Tcl
Tcl
Read Data from an SSH Channel
See more SSH Examples
Demonstrates the Chilkat Ssh.ChannelRead method, which reads incoming SSH channel messages. The only argument is the channel number; the ReadTimeoutMs property limits how long it waits. Received data is appended to the channel's receive buffer, then retrieved with a GetReceived... method.
The integer return value is -1 on error, -2 if nothing arrived before the timeout, 0 if EOF or channel close was processed, or — when greater than 0 — the total bytes currently buffered for the channel (not the bytes newly received by this call).
Background: SSH reading is a two-step pattern that trips people up at first: the
Channel* read/poll methods move bytes from the network into an internal per-channel buffer, and a separate GetReceived... call hands that buffer to your code. A positive return is therefore a buffer total, not a count of what this call fetched. Note that a 0 return means EOF or close — buffered data can still remain and should be drained.Chilkat Tcl Downloads
load ./chilkat.dll
set success 0
# Demonstrates the Ssh.ChannelRead method, which reads incoming SSH channel messages. The only
# argument is the channel number. The ReadTimeoutMs property limits how long it waits.
set ssh [new_CkSsh]
set sshPort 22
set success [CkSsh_Connect $ssh "ssh.example.com" $sshPort]
if {$success == 0} then {
puts [CkSsh_lastErrorText $ssh]
delete_CkSsh $ssh
exit
}
# Normally you would not hard-code the password in source. You should instead obtain it
# from an interactive prompt, environment variable, or a secrets vault.
set password "mySshPassword"
set success [CkSsh_AuthenticatePw $ssh "mySshLogin" $password]
if {$success == 0} then {
puts [CkSsh_lastErrorText $ssh]
delete_CkSsh $ssh
exit
}
set channelNum [CkSsh_OpenSessionChannel $ssh]
if {$channelNum < 0} then {
puts [CkSsh_lastErrorText $ssh]
delete_CkSsh $ssh
exit
}
# Start a shell WITHOUT requesting a PTY. Without a PTY the shell sees it is connected to a
# raw data pipe and enters non-interactive mode: it suppresses the command prompt, does not
# echo typed commands back, and skips interactive login scripts. The output is therefore
# clean and easily parseable -- but because no prompt is ever sent, code must not wait for
# one. Instead, have each command print a predictable marker, or wait for EOF / channel close.
set success [CkSsh_SendReqShell $ssh $channelNum]
if {$success == 0} then {
puts [CkSsh_lastErrorText $ssh]
delete_CkSsh $ssh
exit
}
set success [CkSsh_ChannelSendString $ssh $channelNum "uname -a\n" "utf-8"]
if {$success == 0} then {
puts [CkSsh_lastErrorText $ssh]
delete_CkSsh $ssh
exit
}
# The return value is: -1 on error, -2 if nothing arrived before the timeout, 0 if EOF or
# channel close was processed, or the total number of buffered bytes when greater than 0.
set result [CkSsh_ChannelRead $ssh $channelNum]
if {$result == -1} then {
puts [CkSsh_lastErrorText $ssh]
delete_CkSsh $ssh
exit
}
if {$result == -2} then {
puts "No channel data arrived before the timeout."
}
if {$result == 0} then {
puts "EOF or channel close was processed. Buffered data may still remain."
}
if {$result > 0} then {
puts "Total bytes buffered: $result"
}
set output [CkSsh_getReceivedText $ssh $channelNum "utf-8"]
if {[CkSsh_get_LastMethodSuccess $ssh] == 0} then {
puts [CkSsh_lastErrorText $ssh]
delete_CkSsh $ssh
exit
}
puts "$output"
CkSsh_Disconnect $ssh
delete_CkSsh $ssh