服务端如何将一个大视频文件做切分，分段响应给客户端，让浏览器可以渐进式地播放。
Spring Boot实现HTTP分片下载断点续传，从而实现H5页面的大视频播放问题，实现渐进式播放，每次只播放需要播放的内容就可以了，不需要加载整个文件到内存中。文章来源：https://www.toymoban.com/news/detail-626958.html
文件的断点续传、文件多线程并发下载（迅雷就是这么玩的）等。文章来源地址https://www.toymoban.com/news/detail-626958.html
<dependencyManagement>
<dependencies>
<dependency>
<groupId>cn.hutool</groupId>
<artifactId>hutool-bom</artifactId>
<version>5.8.18</version>
<type>pom</type>
<scope>import</scope>
</dependency>
</dependencies>
</dependencyManagement>
<dependencies>
<dependency>
<groupId>cn.hutool</groupId>
<artifactId>hutool-core</artifactId>
</dependency>
<dependency>
<groupId>org.projectlombok</groupId>
<artifactId>lombok</artifactId>
<optional>true</optional>
</dependency>
</dependencies>
代码实现
ResourceController
package com.example.insurance.controller;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.util.List;
import com.example.insurance.common.ContentRange;
import com.example.insurance.common.MediaContentUtil;
import com.example.insurance.common.NioUtils;
import lombok.extern.slf4j.Slf4j;
import org.springframework.http.HttpHeaders;
import org.springframework.http.HttpRange;
import org.springframework.http.HttpStatus;
import org.springframework.util.CollectionUtils;
import org.springframework.util.StopWatch;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.PathVariable;
import org.springframework.web.bind.annotation.RequestHeader;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RestController;
/**
* 内容资源控制器
*/
@SuppressWarnings("unused")
@Slf4j
@RestController("resourceController")
@RequestMapping(path = "/resource")
public class ResourceController {
/**
* 获取文件内容
*
* @param fileName 内容文件名称
* @param response 响应对象
*/
@GetMapping("/media/{fileName}")
public void getMedia(@PathVariable String fileName, HttpServletRequest request, HttpServletResponse response,
@RequestHeader HttpHeaders headers) {
// printRequestInfo(fileName, request, headers);
String filePath = MediaContentUtil.filePath();
try {
this.download(fileName, filePath, request, response, headers);
} catch (Exception e) {
log.error("getMedia error, fileName={}", fileName, e);
}
}
/**
* 获取封面内容
*
* @param fileName 内容封面名称
* @param response 响应对象
*/
@GetMapping("/cover/{fileName}")
public void getCover(@PathVariable String fileName, HttpServletRequest request, HttpServletResponse response,
@RequestHeader HttpHeaders headers) {
// printRequestInfo(fileName, request, headers);
String filePath = MediaContentUtil.filePath();
try {
this.download(fileName, filePath, request, response, headers);
} catch (Exception e) {
log.error("getCover error, fileName={}", fileName, e);
}
}
// ======= internal =======
private static void printRequestInfo(String fileName, HttpServletRequest request, HttpHeaders headers) {
String requestUri = request.getRequestURI();
String queryString = request.getQueryString();
log.debug("file={}, url={}?{}", fileName, requestUri, queryString);
log.info("headers={}", headers);
}
/**
* 设置请求响应状态、头信息、内容类型与长度 等。
* <pre>
* <a href="https://www.rfc-editor.org/rfc/rfc7233">
* HTTP/1.1 Range Requests</a>
* 2. Range Units
* 4. Responses to a Range Request
*
* <a href="https://www.rfc-editor.org/rfc/rfc2616.html">
* HTTP/1.1</a>
* 10.2.7 206 Partial Content
* 14.5 Accept-Ranges
* 14.13 Content-Length
* 14.16 Content-Range
* 14.17 Content-Type
* 19.5.1 Content-Disposition
* 15.5 Content-Disposition Issues
*
* <a href="https://www.rfc-editor.org/rfc/rfc2183">
* Content-Disposition</a>
* 2. The Content-Disposition Header Field
* 2.1 The Inline Disposition Type
* 2.3 The Filename Parameter
* </pre>
*
* @param response 请求响应对象
* @param fileName 请求的文件名称
* @param contentType 内容类型
* @param contentRange 内容范围对象
*/
private static void setResponse(
HttpServletResponse response, String fileName, String contentType,
ContentRange contentRange) {
// http状态码要为206：表示获取部分内容
response.setStatus(HttpStatus.PARTIAL_CONTENT.value());
// 支持断点续传，获取部分字节内容
// Accept-Ranges：bytes，表示支持Range请求
response.setHeader(HttpHeaders.ACCEPT_RANGES, ContentRange.BYTES_STRING);
// inline表示浏览器直接使用，attachment表示下载，fileName表示下载的文件名
response.setHeader(HttpHeaders.CONTENT_DISPOSITION,
"inline;filename=" + MediaContentUtil.encode(fileName));
// Content-Range，格式为：[要下载的开始位置]-[结束位置]/[文件总大小]
// Content-Range: bytes 0-10/3103，格式为bytes 开始-结束/全部
response.setHeader(HttpHeaders.CONTENT_RANGE, contentRange.toContentRange());
response.setContentType(contentType);
// Content-Length: 11，本次内容的大小
response.setContentLengthLong(contentRange.applyAsContentLength());
}
/**
* <a href="https://www.jianshu.com/p/08db5ba3bc95">
* Spring Boot 处理 HTTP Headers</a>
*/
private void download(
String fileName, String path, HttpServletRequest request, HttpServletResponse response,
HttpHeaders headers)
throws IOException {
Path filePath = Paths.get(path + fileName);
if (!Files.exists(filePath)) {
log.warn("file not exist, filePath={}", filePath);
return;
}
long fileLength = Files.size(filePath);
// long fileLength2 = filePath.toFile().length() - 1;
// // fileLength=1184856, fileLength2=1184855
// log.info("fileLength={}, fileLength2={}", fileLength, fileLength2);
// 内容范围
ContentRange contentRange = applyAsContentRange(headers, fileLength, request);
// 要下载的长度
long contentLength = contentRange.applyAsContentLength();
log.debug("contentRange={}, contentLength={}", contentRange, contentLength);
// 文件类型
String contentType = request.getServletContext().getMimeType(fileName);
// mimeType=video/mp4, CONTENT_TYPE=null
log.debug("mimeType={}, CONTENT_TYPE={}", contentType, request.getContentType());
setResponse(response, fileName, contentType, contentRange);
// 耗时指标统计
StopWatch stopWatch = new StopWatch("downloadFile");
stopWatch.start(fileName);
try {
// case-1.参考网上他人的实现
// if (fileLength >= Integer.MAX_VALUE) {
// NioUtils.copy(filePath, response, contentRange);
// } else {
// NioUtils.copyByChannelAndBuffer(filePath, response, contentRange);
// }
// case-2.使用现成API
NioUtils.copyByBio(filePath, response, contentRange);
// NioUtils.copyByNio(filePath, response, contentRange);
// case-3.视频分段渐进式播放
// if (contentType.startsWith("video")) {
// NioUtils.copyForBufferSize(filePath, response, contentRange);
// } else {
// // 图片、PDF等文件
// NioUtils.copyByBio(filePath, response, contentRange);
// }
} finally {
stopWatch.stop();
log.info("download file, fileName={}, time={} ms", fileName, stopWatch.getTotalTimeMillis());
}
}
private static ContentRange applyAsContentRange(
HttpHeaders headers, long fileLength, HttpServletRequest request) {
/*
* 3.1. Range - HTTP/1.1 Range Requests
* https://www.rfc-editor.org/rfc/rfc7233#section-3.1
* Range: "bytes" "=" first-byte-pos "-" [ last-byte-pos ]
*
* For example:
* bytes=0-
* bytes=0-499
*/
// Range：告知服务端，客户端下载该文件想要从指定的位置开始下载
List<HttpRange> httpRanges = headers.getRange();
String range = request.getHeader(HttpHeaders.RANGE);
// httpRanges=[], range=null
// httpRanges=[448135688-], range=bytes=448135688-
log.debug("httpRanges={}, range={}", httpRanges, range);
// 开始下载位置
long firstBytePos;
// 结束下载位置
long lastBytePos;
if (CollectionUtils.isEmpty(httpRanges)) {
firstBytePos = 0;
lastBytePos = fileLength - 1;
} else {
HttpRange httpRange = httpRanges.get(0);
firstBytePos = httpRange.getRangeStart(fileLength);
lastBytePos = httpRange.getRangeEnd(fileLength);
}
return new ContentRange(firstBytePos, lastBytePos, fileLength);
}
}
NioUtils
package com.example.insurance.common;
import javax.servlet.http.HttpServletResponse;
import java.io.BufferedOutputStream;
import java.io.InputStream;
import java.io.OutputStream;
import java.io.RandomAccessFile;
import java.nio.MappedByteBuffer;
import java.nio.channels.Channels;
import java.nio.channels.FileChannel;
import java.nio.file.Path;
import java.nio.file.StandardOpenOption;
import cn.hutool.core.io.IORuntimeException;
import cn.hutool.core.io.IoUtil;
import cn.hutool.core.io.NioUtil;
import cn.hutool.core.io.StreamProgress;
import cn.hutool.core.io.unit.DataSize;
import lombok.extern.slf4j.Slf4j;
import org.apache.catalina.connector.ClientAbortException;
/**
* NIO相关工具封装，主要针对Channel读写、拷贝等封装
*/
@Slf4j
public final class NioUtils {
/**
* 缓冲区大小 16KB
*
* @see NioUtil#DEFAULT_BUFFER_SIZE
* @see NioUtil#DEFAULT_LARGE_BUFFER_SIZE
*/
// private static final int BUFFER_SIZE = NioUtil.DEFAULT_MIDDLE_BUFFER_SIZE;
private static final int BUFFER_SIZE = (int) DataSize.ofKilobytes(16L).toBytes();
/**
* <pre>
* <a href="https://blog.csdn.net/qq_32099833/article/details/109703883">
* Java后端实现视频分段渐进式播放</a>
* 服务端如何将一个大的视频文件做切分，分段响应给客户端，让浏览器可以渐进式地播放。
* 文件的断点续传、文件多线程并发下载（迅雷就是这么玩的）等。
*
* <a href="https://blog.csdn.net/qq_32099833/article/details/109630499">
* 大文件分片上传前后端实现</a>
* </pre>
*/
public static void copyForBufferSize(
Path filePath, HttpServletResponse response, ContentRange contentRange) {
String fileName = filePath.getFileName().toString();
RandomAccessFile randomAccessFile = null;
OutputStream outputStream = null;
try {
// 随机读文件
randomAccessFile = new RandomAccessFile(filePath.toFile(), "r");
// 移动访问指针到指定位置
randomAccessFile.seek(contentRange.getStart());
// 注意：缓冲区大小 2MB，视频加载正常；1MB时有部分视频加载失败
int bufferSize = BUFFER_SIZE;
//获取响应的输出流
outputStream = new BufferedOutputStream(response.getOutputStream(), bufferSize);
// 每次请求只返回1MB的视频流
byte[] buffer = new byte[bufferSize];
int len = randomAccessFile.read(buffer);
//设置此次相应返回的数据长度
response.setContentLength(len);
// 将这1MB的视频流响应给客户端
outputStream.write(buffer, 0, len);
log.info("file download complete, fileName={}, contentRange={}",
fileName, contentRange.toContentRange());
} catch (ClientAbortException | IORuntimeException e) {
// 捕获此异常表示用户停止下载
log.warn("client stop file download, fileName={}", fileName);
} catch (Exception e) {
log.error("file download error, fileName={}", fileName, e);
} finally {
IoUtil.close(outputStream);
IoUtil.close(randomAccessFile);
}
}
/**
* 拷贝流，拷贝后关闭流。
*
* @param filePath 源文件路径
* @param response 请求响应
* @param contentRange 内容范围
*/
public static void copyByBio(
Path filePath, HttpServletResponse response, ContentRange contentRange) {
String fileName = filePath.getFileName().toString();
InputStream inputStream = null;
OutputStream outputStream = null;
try {
RandomAccessFile randomAccessFile = new RandomAccessFile(filePath.toFile(), "r");
randomAccessFile.seek(contentRange.getStart());
inputStream = Channels.newInputStream(randomAccessFile.getChannel());
outputStream = new BufferedOutputStream(response.getOutputStream(), BUFFER_SIZE);
StreamProgress streamProgress = new StreamProgressImpl(fileName);
long transmitted = IoUtil.copy(inputStream, outputStream, BUFFER_SIZE, streamProgress);
log.info("file download complete, fileName={}, transmitted={}", fileName, transmitted);
} catch (ClientAbortException | IORuntimeException e) {
// 捕获此异常表示用户停止下载
log.warn("client stop file download, fileName={}", fileName);
} catch (Exception e) {
log.error("file download error, fileName={}", fileName, e);
} finally {
IoUtil.close(outputStream);
IoUtil.close(inputStream);
}
}
/**
* 拷贝流，拷贝后关闭流。
* <pre>
* <a href="https://www.cnblogs.com/czwbig/p/10035631.html">
* Java NIO 学习笔记（一）----概述，Channel/Buffer</a>
* </pre>
*
* @param filePath 源文件路径
* @param response 请求响应
* @param contentRange 内容范围
*/
public static void copyByNio(
Path filePath, HttpServletResponse response, ContentRange contentRange) {
String fileName = filePath.getFileName().toString();
InputStream inputStream = null;
OutputStream outputStream = null;
try {
RandomAccessFile randomAccessFile = new RandomAccessFile(filePath.toFile(), "r");
randomAccessFile.seek(contentRange.getStart());
inputStream = Channels.newInputStream(randomAccessFile.getChannel());
outputStream = new BufferedOutputStream(response.getOutputStream(), BUFFER_SIZE);
StreamProgress streamProgress = new StreamProgressImpl(fileName);
long transmitted = NioUtil.copyByNIO(inputStream, outputStream,
BUFFER_SIZE, streamProgress);
log.info("file download complete, fileName={}, transmitted={}", fileName, transmitted);
} catch (ClientAbortException | IORuntimeException e) {
// 捕获此异常表示用户停止下载
log.warn("client stop file download, fileName={}", fileName);
} catch (Exception e) {
log.error("file download error, fileName={}", fileName, e);
} finally {
IoUtil.close(outputStream);
IoUtil.close(inputStream);
}
}
/**
* <pre>
* <a href="https://blog.csdn.net/lovequanquqn/article/details/104562945">
* SpringBoot Java实现Http方式分片下载断点续传+实现H5大视频渐进式播放</a>
* SpringBoot 实现Http分片下载断点续传，从而实现H5页面的大视频播放问题，实现渐进式播放，每次只播放需要播放的内容就可以了，不需要加载整个文件到内存中。
* 二、Http分片下载断点续传实现
* 四、缓存文件定时删除任务
* </pre>
*/
public static void copy(Path filePath, HttpServletResponse response, ContentRange contentRange) {
String fileName = filePath.getFileName().toString();
// 要下载的长度
long contentLength = contentRange.applyAsContentLength();
BufferedOutputStream outputStream = null;
RandomAccessFile randomAccessFile = null;
// 已传送数据大小
long transmitted = 0;
try {
randomAccessFile = new RandomAccessFile(filePath.toFile(), "r");
randomAccessFile.seek(contentRange.getStart());
outputStream = new BufferedOutputStream(response.getOutputStream(), BUFFER_SIZE);
// 把数据读取到缓冲区中
byte[] buffer = new byte[BUFFER_SIZE];
int len = BUFFER_SIZE;
//warning：判断是否到了最后不足4096（buffer的length）个byte这个逻辑（(transmitted + len) <= contentLength）要放前面
//不然会会先读取randomAccessFile，造成后面读取位置出错;
while ((transmitted + len) <= contentLength && (len = randomAccessFile.read(buffer)) != -1) {
outputStream.write(buffer, 0, len);
transmitted += len;
log.info("fileName={}, transmitted={}", fileName, transmitted);
}
//处理不足buffer.length部分
if (transmitted < contentLength) {
len = randomAccessFile.read(buffer, 0, (int) (contentLength - transmitted));
outputStream.write(buffer, 0, len);
transmitted += len;
log.info("fileName={}, transmitted={}", fileName, transmitted);
}
log.info("file download complete, fileName={}, transmitted={}", fileName, transmitted);
} catch (ClientAbortException e) {
// 捕获此异常表示用户停止下载
log.warn("client stop file download, fileName={}, transmitted={}", fileName, transmitted);
} catch (Exception e) {
log.error("file download error, fileName={}, transmitted={}", fileName, transmitted, e);
} finally {
IoUtil.close(outputStream);
IoUtil.close(randomAccessFile);
}
}
/**
* 通过数据传输通道和缓冲区读取文件数据。
* <pre>
* 当文件长度超过{@link Integer#MAX_VALUE}时，
* 使用{@link FileChannel#map(FileChannel.MapMode, long, long)}报如下异常。
* java.lang.IllegalArgumentException: Size exceeds Integer.MAX_VALUE
* at sun.nio.ch.FileChannelImpl.map(FileChannelImpl.java:863)
* at com.example.insurance.controller.ResourceController.download(ResourceController.java:200)
* </pre>
*
* @param filePath 源文件路径
* @param response 请求响应
* @param contentRange 内容范围
*/
public static void copyByChannelAndBuffer(
Path filePath, HttpServletResponse response, ContentRange contentRange) {
String fileName = filePath.getFileName().toString();
// 要下载的长度
long contentLength = contentRange.applyAsContentLength();
BufferedOutputStream outputStream = null;
FileChannel inChannel = null;
// 已传送数据大小
long transmitted = 0;
long firstBytePos = contentRange.getStart();
long fileLength = contentRange.getLength();
try {
inChannel = FileChannel.open(filePath, StandardOpenOption.READ, StandardOpenOption.WRITE);
// 建立直接缓冲区
MappedByteBuffer inMap = inChannel.map(FileChannel.MapMode.READ_ONLY, firstBytePos, fileLength);
outputStream = new BufferedOutputStream(response.getOutputStream(), BUFFER_SIZE);
// 把数据读取到缓冲区中
byte[] buffer = new byte[BUFFER_SIZE];
int len = BUFFER_SIZE;
// warning：判断是否到了最后不足4096（buffer的length）个byte这个逻辑（(transmitted + len) <= contentLength）要放前面
// 不然会会先读取file，造成后面读取位置出错
while ((transmitted + len) <= contentLength) {
inMap.get(buffer);
outputStream.write(buffer, 0, len);
transmitted += len;
log.info("fileName={}, transmitted={}", fileName, transmitted);
}
// 处理不足buffer.length部分
if (transmitted < contentLength) {
len = (int) (contentLength - transmitted);
buffer = new byte[len];
inMap.get(buffer);
outputStream.write(buffer, 0, len);
transmitted += len;
log.info("fileName={}, transmitted={}", fileName, transmitted);
}
log.info("file download complete, fileName={}, transmitted={}", fileName, transmitted);
} catch (ClientAbortException e) {
// 捕获此异常表示用户停止下载
log.warn("client stop file download, fileName={}, transmitted={}", fileName, transmitted);
} catch (Exception e) {
log.error("file download error, fileName={}, transmitted={}", fileName, transmitted, e);
} finally {
IoUtil.close(outputStream);
IoUtil.close(inChannel);
}
}
}
ContentRange
package com.example.insurance.common;
import lombok.AllArgsConstructor;
import lombok.Getter;
/**
* 内容范围对象
* <pre>
* <a href="https://www.rfc-editor.org/rfc/rfc7233#section-4.2">
* 4.2. Content-Range - HTTP/1.1 Range Requests</a>
* Content-Range: "bytes" first-byte-pos "-" last-byte-pos "/" complete-length
*
* For example:
* Content-Range: bytes 0-499/1234
* </pre>
*
* @see org.apache.catalina.servlets.DefaultServlet.Range
*/
@Getter
@AllArgsConstructor
public class ContentRange {
/**
* 第一个字节的位置
*/
private final long start;
/**
* 最后一个字节的位置
*/
private long end;
/**
* 内容完整的长度/总长度
*/
private final long length;
public static final String BYTES_STRING = "bytes";
/**
* 组装内容范围的响应头。
* <pre>
* <a href="https://www.rfc-editor.org/rfc/rfc7233#section-4.2">
* 4.2. Content-Range - HTTP/1.1 Range Requests</a>
* Content-Range: "bytes" first-byte-pos "-" last-byte-pos "/" complete-length
*
* For example:
* Content-Range: bytes 0-499/1234
* </pre>
*
* @return 内容范围的响应头
*/
public String toContentRange() {
return BYTES_STRING + ' ' + start + '-' + end + '/' + length;
// return "bytes " + start + "-" + end + "/" + length;
}
/**
* 计算内容完整的长度/总长度。
*
* @return 内容完整的长度/总长度
*/
public long applyAsContentLength() {
return end - start + 1;
}
/**
* Validate range.
*
* @return true if the range is valid, otherwise false
*/
public boolean validate() {
if (end >= length) {
end = length - 1;
}
return (start >= 0) && (end >= 0) && (start <= end) && (length > 0);
}
@Override
public String toString() {
return "firstBytePos=" + start +
", lastBytePos=" + end +
", fileLength=" + length;
}
}
StreamProgressImpl
package com.example.insurance.common;
import cn.hutool.core.io.StreamProgress;
import lombok.AllArgsConstructor;
import lombok.extern.slf4j.Slf4j;
/**
* 数据流进度条
*/
@Slf4j
@AllArgsConstructor
public class StreamProgressImpl implements StreamProgress {
private final String fileName;
@Override
public void start() {
log.info("start progress {}", fileName);
}
@Override
public void progress(long total, long progressSize) {
log.debug("progress {}, total={}, progressSize={}", fileName, total, progressSize);
}
@Override
public void finish() {
log.info("finish progress {}", fileName);
}
}
MediaContentUtil
package com.example.insurance.common;
import java.net.URLDecoder;
import java.net.URLEncoder;
import java.nio.charset.StandardCharsets;
/**
* 文件内容辅助方法集
*/
public final class MediaContentUtil {
public static String filePath() {
String osName = System.getProperty("os.name");
String filePath = "/data/files/";
if (osName.startsWith("Windows")) {
filePath = "D:\" + filePath;
}
// else if (osName.startsWith("Linux")) {
// filePath = MediaContentConstant.FILE_PATH;
// }
else if (osName.startsWith("Mac") || osName.startsWith("Linux")) {
filePath = "/home/admin" + filePath;
}
return filePath;
}
public static String encode(String fileName) {
return URLEncoder.encode(fileName, StandardCharsets.UTF_8);
}
public static String decode(String fileName) {
return URLDecoder.decode(fileName, StandardCharsets.UTF_8);
}
}
到了这里，关于Spring Boot实现HTTP大文件断点续传分片下载-大视频分段渐进式播放的文章就介绍完了。如果您还想了解更多内容，请在右上角搜索TOY模板网以前的文章或继续浏览下面的相关文章，希望大家以后多多支持TOY模板网！
