first with full handle over rpm

Signed-off-by: GnomeZworc <nicolas.boufidjeline@g3e.fr>

internal/clone/rpm/scanner.go

@@ -0,0 +1,143 @@
+package rpm
+
+import (
+	"compress/gzip"
+	"context"
+	"encoding/xml"
+	"fmt"
+	"io"
+	"net/http"
+	"os"
+	"path/filepath"
+	"strings"
+	"time"
+)
+
+type NewPackage struct {
+	Name         string
+	Version      string
+	Arch         string
+	Location     string
+	Checksum     string
+	ChecksumType string
+	Size         int64
+}
+
+type Scanner struct {
+	SourceURL  string
+	LocalDir   string
+	HTTPClient *http.Client
+}
+
+func NewScanner(sourceURL, localDir string) *Scanner {
+	return &Scanner{
+		SourceURL:  strings.TrimRight(sourceURL, "/"),
+		LocalDir:   localDir,
+		HTTPClient: &http.Client{Timeout: 5 * time.Minute},
+	}
+}
+
+// Scan fetches the remote package list and returns packages not present on disk.
+func (sc *Scanner) Scan(ctx context.Context) ([]NewPackage, error) {
+	repomd, err := sc.fetchRepoMD(ctx)
+	if err != nil {
+		return nil, fmt.Errorf("fetch repomd.xml: %w", err)
+	}
+
+	var primaryEntry *RepoMDEntry
+	for i := range repomd.Data {
+		if repomd.Data[i].Type == "primary" {
+			primaryEntry = &repomd.Data[i]
+			break
+		}
+	}
+	if primaryEntry == nil {
+		return nil, fmt.Errorf("no primary entry in repomd.xml")
+	}
+
+	packages, err := sc.fetchPrimary(ctx, *primaryEntry)
+	if err != nil {
+		return nil, fmt.Errorf("fetch primary.xml: %w", err)
+	}
+
+	var missing []NewPackage
+	for _, pkg := range packages {
+		localPath := filepath.Join(sc.LocalDir, filepath.FromSlash(pkg.Location.Href))
+		if _, err := os.Stat(localPath); os.IsNotExist(err) {
+			ver := pkg.Version.Ver + "-" + pkg.Version.Rel
+			if pkg.Version.Epoch != "0" && pkg.Version.Epoch != "" {
+				ver = pkg.Version.Epoch + ":" + ver
+			}
+			missing = append(missing, NewPackage{
+				Name:         pkg.Name,
+				Version:      ver,
+				Arch:         pkg.Arch,
+				Location:     pkg.Location.Href,
+				Checksum:     pkg.Checksum.Value,
+				ChecksumType: pkg.Checksum.Type,
+				Size:         pkg.Size.Package,
+			})
+		}
+	}
+	return missing, nil
+}
+
+func (sc *Scanner) fetchRepoMD(ctx context.Context) (*RepoMD, error) {
+	data, err := sc.fetchBytes(ctx, sc.SourceURL+"/repodata/repomd.xml")
+	if err != nil {
+		return nil, err
+	}
+	var repomd RepoMD
+	if err := xml.Unmarshal(data, &repomd); err != nil {
+		return nil, fmt.Errorf("parse repomd.xml: %w", err)
+	}
+	return &repomd, nil
+}
+
+func (sc *Scanner) fetchPrimary(ctx context.Context, entry RepoMDEntry) ([]Package, error) {
+	data, err := sc.fetchBytes(ctx, sc.SourceURL+"/"+entry.Location.Href)
+	if err != nil {
+		return nil, err
+	}
+	if entry.Checksum.Type == "sha256" {
+		if err := verifyChecksum(data, entry.Checksum.Value); err != nil {
+			return nil, fmt.Errorf("primary.xml.gz: %w", err)
+		}
+	}
+
+	// Save upstream primary to disk so RegenerateMetadata can use it as source.
+	dest := filepath.Join(sc.LocalDir, filepath.FromSlash(entry.Location.Href))
+	if err := os.MkdirAll(filepath.Dir(dest), 0o755); err == nil {
+		if err := writeFile(dest, data); err == nil {
+			marker := filepath.Join(sc.LocalDir, "repodata", PrimarySourceFile)
+			_ = os.WriteFile(marker, []byte(entry.Location.Href), 0o644)
+		}
+	}
+
+	gz, err := gzip.NewReader(strings.NewReader(string(data)))
+	if err != nil {
+		return nil, fmt.Errorf("open gzip: %w", err)
+	}
+	defer gz.Close()
+	var primary PrimaryMetadata
+	if err := xml.NewDecoder(gz).Decode(&primary); err != nil {
+		return nil, fmt.Errorf("parse primary.xml: %w", err)
+	}
+	return primary.Packages, nil
+}
+
+func (sc *Scanner) fetchBytes(ctx context.Context, url string) ([]byte, error) {
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, url, nil)
+	if err != nil {
+		return nil, err
+	}
+	resp, err := sc.HTTPClient.Do(req)
+	if err != nil {
+		return nil, err
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		return nil, fmt.Errorf("HTTP %d for %s", resp.StatusCode, url)
+	}
+	return io.ReadAll(resp.Body)
+}