URL
https://opencores.org/ocsvn/openrisc/openrisc/trunk
Subversion Repositories openrisc
[/] [openrisc/] [trunk/] [gnu-dev/] [or1k-gcc/] [libgo/] [go/] [encoding/] [ascii85/] [ascii85.go] - Rev 758
Go to most recent revision | Compare with Previous | Blame | View Log
// Copyright 2009 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package ascii85 implements the ascii85 data encoding
// as used in the btoa tool and Adobe's PostScript and PDF document formats.
package ascii85
import (
"io"
"strconv"
)
/*
* Encoder
*/
// Encode encodes src into at most MaxEncodedLen(len(src))
// bytes of dst, returning the actual number of bytes written.
//
// The encoding handles 4-byte chunks, using a special encoding
// for the last fragment, so Encode is not appropriate for use on
// individual blocks of a large data stream. Use NewEncoder() instead.
//
// Often, ascii85-encoded data is wrapped in <~ and ~> symbols.
// Encode does not add these.
func Encode(dst, src []byte) int {
if len(src) == 0 {
return 0
}
n := 0
for len(src) > 0 {
dst[0] = 0
dst[1] = 0
dst[2] = 0
dst[3] = 0
dst[4] = 0
// Unpack 4 bytes into uint32 to repack into base 85 5-byte.
var v uint32
switch len(src) {
default:
v |= uint32(src[3])
fallthrough
case 3:
v |= uint32(src[2]) << 8
fallthrough
case 2:
v |= uint32(src[1]) << 16
fallthrough
case 1:
v |= uint32(src[0]) << 24
}
// Special case: zero (!!!!!) shortens to z.
if v == 0 && len(src) >= 4 {
dst[0] = 'z'
dst = dst[1:]
n++
continue
}
// Otherwise, 5 base 85 digits starting at !.
for i := 4; i >= 0; i-- {
dst[i] = '!' + byte(v%85)
v /= 85
}
// If src was short, discard the low destination bytes.
m := 5
if len(src) < 4 {
m -= 4 - len(src)
src = nil
} else {
src = src[4:]
}
dst = dst[m:]
n += m
}
return n
}
// MaxEncodedLen returns the maximum length of an encoding of n source bytes.
func MaxEncodedLen(n int) int { return (n + 3) / 4 * 5 }
// NewEncoder returns a new ascii85 stream encoder. Data written to
// the returned writer will be encoded and then written to w.
// Ascii85 encodings operate in 32-bit blocks; when finished
// writing, the caller must Close the returned encoder to flush any
// trailing partial block.
func NewEncoder(w io.Writer) io.WriteCloser { return &encoder{w: w} }
type encoder struct {
err error
w io.Writer
buf [4]byte // buffered data waiting to be encoded
nbuf int // number of bytes in buf
out [1024]byte // output buffer
}
func (e *encoder) Write(p []byte) (n int, err error) {
if e.err != nil {
return 0, e.err
}
// Leading fringe.
if e.nbuf > 0 {
var i int
for i = 0; i < len(p) && e.nbuf < 4; i++ {
e.buf[e.nbuf] = p[i]
e.nbuf++
}
n += i
p = p[i:]
if e.nbuf < 4 {
return
}
nout := Encode(e.out[0:], e.buf[0:])
if _, e.err = e.w.Write(e.out[0:nout]); e.err != nil {
return n, e.err
}
e.nbuf = 0
}
// Large interior chunks.
for len(p) >= 4 {
nn := len(e.out) / 5 * 4
if nn > len(p) {
nn = len(p)
}
nn -= nn % 4
if nn > 0 {
nout := Encode(e.out[0:], p[0:nn])
if _, e.err = e.w.Write(e.out[0:nout]); e.err != nil {
return n, e.err
}
}
n += nn
p = p[nn:]
}
// Trailing fringe.
for i := 0; i < len(p); i++ {
e.buf[i] = p[i]
}
e.nbuf = len(p)
n += len(p)
return
}
// Close flushes any pending output from the encoder.
// It is an error to call Write after calling Close.
func (e *encoder) Close() error {
// If there's anything left in the buffer, flush it out
if e.err == nil && e.nbuf > 0 {
nout := Encode(e.out[0:], e.buf[0:e.nbuf])
e.nbuf = 0
_, e.err = e.w.Write(e.out[0:nout])
}
return e.err
}
/*
* Decoder
*/
type CorruptInputError int64
func (e CorruptInputError) Error() string {
return "illegal ascii85 data at input byte " + strconv.FormatInt(int64(e), 10)
}
// Decode decodes src into dst, returning both the number
// of bytes written to dst and the number consumed from src.
// If src contains invalid ascii85 data, Decode will return the
// number of bytes successfully written and a CorruptInputError.
// Decode ignores space and control characters in src.
// Often, ascii85-encoded data is wrapped in <~ and ~> symbols.
// Decode expects these to have been stripped by the caller.
//
// If flush is true, Decode assumes that src represents the
// end of the input stream and processes it completely rather
// than wait for the completion of another 32-bit block.
//
// NewDecoder wraps an io.Reader interface around Decode.
//
func Decode(dst, src []byte, flush bool) (ndst, nsrc int, err error) {
var v uint32
var nb int
for i, b := range src {
if len(dst)-ndst < 4 {
return
}
switch {
case b <= ' ':
continue
case b == 'z' && nb == 0:
nb = 5
v = 0
case '!' <= b && b <= 'u':
v = v*85 + uint32(b-'!')
nb++
default:
return 0, 0, CorruptInputError(i)
}
if nb == 5 {
nsrc = i + 1
dst[ndst] = byte(v >> 24)
dst[ndst+1] = byte(v >> 16)
dst[ndst+2] = byte(v >> 8)
dst[ndst+3] = byte(v)
ndst += 4
nb = 0
v = 0
}
}
if flush {
nsrc = len(src)
if nb > 0 {
// The number of output bytes in the last fragment
// is the number of leftover input bytes - 1:
// the extra byte provides enough bits to cover
// the inefficiency of the encoding for the block.
if nb == 1 {
return 0, 0, CorruptInputError(len(src))
}
for i := nb; i < 5; i++ {
// The short encoding truncated the output value.
// We have to assume the worst case values (digit 84)
// in order to ensure that the top bits are correct.
v = v*85 + 84
}
for i := 0; i < nb-1; i++ {
dst[ndst] = byte(v >> 24)
v <<= 8
ndst++
}
}
}
return
}
// NewDecoder constructs a new ascii85 stream decoder.
func NewDecoder(r io.Reader) io.Reader { return &decoder{r: r} }
type decoder struct {
err error
readErr error
r io.Reader
end bool // saw end of message
buf [1024]byte // leftover input
nbuf int
out []byte // leftover decoded output
outbuf [1024]byte
}
func (d *decoder) Read(p []byte) (n int, err error) {
if len(p) == 0 {
return 0, nil
}
if d.err != nil {
return 0, d.err
}
for {
// Copy leftover output from last decode.
if len(d.out) > 0 {
n = copy(p, d.out)
d.out = d.out[n:]
return
}
// Decode leftover input from last read.
var nn, nsrc, ndst int
if d.nbuf > 0 {
ndst, nsrc, d.err = Decode(d.outbuf[0:], d.buf[0:d.nbuf], d.readErr != nil)
if ndst > 0 {
d.out = d.outbuf[0:ndst]
d.nbuf = copy(d.buf[0:], d.buf[nsrc:d.nbuf])
continue // copy out and return
}
}
// Out of input, out of decoded output. Check errors.
if d.err != nil {
return 0, d.err
}
if d.readErr != nil {
d.err = d.readErr
return 0, d.err
}
// Read more data.
nn, d.readErr = d.r.Read(d.buf[d.nbuf:])
d.nbuf += nn
}
panic("unreachable")
}
Go to most recent revision | Compare with Previous | Blame | View Log