An alternative implementation of [chan copy] that operates in precisely

those circumstances where the incumbent implementation is known to perform
a byte-for-byte copy from input channel to output channel.

In that case, the new implementation ceases the activity of copying every
byte from buffer to buffer to buffer to buffer to effect the copy.  Instead
it arranges for the very same buffer the input channel read driver read into
to be the buffer the output channel write driver writes out from.  No data
copying (or even examining) at all.  When the proper conditions apply, this
ought to be noticeably faster.

1 files changed, 254 insertions, 7 deletions

diff --git a/generic/tclIO.c b/generic/tclIO.c
index d7ed3ac..d9d37a2 100644
--- a/generic/tclIO.c
+++ b/generic/tclIO.c
@@ -182,6 +182,14 @@ static int		CloseChannelPart(Tcl_Interp *interp, Channel *chanPtr,
 static int		CloseWrite(Tcl_Interp *interp, Channel *chanPtr);
 static void		CommonGetsCleanup(Channel *chanPtr);
 static int		CopyData(CopyState *csPtr, int mask);
+static int		MoveBytes(CopyState *csPtr);
+
+static void		MBCallback(CopyState *csPtr, Tcl_Obj *errObj);
+static void		MBError(CopyState *csPtr, int mask, int errorCode);
+static int		MBRead(CopyState *csPtr);
+static int		MBWrite(CopyState *csPtr);
+static void		MBEvent(ClientData clientData, int mask);
+
 static void		CopyEventProc(ClientData clientData, int mask);
 static void		CreateScriptRecord(Tcl_Interp *interp,
 			    Channel *chanPtr, int mask, Tcl_Obj *scriptPtr);
@@ -8778,6 +8786,7 @@ TclCopyChannel(
     int readFlags, writeFlags;
     CopyState *csPtr;
     int nonBlocking = (cmdPtr) ? CHANNEL_NONBLOCKING : 0;
+    int moveBytes;
 
     inStatePtr = inPtr->state;
     outStatePtr = outPtr->state;
@@ -8829,13 +8838,24 @@ TclCopyChannel(
 	    | CHANNEL_UNBUFFERED;
 
     /*
+     * Test for conditions where we know we can just move bytes from input
+     * channel to output channel with no transformation or even examination
+     * of the bytes themselves.
+     */
+
+    moveBytes = inStatePtr->inEofChar == '\0'	/* No eofChar to stop input */
+	    && inStatePtr->inputTranslation == TCL_TRANSLATE_LF
+	    && outStatePtr->outputTranslation == TCL_TRANSLATE_LF
+	    && inStatePtr->encoding == outStatePtr->encoding;
+
+    /*
      * Allocate a new CopyState to maintain info about the current copy in
      * progress. This structure will be deallocated when the copy is
      * completed.
      */
 
-    csPtr = ckalloc(sizeof(CopyState) + inStatePtr->bufSize);
-    csPtr->bufSize = inStatePtr->bufSize;
+    csPtr = ckalloc(sizeof(CopyState) + !moveBytes * inStatePtr->bufSize);
+    csPtr->bufSize = !moveBytes * inStatePtr->bufSize;
     csPtr->readPtr = inPtr;
     csPtr->writePtr = outPtr;
     csPtr->readFlags = readFlags;
@@ -8851,6 +8871,10 @@ TclCopyChannel(
     inStatePtr->csPtrR  = csPtr;
     outStatePtr->csPtrW = csPtr;
 
+    if (moveBytes) {
+	return MoveBytes(csPtr);
+    }
+
     /*
      * Special handling of -size 0 async transfers, so that the -command is
      * still called asynchronously.
@@ -8885,6 +8909,225 @@ TclCopyChannel(
  *----------------------------------------------------------------------
  */
 
+static void
+MBCallback(
+    CopyState *csPtr,
+    Tcl_Obj *errObj)
+{
+    Tcl_Obj *cmdPtr = Tcl_DuplicateObj(csPtr->cmdPtr);
+    Tcl_WideInt total = csPtr->total;
+    Tcl_Interp *interp = csPtr->interp;
+    int code;
+
+    Tcl_IncrRefCount(cmdPtr);
+    StopCopy(csPtr);
+
+    /* TODO: What if cmdPtr is not a list?! */
+
+    Tcl_ListObjAppendElement(NULL, cmdPtr, Tcl_NewWideIntObj(total));
+    if (errObj) {
+	Tcl_ListObjAppendElement(NULL, cmdPtr, errObj);
+    }
+
+    Tcl_Preserve(interp);
+    code = Tcl_EvalObjEx(interp, cmdPtr, TCL_EVAL_GLOBAL);
+    if (code != TCL_OK) {
+	Tcl_BackgroundException(interp, code);
+    }
+    Tcl_Release(interp);
+    TclDecrRefCount(cmdPtr);
+}
+
+static void
+MBError(
+    CopyState *csPtr,
+    int mask,
+    int errorCode)
+{
+    Tcl_Channel inChan = (Tcl_Channel) csPtr->readPtr;
+    Tcl_Channel outChan = (Tcl_Channel) csPtr->writePtr;
+    Tcl_Obj *errObj;
+
+    Tcl_SetErrno(errorCode);
+
+    errObj = Tcl_ObjPrintf( "error %sing \"%s\": %s",
+	    (mask & TCL_READABLE) ? "read" : "writ",
+	    Tcl_GetChannelName((mask & TCL_READABLE) ? inChan : outChan),
+	    Tcl_PosixError(csPtr->interp));
+
+    if (csPtr->cmdPtr) {
+	MBCallback(csPtr, errObj);
+    } else {
+	Tcl_SetObjResult(csPtr->interp, errObj);
+	StopCopy(csPtr);
+    }
+}
+
+static void
+MBEvent(
+    ClientData clientData,
+    int mask)
+{
+    CopyState *csPtr = (CopyState *) clientData;
+    Tcl_Channel inChan = (Tcl_Channel) csPtr->readPtr;
+    Tcl_Channel outChan = (Tcl_Channel) csPtr->writePtr;
+    ChannelState *inStatePtr = csPtr->readPtr->state;
+
+    if (mask & TCL_WRITABLE) {
+	Tcl_DeleteChannelHandler(inChan, MBEvent, csPtr);
+	Tcl_DeleteChannelHandler(outChan, MBEvent, csPtr);
+	switch (MBWrite(csPtr)) {
+	case TCL_OK:
+	    MBCallback(csPtr, NULL);
+	    break;
+	case TCL_CONTINUE:
+	    Tcl_CreateChannelHandler(inChan, TCL_READABLE, MBEvent, csPtr);
+	    break;
+	}
+    } else if (mask & TCL_READABLE) {
+	if (TCL_OK == MBRead(csPtr)) {
+	    /* When at least one full buffer is present, stop reading. */
+	    if (IsBufferFull(inStatePtr->inQueueHead)
+		    || !Tcl_InputBlocked(inChan)) {
+		Tcl_DeleteChannelHandler(inChan, MBEvent, csPtr);
+	    }
+
+	    /* Successful read -- set up to write the bytes we read */
+	    Tcl_CreateChannelHandler(outChan, TCL_WRITABLE, MBEvent, csPtr);
+	}
+    }
+}
+
+static int 
+MBRead(
+    CopyState *csPtr)
+{
+    ChannelState *inStatePtr = csPtr->readPtr->state;
+    ChannelBuffer *bufPtr = inStatePtr->inQueueHead;
+    int code;
+
+    if (bufPtr && BytesLeft(bufPtr) > 0) {
+	return TCL_OK;
+    }
+
+    code = GetInput(inStatePtr->topChanPtr);
+    if (code == 0) {
+	return TCL_OK;
+    } else {
+	MBError(csPtr, TCL_READABLE, code);
+	return TCL_ERROR;
+    }
+}
+
+static int 
+MBWrite(
+    CopyState *csPtr)
+{
+    ChannelState *inStatePtr = csPtr->readPtr->state;
+    ChannelState *outStatePtr = csPtr->writePtr->state;
+    ChannelBuffer *bufPtr = inStatePtr->inQueueHead;
+    ChannelBuffer *tail = NULL;
+    int code, inBytes = 0;
+
+    /* Count up number of bytes waiting in the input queue */
+    while (bufPtr) {
+	inBytes += BytesLeft(bufPtr);
+	tail = bufPtr;
+	if (csPtr->toRead != -1 && csPtr->toRead < inBytes) {
+	    /* Queue has enough bytes to complete the copy */
+	    break;
+	}
+	bufPtr = bufPtr->nextPtr;
+    }
+
+    if (bufPtr) {
+	/* Split the overflowing buffer in two */
+	int extra = inBytes - csPtr->toRead;
+
+	bufPtr = AllocChannelBuffer(extra);
+
+	tail->nextAdded -= extra;
+	memcpy(InsertPoint(bufPtr), InsertPoint(tail), extra);
+	bufPtr->nextAdded += extra;
+	bufPtr->nextPtr = tail->nextPtr;
+	tail->nextPtr = NULL;
+	inBytes = csPtr->toRead;
+    }
+
+    /* Update the byte counts */
+    if (csPtr->toRead != -1) {
+	csPtr->toRead -= inBytes;
+    }
+    csPtr->total += inBytes;
+
+    /* Move buffers from input to output channels */
+    if (outStatePtr->outQueueTail) {
+	outStatePtr->outQueueTail->nextPtr = inStatePtr->inQueueHead;
+    } else {
+	outStatePtr->outQueueHead = inStatePtr->inQueueHead;
+    }
+    outStatePtr->outQueueTail = tail;
+    inStatePtr->inQueueHead = bufPtr;
+    if (bufPtr == NULL) {
+	inStatePtr->inQueueTail = NULL;
+    }
+
+    code = FlushChannel(csPtr->interp, outStatePtr->topChanPtr, 0);
+    if (code) {
+	MBError(csPtr, TCL_WRITABLE, code);
+	return TCL_ERROR;
+    }
+    if (csPtr->toRead == 0 || GotFlag(inStatePtr, CHANNEL_EOF)) {
+	return TCL_OK;
+    }
+    return TCL_CONTINUE;
+}
+
+static int
+MoveBytes(
+    CopyState *csPtr)		/* State of copy operation. */
+{
+    ChannelState *outStatePtr = csPtr->writePtr->state;
+    ChannelBuffer *bufPtr = outStatePtr->curOutPtr;
+    int errorCode;
+
+    if (bufPtr && BytesLeft(bufPtr)) {
+	/* If we start with unflushed bytes in the destination
+	 * channel, flush them out of the way first. */
+
+	errorCode = FlushChannel(csPtr->interp, outStatePtr->topChanPtr, 0);
+	if (errorCode != 0) {
+	    MBError(csPtr, TCL_WRITABLE, errorCode);
+	    return TCL_ERROR;
+	}
+    }
+
+    if (csPtr->cmdPtr) {
+	Tcl_Channel inChan = (Tcl_Channel) csPtr->readPtr;
+	Tcl_CreateChannelHandler(inChan, TCL_READABLE, MBEvent, csPtr);
+	return TCL_OK;
+    }
+
+    while (1) {
+	int code;
+
+	if (TCL_ERROR == MBRead(csPtr)) {
+	    return TCL_ERROR;
+	}
+	code = MBWrite(csPtr);
+	if (code == TCL_OK) {
+	    Tcl_SetObjResult(csPtr->interp, Tcl_NewWideIntObj(csPtr->total));
+	    StopCopy(csPtr);
+	    return TCL_OK;
+	}
+	if (code == TCL_ERROR) {
+	    return TCL_ERROR;
+	}
+	/* code == TCL_CONTINUE --> continue the loop */
+    }
+    return TCL_OK;	/* Silence compiler warnings */
+}
+
 static int
 CopyData(
     CopyState *csPtr,		/* State of copy operation. */
@@ -9410,12 +9653,16 @@ StopCopy(
     CopyState *csPtr)		/* State for bg copy to stop . */
 {
     ChannelState *inStatePtr, *outStatePtr;
+    Tcl_Channel inChan, outChan;
+
     int nonBlocking;
 
     if (!csPtr) {
 	return;
     }
 
+    inChan = (Tcl_Channel) csPtr->readPtr;
+    outChan = (Tcl_Channel) csPtr->writePtr;
     inStatePtr = csPtr->readPtr->state;
     outStatePtr = csPtr->writePtr->state;
 
@@ -9440,12 +9687,12 @@ StopCopy(
 	    csPtr->writeFlags & (CHANNEL_LINEBUFFERED | CHANNEL_UNBUFFERED);
 
     if (csPtr->cmdPtr) {
-	Tcl_DeleteChannelHandler((Tcl_Channel) csPtr->readPtr, CopyEventProc,
-		csPtr);
-	if (csPtr->readPtr != csPtr->writePtr) {
-	    Tcl_DeleteChannelHandler((Tcl_Channel) csPtr->writePtr,
-		    CopyEventProc, csPtr);
+	Tcl_DeleteChannelHandler(inChan, CopyEventProc, csPtr);
+	if (inChan != outChan) {
+	    Tcl_DeleteChannelHandler(outChan, CopyEventProc, csPtr);
 	}
+	Tcl_DeleteChannelHandler(inChan, MBEvent, csPtr);
+	Tcl_DeleteChannelHandler(outChan, MBEvent, csPtr);
 	TclDecrRefCount(csPtr->cmdPtr);
     }
     inStatePtr->csPtrR = NULL;