The stall is exposed in the test, though the test itself asserts before
it stalls.
The test is constructed to replicate the stalling state of a complex
Passthrough usecase since I was not able to reliable trigger the stall.
Some of the preconditions for triggering the stall are:
* rs.length >= rs.highWaterMark
* !rs.needReadable
* _transform() handler that can return empty transforms
* multiple sync write() calls
Combined this can trigger a case where rs.reading is not cleared when
further progress requires this. The fix is to always clear rs.reading.
Now that highWaterMark increases when there are large reads, this
greatly reduces the number of calls necessary to _read(size), assuming
that _read actually respects the size argument.
This makes it so that `stream.push(chunk)` is the only way to signal the
end of reading, removing the confusing disparity between the
callback-style _read method, and the fact that most real-world streams
do not have a 1:1 corollation between the "please give me data" event,
and the actual arrival of a chunk of data.
It is still possible, of course, to implement a `CallbackReadable` on
top of this. Simply provide a method like this as the callback:
function readCallback(er, chunk) {
if (er)
stream.emit('error', er);
else
stream.push(chunk);
}
However, *only* fs streams actually would behave in this way, so it
makes not a lot of sense to make TCP, TLS, HTTP, and all the rest have
to bend into this uncomfortable paradigm.
The Readable and Writable classes will nextTick certain things
if in sync mode. The sync flag gets unset after a call to _read
or _write. However, most of these behaviors should also be
deferred until nextTick if no reads have been made (for example,
the automatic '_read up to hwm' behavior on Readable.push(chunk))
Set the sync flag to true in the constructor, so that it will not
trigger an immediate 'readable' event, call to _read, before the
user has had a chance to set a _read method implementation.
It seems like a good idea on the face of it, but lowWaterMarks are
actually not useful, and in practice should always be set to zero.
It would be worthwhile for writers if we actually did some kind of
writev() type of thing, but actually this just delays calling write()
and the overhead of doing a bunch of Buffer copies is not worth the
slight benefit of calling write() fewer times.
The refactor in b43e544140 to use
stream.push() in Transform inadvertently caused it to immediately
consume all the written data, regardless of whether or not the readable
side was being consumed.
Only pull data through the _transform() process when the readable side
is being consumed.
Fix#4667
This also slightly changes the semantics, in that a 'readable'
event may be triggered by the first write() call, even if a
user has not yet called read().
This happens because the Transform _write() handler is calling
read(0) to start the flow of data. Technically, the new behavior
is more 'correct', since it is more in line with the semantics
of the 'readable' event in other streams.
Gil Pedersen
isaacs