core/tsrc/fshell-last-test.script
author Tom Sutcliffe <thomas.sutcliffe@accenture.com>
Sun, 17 Oct 2010 18:43:12 +0100
changeset 69 849a0b46c767
parent 0 7f656887cf89
permissions -rw-r--r--
Fixed lots of issues with installing a low-caps version of fshell from SIS file. * Fixed issue in CCommandFactory whereby some APIs like GetCommandInfoL could trigger allocations on the wrong heap or signals to the wrong thread. The symptoms were often seen as a crash in the which_00 thread when running ciftest. * Lots of build fixes for when FSHELL_PROTECTED_UIDS isn't defined and when all capabilities aren't available. * Added new platform.mmh macro FSHELL_OPEN_SIGNED. * Open signing of fshell SIS files is now supported for production S60 handsets. Build fshell with the FSHELL_OPEN_SIGNED macro defined (and without defining FSHELL_CAP_ALL or FSHELL_PROTECTED_UIDS) in your platform.mmh and submit \epoc32\fshell\fshell.unsigned.sis to https://www.symbiansigned.com/app/page/public/openSignedOnline.do . The following commands are not available when using Open Signing due to Platform Security restrictions: fdb; kerninfo; chunkinfo; svrinfo; objinfo; sudo; fsck; localdrive; ramdefrag; readmem; reboot; setcritical; setpriority. Others such as chkdeps, e32header, ps, and fshell itself will run but in a restricted capacity (for example, fshell will no longer allow you to modify files in the \sys\bin directory). * Removed commands objinfo, svrinfo, chunkinfo, readmem, fsck completely when memory access isn't present - previously they would still appear in the help but would give an error if you tried to run them.

#!fshell
# fshell-last-test.script
# 
# Copyright (c) 2010 Accenture. All rights reserved.
# This component and the accompanying materials are made available
# under the terms of the "Eclipse Public License v1.0"
# which accompanies this distribution, and is available
# at the URL "http://www.eclipse.org/legal/epl-v10.html".
# 
# Initial Contributors:
# Accenture - Initial contribution
#

# Test that the behaviour of the 'last' attribute behaves the way we expect, in various awkward cases involving pipes and pipelines

source $SCRIPT_PATH\errordef.script

export TEST_TEXT "Something with spaces and 'inner quotes' and --options"

echo -n "Something with spaces and 'inner quotes' and --options" | export -s RESULT
var RESULT == "$TEST_TEXT" || $Error

echo -n Something with spaces and 'inner quotes' and --options | export -s RESULT
var RESULT == "$TEST_TEXT" || $Error

fshell -e "echo -n a && echo -n b" | export -s RESULT
var RESULT == "ab" || $Error

repeat 1 "fshell -e 'echo -n c && echo -n d'" | export -s RESULT
var RESULT == "cd" || $Error

repeat 3 echo a && echo b | export -s RESULT
var RESULT == b^r^n || $Error # Apparently pipe only operates on a single pipeline stage

# Put the repeat inside an fshell -e merely so we can pipe the whole lot into export
fshell -e "repeat 3 echo -n a && echo -n b" | export -s RESULT
var RESULT == aaab || $Error # last doesn't include &&

fshell -e "repeat 3 'echo -n a && echo -n b'" | export -s RESULT
var RESULT == ababab || $Error

fshell -e "repeat 1 echo -n 'quotation ^^'soup^^''" | export -s RESULT
var RESULT == "quotation 'soup'" || $Error

export ARG1 "Arg 1"
export ARG2 "2222"
export ARG3 "Arg '3' has inner quotes"
fshell $SCRIPT_PATH\checkargs.script "Arg 1" 2222 "Arg '3' has inner quotes"

# Check that escape sequences inside a "last" command are handled correctly (ie aren't expanded)
echo hello ^t world | export -s RESULT
var RESULT == "hello ^^t world^r^n" || $Error

# Check that variable expansion inside a "last" command is handled correctly
echo -n dollar questionmark is: $? | export -s RESULT
var RESULT == "dollar questionmark is: 0" || $Error