feat: use ContentType to control HTML extension

cmd/goget/main.go

@@ -130,9 +130,12 @@ func runPostDownload(
 			fmt.Fprintf(os.Stderr, "Warning: failed to close writer: %v\n", cerr)
 		}
 
-		// Adjust extension: append .html to HTML files without extension
+		// Adjust extension: append .html to HTML files without extension.
+		// When ContentType is available (WebDAV, Gopher), only apply to
+		// text/html responses. When ContentType is empty (HTTP legacy
+		// path), preserve the existing wget-compatible behaviour.
 		if *f.AdjustExt && *f.Output != "" {
-			if filepath.Ext(*f.Output) == "" {
+			if filepath.Ext(*f.Output) == "" && isHTMLContentType(result.ContentType) {
 				newPath := *f.Output + ".html"
 				if err := os.Rename(*f.Output, newPath); err == nil {
 					*f.Output = newPath
@@ -289,6 +292,20 @@ func runPostDownload(
 	return 0
 }
 
+// isHTMLContentType returns true when ct indicates HTML content.
+// An empty string is treated as HTML for backward compatibility with
+// protocol handlers that do not populate ContentType (e.g. HTTP).
+func isHTMLContentType(ct string) bool {
+	if ct == "" {
+		return true
+	}
+	// Strip parameters (charset, boundary, …).
+	if i := strings.IndexByte(ct, ';'); i >= 0 {
+		ct = ct[:i]
+	}
+	return strings.TrimSpace(ct) == "text/html"
+}
+
 // splitCommand splits a shell-like command string into tokens respecting double quotes.
 // Used to parse --on-complete and --on-complete-url without invoking a shell.
 func splitCommand(raw string) []string {

internal/core/types.go

@@ -103,6 +103,7 @@ type DownloadResult struct {
 	Parallel        bool
 	ChunksCount     int
 	HTTPStatusCode  int           // HTTP status code from response
+	ContentType     string        // Content-Type from response (if available)
 	TraceTimings    *TraceTimings // HTTP tracing breakdown
 }
 

internal/protocol/gopher/gopher.go

@@ -209,9 +209,22 @@ func (p *Protocol) downloadText(ctx context.Context, reader *bufio.Reader, req *
 		IPVersion:       4,
 		Speed:           speed,
 		OutputPath:      req.Output,
+		ContentType:     gopherContentType(firstType),
 	}, nil
 }
 
+// gopherContentType returns the MIME content type for a gopher item type.
+func gopherContentType(t byte) string {
+	switch t {
+	case typeHTML:
+		return "text/html"
+	case typeTextFile, typeInformation:
+		return "text/plain"
+	default:
+		return "application/octet-stream"
+	}
+}
+
 // downloadDirectory formats a directory listing as readable text.
 func (p *Protocol) downloadDirectory(reader *bufio.Reader, req *core.DownloadRequest, startTime time.Time) (*core.DownloadResult, error) {
 	var writer io.Writer
@@ -261,6 +274,7 @@ func (p *Protocol) downloadDirectory(reader *bufio.Reader, req *core.DownloadReq
 		IPVersion:       4,
 		Speed:           speed,
 		OutputPath:      req.Output,
+		ContentType:     "text/plain",
 	}, nil
 }
 
@@ -313,6 +327,7 @@ func (p *Protocol) downloadBinary(ctx context.Context, reader *bufio.Reader, req
 		IPVersion:       4,
 		Speed:           speed,
 		OutputPath:      req.Output,
+		ContentType:     "application/octet-stream",
 	}, nil
 }
 

internal/protocol/webdav/protocol.go

@@ -680,6 +680,7 @@ func (p *Protocol) downloadFile(ctx context.Context, req *core.DownloadRequest,
 		OutputPath:      req.Output,
 		Resumed:         resumed,
 		ChunksCount:     1,
+		ContentType:     resp.Header.Get("Content-Type"),
 	}, nil
 }