package http import ( "bytes" "context" "fmt" "git.bvbej.com/bvbej/base-golang/pkg/downloader/base" "git.bvbej.com/bvbej/base-golang/pkg/downloader/fetcher" "golang.org/x/sync/errgroup" "io" "io/ioutil" "mime" "net/http" "net/http/cookiejar" "net/url" "path" "path/filepath" "strconv" "strings" "time" ) type RequestError struct { Code int Msg string } func NewRequestError(code int, msg string) *RequestError { return &RequestError{Code: code, Msg: msg} } func (re *RequestError) Error() string { return fmt.Sprintf("http request fail,code:%d", re.Code) } type Fetcher struct { *fetcher.DefaultFetcher res *base.Resource opts *base.Options status base.Status clients []*http.Response chunks []*Chunk ctx context.Context cancel context.CancelFunc pauseCh chan any } func NewFetcher() *Fetcher { return &Fetcher{ DefaultFetcher: new(fetcher.DefaultFetcher), pauseCh: make(chan any), } } var protocols = []string{"HTTP", "HTTPS"} func FetcherBuilder() ([]string, func() fetcher.Fetcher) { return protocols, func() fetcher.Fetcher { return NewFetcher() } } func (f *Fetcher) Resolve(req *base.Request) (*base.Resource, error) { httpReq, err := buildRequest(nil, req) if err != nil { return nil, err } client := buildClient() // 只访问一个字节,测试资源是否支持Range请求 httpReq.Header.Set(base.HttpHeaderRange, fmt.Sprintf(base.HttpHeaderRangeFormat, 0, 0)) httpResp, err := client.Do(httpReq) if err != nil { return nil, err } // 拿到响应头就关闭,不用加defer httpResp.Body.Close() res := &base.Resource{ Req: req, Range: false, Files: []*base.FileInfo{}, } if base.HttpCodePartialContent == httpResp.StatusCode { // 返回206响应码表示支持断点下载 res.Range = true // 解析资源大小: bytes 0-1000/1001 => 1001 contentTotal := path.Base(httpResp.Header.Get(base.HttpHeaderContentRange)) if contentTotal != "" { parse, err := strconv.ParseInt(contentTotal, 10, 64) if err != nil { return nil, err } res.TotalSize = parse } } else if base.HttpCodeOK == httpResp.StatusCode { // 返回200响应码,不支持断点下载,通过Content-Length头获取文件大小,获取不到的话可能是chunked编码 contentLength := httpResp.Header.Get(base.HttpHeaderContentLength) if contentLength != "" { parse, err := strconv.ParseInt(contentLength, 10, 64) if err != nil { return nil, err } res.TotalSize = parse } } else { return nil, NewRequestError(httpResp.StatusCode, httpResp.Status) } file := &base.FileInfo{ Size: res.TotalSize, } contentDisposition := httpResp.Header.Get(base.HttpHeaderContentDisposition) if contentDisposition != "" { _, params, _ := mime.ParseMediaType(contentDisposition) filename := params["filename"] if filename != "" { file.Name = filename } } // Get file filename by URL if file.Name == "" && strings.Count(req.URL, "/") > 2 { file.Name = filepath.Base(req.URL) } // unknown file filename if file.Name == "" { file.Name = "unknown" } res.Files = append(res.Files, file) return res, nil } func (f *Fetcher) Create(res *base.Resource, opts *base.Options) error { f.res = res f.opts = opts f.status = base.DownloadStatusReady return nil } func (f *Fetcher) Start() (err error) { // 创建文件 name := f.filename() _, err = f.Ctl.Touch(name, f.res.TotalSize) if err != nil { return err } f.status = base.DownloadStatusStart if f.res.Range { // 每个连接平均需要下载的分块大小 chunkSize := f.res.TotalSize / int64(f.opts.Connections) f.chunks = make([]*Chunk, f.opts.Connections) f.clients = make([]*http.Response, f.opts.Connections) for i := 0; i < f.opts.Connections; i++ { var ( begin = chunkSize * int64(i) end int64 ) if i == f.opts.Connections-1 { // 最后一个分块需要保证把文件下载完 end = f.res.TotalSize - 1 } else { end = begin + chunkSize - 1 } chunk := NewChunk(begin, end) f.chunks[i] = chunk } } else { // 只支持单连接下载 f.chunks = make([]*Chunk, 1) f.clients = make([]*http.Response, 1) f.chunks[0] = NewChunk(0, 0) } f.fetch() return } func (f *Fetcher) Pause() (err error) { if base.DownloadStatusStart != f.status { return } f.status = base.DownloadStatusPause f.cancel() <-f.pauseCh return } func (f *Fetcher) Continue() (err error) { if base.DownloadStatusStart == f.status || base.DownloadStatusDone == f.status { return } f.status = base.DownloadStatusStart var name = f.filename() _, err = f.Ctl.Open(name) if err != nil { return err } f.fetch() return } func (f *Fetcher) Progress() fetcher.Progress { p := make(fetcher.Progress, 0) if len(f.chunks) > 0 { total := int64(0) for _, chunk := range f.chunks { total += chunk.Downloaded } p = append(p, total) } return p } func (f *Fetcher) filename() string { // 创建文件 var filename = f.opts.Name if filename == "" { filename = f.res.Files[0].Name } return filepath.Join(f.opts.Path, filename) } func (f *Fetcher) fetch() { f.ctx, f.cancel = context.WithCancel(context.Background()) eg, _ := errgroup.WithContext(f.ctx) for i := 0; i < f.opts.Connections; i++ { j := i //不加这一行会造成越界报错 eg.Go(func() error { return f.fetchChunk(j) }) } go func() { err := eg.Wait() // 下载停止,关闭文件句柄 _ = f.Ctl.Close(f.filename()) if f.status == base.DownloadStatusPause { f.pauseCh <- nil } else { if err != nil { f.status = base.DownloadStatusError } else { f.status = base.DownloadStatusDone } f.DoneCh <- err } }() } func (f *Fetcher) fetchChunk(index int) (err error) { filename := f.filename() chunk := f.chunks[index] httpReq, err := buildRequest(f.ctx, f.res.Req) if err != nil { return err } var ( client = buildClient() buf = make([]byte, 8192) ) // 重试10次 for i := 0; i < 10; i++ { // 如果下载完成直接返回 if chunk.Status == base.DownloadStatusDone { return } // 如果已暂停直接跳出 if f.status == base.DownloadStatusPause { break } var ( resp *http.Response retry bool ) if f.res.Range { httpReq.Header.Set(base.HttpHeaderRange, fmt.Sprintf(base.HttpHeaderRangeFormat, chunk.Begin+chunk.Downloaded, chunk.End)) } else { chunk.Downloaded = 0 } err = func() error { resp, err = client.Do(httpReq) if err != nil { return err } f.clients[index] = resp if resp.StatusCode != base.HttpCodeOK && resp.StatusCode != base.HttpCodePartialContent { err = NewRequestError(resp.StatusCode, resp.Status) return err } return nil }() if err != nil { //请求失败3s后重试 time.Sleep(time.Second * 3) continue } // 请求成功就重置错误次数,连续失败10次才终止 i = 0 retry, err = func() (bool, error) { defer func() { _ = resp.Body.Close() }() var n int for { n, err = resp.Body.Read(buf) if n > 0 { _, err = f.Ctl.Write(filename, chunk.Begin+chunk.Downloaded, buf[:n]) if err != nil { return false, err } chunk.Downloaded += int64(n) } if err != nil { if err != io.EOF { return true, err } break } } return false, nil }() if !retry { // 下载成功,跳出重试 break } } if f.status == base.DownloadStatusPause { chunk.Status = base.DownloadStatusPause } else if chunk.Downloaded >= chunk.End-chunk.Begin+1 { chunk.Status = base.DownloadStatusDone } else { if err != nil { chunk.Status = base.DownloadStatusError } else { chunk.Status = base.DownloadStatusDone } } return } func buildClient() *http.Client { // Cookie handle jar, _ := cookiejar.New(nil) return &http.Client{ Jar: jar, Timeout: time.Second * 10, } } func buildRequest(ctx context.Context, req *base.Request) (httpReq *http.Request, err error) { reqUrl, err := url.Parse(req.URL) if err != nil { return } var ( method string body io.Reader ) headers := make(map[string][]string) if req.Extra == nil { method = http.MethodGet } else { extra := req.Extra.(Extra) if extra.Method != "" { method = extra.Method } else { method = http.MethodGet } if len(extra.Header) > 0 { for k, v := range extra.Header { headers[k] = []string{v} } } if extra.Body != "" { body = ioutil.NopCloser(bytes.NewBufferString(extra.Body)) } } if ctx != nil { httpReq, err = http.NewRequestWithContext(ctx, method, reqUrl.String(), body) } else { httpReq, err = http.NewRequest(method, reqUrl.String(), body) } if err != nil { return } httpReq.Header = headers return httpReq, nil }