getFileName()经常返回null，尤其是遇到一些老邮件或者非标准编码的邮件头时。下面是我整理的解决方案，以及能递归解析EML的通用库推荐：

一、手动解析附件文件名（解决Null问题） #

getFileName()失效的核心原因是邮件的Content-Disposition或Content-Type头里的文件名参数可能用了RFC 2047编码，或者格式不符合标准，导致JavaMail的默认解析逻辑没识别到。我们可以手动解析这些头字段，再处理编码：

完整的文件名获取代码 #

import javax.mail.Part;
import javax.mail.MessagingException;
import javax.mail.internet.MimeUtility;
import java.io.UnsupportedEncodingException;
import java.util.HashMap;
import java.util.Map;

public class AttachmentUtils {
    public static String getAttachmentFileName(Part part) throws MessagingException, UnsupportedEncodingException {
        String fileName = part.getFileName();
        
        // 第一步：尝试默认方法获取，失败则手动解析Content-Disposition头
        if (fileName == null) {
            String contentDisposition = part.getHeader("Content-Disposition", null);
            if (contentDisposition != null) {
                int startIndex = contentDisposition.indexOf("filename=");
                if (startIndex != -1) {
                    String fileNamePart = contentDisposition.substring(startIndex + 9);
                    // 处理带引号的文件名
                    if (fileNamePart.startsWith("\"")) {
                        fileNamePart = fileNamePart.substring(1, fileNamePart.lastIndexOf("\""));
                    } else {
                        // 处理不带引号的文件名，截断到下一个分号
                        int endIndex = fileNamePart.indexOf(";");
                        if (endIndex != -1) {
                            fileNamePart = fileNamePart.substring(0, endIndex);
                        }
                    }
                    // 解码RFC 2047编码的文件名（比如中文、特殊字符）
                    fileName = MimeUtility.decodeText(fileNamePart.trim());
                }
            }
        }
        
        // 第二步：如果还是null，尝试从Content-Type的name参数提取
        if (fileName == null) {
            String contentType = part.getContentType();
            int nameIndex = contentType.indexOf("name=");
            if (nameIndex != -1) {
                String namePart = contentType.substring(nameIndex + 5);
                if (namePart.startsWith("\"")) {
                    namePart = namePart.substring(1, namePart.lastIndexOf("\""));
                } else {
                    int endIndex = namePart.indexOf(";");
                    if (endIndex != -1) {
                        namePart = namePart.substring(0, endIndex);
                    }
                }
                fileName = MimeUtility.decodeText(namePart.trim());
            }
        }
        
        // 第三步：如果文件名没有扩展名，根据Content-Type推断
        if (fileName != null && !fileName.contains(".")) {
            String contentType = part.getContentType();
            Map<String, String> extMapping = new HashMap<>();
            extMapping.put("application/pdf", ".pdf");
            extMapping.put("image/jpeg", ".jpg");
            extMapping.put("image/png", ".png");
            extMapping.put("application/msword", ".doc");
            extMapping.put("application/vnd.openxmlformats-officedocument.wordprocessingml.document", ".docx");
            extMapping.put("text/plain", ".txt");
            // 可以根据需求补充更多类型映射
            
            String extension = extMapping.get(contentType);
            fileName += extension != null ? extension : ".bin";
        }
        
        return fileName;
    }
}

你可以在获取附件时调用这个方法，代替直接用part.getFileName()，基本能解决大部分Null的情况。

二、通用EML递归解析库推荐 #

如果需要处理复杂的EML结构（比如嵌套的Multipart、混合格式正文等），自己写递归逻辑容易踩坑，推荐用成熟的开源库：

1. Apache James Mime4j #

这是Apache专门处理MIME消息的库，对复杂MIME结构的解析、编码处理都比JavaMail更可靠，支持递归解析嵌套的Multipart：

依赖（Maven）

<dependency>
    <groupId>org.apache.james</groupId>
    <artifactId>apache-mime4j-core</artifactId>
    <version>0.8.10</version>
</dependency>
<dependency>
    <groupId>org.apache.james</groupId>
    <artifactId>apache-mime4j-dom</artifactId>
    <version>0.8.10</version>
</dependency>

递归解析示例

import org.apache.james.mime4j.dom.Message;
import org.apache.james.mime4j.dom.Multipart;
import org.apache.james.mime4j.dom.BodyPart;
import org.apache.james.mime4j.dom.TextBody;
import org.apache.james.mime4j.message.DefaultMessageBuilder;
import org.apache.james.mime4j.stream.MimeConfig;

import java.io.FileInputStream;
import java.io.IOException;
import java.io.FileOutputStream;

public class EmlParser {
    public static void parseEmlFile(String emlFilePath) throws IOException {
        MimeConfig config = MimeConfig.DEFAULT;
        DefaultMessageBuilder builder = new DefaultMessageBuilder();
        
        try (FileInputStream fis = new FileInputStream(emlFilePath)) {
            Message message = builder.parseMessage(fis);
            // 提取正文
            extractMessageBody(message);
            // 提取附件
            extractMessageAttachments(message, "save/attachments/");
        }
    }

    private static void extractMessageBody(Message message) throws IOException {
        Object body = message.getBody();
        if (body instanceof TextBody) {
            TextBody textBody = (TextBody) body;
            String content = textBody.getReader().readAllLines().stream()
                    .reduce((line1, line2) -> line1 + "\n" + line2)
                    .orElse("");
            System.out.println("邮件正文：\n" + content);
        } else if (body instanceof Multipart) {
            Multipart multipart = (Multipart) body;
            for (int i = 0; i < multipart.getCount(); i++) {
                extractMessageBody(multipart.getBodyPart(i));
            }
        }
    }

    private static void extractMessageAttachments(Message message, String saveDir) throws IOException {
        Object body = message.getBody();
        if (body instanceof Multipart) {
            Multipart multipart = (Multipart) body;
            for (int i = 0; i < multipart.getCount(); i++) {
                BodyPart part = multipart.getBodyPart(i);
                String disposition = part.getDisposition();
                
                if ((disposition != null && (disposition.equalsIgnoreCase("attachment") || disposition.equalsIgnoreCase("inline"))) 
                    || part.getFilename() != null) {
                    
                    String fileName = part.getFilename();
                    // 兜底处理文件名（虽然Mime4j已经处理得很好，但以防万一）
                    if (fileName == null) {
                        // 可以复用上面AttachmentUtils的逻辑
                    }
                    
                    // 保存附件到指定目录
                    try (FileOutputStream fos = new FileOutputStream(saveDir + fileName)) {
                        part.writeTo(fos);
                    }
                    System.out.println("已保存附件：" + fileName);
                } else {
                    // 递归处理嵌套的Multipart
                    extractMessageAttachments(part, saveDir);
                }
            }
        }
    }
}

2. 基于JavaMail的递归解析 #

如果你不想引入新依赖，也可以用JavaMail自己实现递归解析，封装上面的文件名获取逻辑即可：

import javax.mail.MessagingException;
import javax.mail.Part;
import javax.mail.internet.MimeMessage;
import javax.mail.internet.MimeMultipart;
import java.io.IOException;
import java.io.FileOutputStream;

public class JavaMailEmlParser {
    public static void parseEml(MimeMessage message) throws MessagingException, IOException {
        processPart(message);
    }

    private static void processPart(Part part) throws MessagingException, IOException {
        if (part.isMimeType("multipart/*")) {
            MimeMultipart multipart = (MimeMultipart) part.getContent();
            for (int i = 0; i < multipart.getCount(); i++) {
                processPart(multipart.getBodyPart(i));
            }
        } else {
            String disposition = part.getDisposition();
            if ((disposition != null && (disposition.equalsIgnoreCase(Part.ATTACHMENT) || disposition.equalsIgnoreCase(Part.INLINE)))
                || part.getFileName() != null) {
                
                String fileName = AttachmentUtils.getAttachmentFileName(part);
                try (FileOutputStream fos = new FileOutputStream("save/attachments/" + fileName)) {
                    part.writeTo(fos);
                }
            } else {
                // 处理正文
                String content = part.getContent().toString();
                System.out.println("正文内容：\n" + content);
            }
        }
    }
}

内容的提问来源于stack exchange，提问作者RRM