All LayGO API function names and argument sequences are identical to those documented in the C language API. For instance, the code to open a CID which might be written in C as:
#include <laygo.h>
LCid cid;
cid = lgo_Open("PHYS0");
if (cid < 0)
{
PostOpenError(cid);
}
else
{
ConnectCid(cid);
}
translates directly into the following Tcl code:
package require Laygo
set cid [lgo_Open "PHYS0"]
if {$cid < 0} {
PostOpenError $cid
} else {
ConnectCid $cid
}
For functions which take C pointer types such as LDataBuffer or
LCtlBuffer, substitute a buffer allocated with the LayGO
memory allocation API. The C code to read data:
unsigned char buffer[1024];
LResult result;
result = lgo_Read(cid, buffer, 1024);
if (result > 0)
{
ProcessData(buffer, result);
}
else if (result == lgo_ERROR_EVENT_WAITING)
{
ReadEvent();
}
becomes the Tcl code:
set buffer [lgo_BufferNew 1024]
set result [lgo_Read $cid $buffer 1024]
if ($result > 0) {
ProcessData $buffer $result
} elseif ($result == $Laygo::ERROR_EVENT_WAITING) {
ReadEvent
}
Where a NULL pointer would be used in C, substitute the Tcl
string NULL:
LResult result; result = lgo_ConnectRequest(cid, NULL, 0);
becomes
set result [lgo_ConnectRequest $cid NULL 0]
Finally, where a pointer to an integral type would be used in C to return a value, the Tcl function returns a list of values:
unsigned char buffer[128]; LBufferSize size; LResult result; size = sizeof(buffer); result = lgo_Event(cid, buffer, &size);
becomes in Tcl:
set size 128 set buffer [lgo_BufferNew $size] set event [lgo_Event $cid $buffer $size] set result [lindex $event 0] set size [lindex $event 1]