archive

type LenWriter ¶

type LenWriter struct {
	// contains filtered or unexported fields
}

func (w *LenWriter) Write(p []byte) (nn int, err error)

type RPackageArchive ¶

type RPackageArchive struct {
	// contains filtered or unexported fields
}

RPackageArchive supports rewriting .tar.gz source packages and binaries in a single data stream while (a) calculating the source and destination SHA256 checksums, (b) calculating the source and destination file sizes, (c) rewriting the `DESCRIPTION` file to include the correct `Repository` field, and (d) rewriting the `MD5` file with any corrections required for the updated DESCRIPTION.

IMPORTANT: If you make improvements to `RPackageArchive`, please also update `RPackageZipArchive` in `archive_zip.go`. We decided to maintain the code separately since there are some substantial differences in reading ZIP vs. TAR archives.

func NewRPackageArchive(bufferSize, gzipLevel int) *RPackageArchive

func (a *RPackageArchive) GetReadme(stream io.Reader, wReadme io.Writer) (bool, error)

func (a *RPackageArchive) RewriteBinary(r io.Reader, w io.Writer) (results *Results, err error)

func (a *RPackageArchive) RewriteWithReadme(r io.Reader, w, wReadme io.Writer) (results *Results, err error)

type RPackageZipArchive ¶

type RPackageZipArchive struct {
	// contains filtered or unexported fields
}

RPackageZipArchive is very similar to RPackageArchive (see `archive.go`). However, ZIP reading requires random access (io.ReaderAt), so we had to create a separate util. Instead of passing a stream as `r`, instead save the ZIP archive to a temporary file and then pass an `*os.File` as `r`. The file will be scanned twice: once to calculate the original checksum and size, and once to rewrite to the destination.

IMPORTANT: If you make improvements to `RPackageZipArchive`, please also update `RPackageArchive` in `archive.go`.

func NewRPackageZipArchive(bufferSize int) *RPackageZipArchive

func (a *RPackageZipArchive) RewriteBinary(r *os.File, w io.Writer) (results *Results, err error)

type Results ¶

type Results struct {
	OriginalSize      int64
	RewrittenSize     int64
	OriginalChecksum  string
	RewrittenChecksum string
	Description       string
	ReadmeMarkdown    bool
}

type RewriteResults ¶

type RewriteResults struct {
	Results
	RewrittenPath       string
	ExtractedReadmePath string
}