bgexec - Run programs in the background while handling Tk events.

blt::bgexec varName ?option value?... program ?arg?...

The bgexec command executes programs in the background, allowing Tk to handle events. A global Tcl variable varName is set when the program has completed.

Tcl's exec command is very useful for gathering information from the operating system. It runs a program and returns the output as its result. This works well for Tcl-only applications. But for Tk applications, a problem occurs when the program takes time to process. Let's say we want the get the disk usage of a directory. We'll use the Unix program CWdu to get the summary. set out [exec du -s $dir] puts "Disk usage for $dir is $out" While CWdu is running, scrollbars won't respond. None of the Tk widgets will be redrawn properly. The send command won't work. And the worst part is that the application appears hung up or dead. The problem is that while exec is waiting for du to finish, Tk is not able to handle X events.

The bgexec command performs the same functions as exec, but also allows Tk to handle events. You can execute a long-running program and the Tk widgets will behave normally. When the program finishes, its output and the exit status are written to Tcl variables. This makes it easy to monitor and save the output of a program.

Here is the disk usage example again, this time using bgexec. The syntax to invoke "du" is exactly the same as the previous example, when we used exec. global myStatus myOutput blt::bgexec myStatus -output myOutput du -s $dir puts "Disk usage for $dir is $myOutput" Two global variables, CWmyStatus and CWmyOutput, will be set by bgexec when CWdu has completed. CWMyStatus will contain the program's exit status. CWMyOutput, specified by the -output option, will store the output of the program.

You can also terminate the program by setting the variable CWmyStatus. If CWmyStatus is set before CWdu has completed, the process is killed. Under Unix, this is done sending by a configurable signal (by default it's SIGKILL). Under Win32, this is done by calling TerminateProcess. It makes no difference what CWmyStatus is set to. set myStatus {} There are several bgexec options to collect different types of information. global myStatus myOutput myErrs blt::bgexec myStatus -output myOutput -error myErrs du -s $dir The -error option is similar to -output. It sets a global variable when the program completes. The variable will contain any data written to stderr by the program.

The -output and -error variables are set only after the program completes. But if the program takes a long time, to run you may want to receive its partial output. You can gather data as it becomes available using the -onoutput option. It specifies a Tcl command prefix. Whenever new data is available, this command is executed, with the data appended as an argument to the command. proc GetInfo { data } { puts $data } blt::bgexec myStatus -onoutput GetInfo du -s $dir When output is available, the procedure CWGetInfo is called. The -onerror option performs a similar function for the stderr data stream.

Like exec, bgexec returns an error if the exit code of the program is not zero. If you think you may get a non-zero exit code, you might want to invoke bgexec from within a catch. catch { blt::bgexec myStatus -output myOutput du -s $dir } By default, bgexec will wait for the program to finish. But you can detach the program making ampersand (&) the last argument on the command line. global myStatus myOutput blt::bgexec myStatus -output myOutput du -s $dir & Bgexec will return immediately and its result will be a list of the spawned process ids. If at some point you need to wait for the program to finish up, you can use tkwait. When the program finishes, the variable CWmyStatus will be written to, breaking out the tkwait command. global myStatus myOutput blt::bgexec myStatus -output myOutput du -s $dir & ... tkwait variable myStatus

The bgexec command takes the following form: blt::bgexec varName ?option value?... program ?arg?... VarName is the name of a global variable which is set when program has finished executing. The exit status of will be stored in varName. The exit status is a list of a status token, the process-id of the program, the exit code, and a status message. You can also prematurely terminate the program by setting varName. Under Unix, the program will be sent a signal to terminate it (by default the signal is a SIGKILL; see the -killsignal option).

Program is the name of the program to be executed and args are any extra arguments for program. The syntax of program and args is the same as the exec command. So you can redirect I/O, execute pipelines, etc. (see the exec manual for further information) just like exec. If the last argument is an ampersand (&), the program will be run detached, and bgexec will return immediately. VarName will still be set with the return status when program completes.

Option refers to the switch name always beginning with a dash (-). Value is the value of the option. Option-value pairs are terminated either by the program name, or double dashes (--). The following options are available for bgexec:

-decodeerror encodingName

Specifies the encoding of the stderr channel. This affects only data returned to the Tcl interpreter. No translation is done on file redirection.

For example if data is to be converted from Unicode for use in Tcl, you would use the "unicode" encoding. The default is that no tranlation is performed.

-decodeoutput encodingName

Specifies the encoding of the stdout channels. This affects only data returned to the Tcl interpreter. No translation is done on file redirection.

For example if data is to be converted from Unicode for use in Tcl, you would use the "unicode" encoding. The default is that no tranlation is performed.

-error varName

Specifies that a global variable varName is to be set with the contents of stderr after the program has completed.

-keepnewline boolean

Specifies that a trailing newline should be retained in the output. If boolean is true, the trailing newline is truncated from the output of the -onoutput and -output variables. The default value is CWtrue.

-killsignal signal

Specifies the signal to be sent to the program when terminating. This is available only under Unix. Signal can either be a number (typically 1-32) or a mnemonic (such as SIGINT). If signal is the empty string, then no signal is sent. The default signal is CW9 (SIGKILL).

-lasterror varName

Specifies a variable varName that is updated whenever data becomes available from standard error of the program. VarName is a global variable. Unlike the -error option, data is available as soon as it arrives.

-lastoutput varName

Specifies a variable varName that is updated whenever data becomes available from standard output of the program. VarName is a global variable. Unlike the -output option, data is available as soon as it arrives.

-linebuffered boolean

Specifies that updates should be made on a line-by-line basis. Normally when new data is available bgexec will set the variable (-lastoutput and -lasterror options) or invoke the command (-onoutput and -onerror options) delivering all the new data currently available. If boolean is true, only one line at a time will be delivered. This can be useful when you want to process the output on a line-by-line basis. The default value is CWfalse.

-output varName

Specifies that a global variable varName is to be set with the output of the program, once it has completed. If this option is not set, no output will be accumulated.

-onerror command

Specifies the start of a Tcl command that will be executed whenever new data is available from standard error. The data is appended to the command as an extra argument before it is executed.

-onoutput command

Specifies the start of a Tcl command that will be executed whenever new data is available from standard output. The data is appended to the command as an extra argument before it is executed.

-update varName

Deprecated. This option is replaced by -lasterror.

--

This marks the end of the options. The following argument will be considered the name of a program even if it starts with a dash (-).

Because bgexec allows Tk to handle events while a program is running, it's possible for an application to preempt itself with further user-interactions. Let's say your application has a button that runs the disk usage example. And while the CWdu program is running, the user accidently presses the button again. A second bgexec program will preempt the first. What this means is that the first program can not finish until the second program has completed.

Care must be taken to prevent an application from preempting itself by blocking further user-interactions (such as button clicks). The BLT busy command is very useful for just these situations. See the busy manual for details.

Since Tk 4.0, a subset of bgexec can be also achieved using the fileevent command. The steps for running a program in the background are:

Execute the program with the open command (using the "|" syntax) and save the file handle. global fileId set fileId [open "|du -s $dir" r] Next register a Tcl code snippet with fileevent to be run whenever output is available on the file handle. The code snippet will read from the file handle and save the output in a variable. fileevent fileId readable { if { [gets $fileId line] < 0 } { close $fileId set output $temp unset fileId temp } else { append temp $line } }

The biggest advantage of bgexec is that, unlike fileevent, it requires no additional Tcl code to run a program. It's simpler and less error prone. You don't have to worry about non-blocking I/O. It's handled tranparently for you.

Bgexec runs programs that fileevent can not. Fileevent assumes that the when stdout is closed the program has completed. But some programs, like the Unix CWcompress program, reopen stdout, fooling fileevent into thinking the program has terminated. In the example above, we assume that the program will write and flush its output line-by-line. However running another program, your application may block in the gets command reading a partial line.

Bgexec lets you get back the exit status of the program. It also allows you to collect data from both stdout and stderr simultaneously. Finally, since data collection is handled in C code, bgexec is faster. You get back to the Tk event loop more quickly, making your application seem more responsive.

busy, exec, tkwait

exec, background, busy