</p>

<h3 id="Channel_types">Channel types</h3>

<p>
A channel provides a mechanism for
<a href="#Go_statements">concurrently executing functions</a>
to communicate by
<a href="#Send_statements">sending</a> and
<a href="#Receive_operator">receiving</a>
values of a specified element type.
The value of an uninitialized channel is <code>nil</code>.
</p>

<pre class="ebnf">

### New unique package

The new [unique](/pkg/unique) package provides facilities for
canonicalizing values (like "interning" or "hash-consing").

Any value of comparable type may be canonicalized with the new
`Make[T]` function, which produces a reference to a canonical copy of
the value in the form of a `Handle[T]`.
Two `Handle[T]` are equal if and only if the values used to produce the
handles are equal, allowing programs to deduplicate values and reduce

	// is too large we are basically just thrashing the CPU D-cache.
	// So if the result length is larger than an empirically-found
	// limit (8KB), we stop growing the source string once the limit
	// is reached and keep reusing the same source string - that
	// should therefore be always resident in the L1 cache - until we
	// have completed the construction of the result.
	// This yields significant speedups (up to +100%) in cases where

</p>

<h3 id="Channel_types">Channel types</h3>

<p>
A channel provides a mechanism for
<a href="#Go_statements">concurrently executing functions</a>
to communicate by
<a href="#Send_statements">sending</a> and
<a href="#Receive_operator">receiving</a>
values of a specified element type.
The value of an uninitialized channel is <code>nil</code>.
</p>

<pre class="ebnf">

pkg syscall (windows-386), const ERROR_FILE_EXISTS = 80
pkg syscall (windows-386), const ERROR_FILE_NOT_FOUND = 2
pkg syscall (windows-386), const ERROR_INSUFFICIENT_BUFFER = 122
pkg syscall (windows-386), const ERROR_IO_PENDING = 997
pkg syscall (windows-386), const ERROR_MOD_NOT_FOUND = 126
pkg syscall (windows-386), const ERROR_NOT_FOUND = 1168
pkg syscall (windows-386), const ERROR_NOT_FOUND Errno

go.exit = process.exit;
WebAssembly.instantiate(fs.readFileSync(process.argv[2]), go.importObject).then((result) => {
	process.on("exit", (code) => { // Node.js exits if no event handler is pending
		if (code === 0 && !go.exited) {
			// deadlock, make Go print error and stack traces
			go._pendingEvent = { id: 0 };
			go._resume();
		}
	});
	return go.run(result.instance);
}).catch((err) => {

package zip

import (
	"compress/flate"
	"errors"
	"io"
	"sync"
)

// A Compressor returns a new compressing writer, writing to w.
// The WriteCloser's Close method must be used to flush pending data to w.
// The Compressor itself must be safe to invoke from multiple goroutines
// simultaneously, but each returned writer will be used only by
// one goroutine at a time.

		}
		pre = cur
	}
	return true
}

// alignSparseEntries mutates src and returns dst where each fragment's
// starting offset is aligned up to the nearest block edge, and each
// ending offset is aligned down to the nearest block edge.
//
// Even though the Go tar Reader and the BSD tar utility can handle entries
// with arbitrary offsets and lengths, the GNU tar utility can only handle

type readSeeker struct{ io.ReadSeeker }
type readBadSeeker struct{ io.ReadSeeker }

func (rbs *readBadSeeker) Seek(int64, int) (int64, error) { return 0, fmt.Errorf("illegal seek") }

// TestReadTruncation test the ending condition on various truncated files and
// that truncated files are still detected even if the underlying io.Reader
// satisfies io.Seeker.
func TestReadTruncation(t *testing.T) {
	var ss []string

	// CFTypeRef is a supertype of those, so it can have bad pointers in it as well.
	// We return true for the other *Ref types just so casting between them is easier.
	// We identify the correct set of types as those ending in Ref and for which
	// there exists a corresponding GetTypeID function.
	// See comment below for details about the bad pointers.
	if goos != "darwin" && goos != "ios" {
		return false
	}
	s := dt.Name