make tests: Locale and whitespaces, ENHANCEMENT

[王铭烨]

Configuration Information:
Machine: x86_64
OS: cygwin
Compiler: gcc
Compilation CFLAGS:  -DPROGRAM='bash.exe' -DCONF_HOSTTYPE='x86_64' 
-DCONF_OSTYPE='cygwin' -DCONF_MACHTYPE='x86_64-unknown-cygwin' 
-DCONF_VENDOR='unknown' -DLOCALEDIR='/usr/share/locale' -DPACKAGE='bash' 
-DSHELL -DHAVE_CONFIG_H -DRECYCLES_PIDS   -I.  
-I/usr/src/bash-4.3.39-2.x86_64/src/bash-4.3 
-I/usr/src/bash-4.3.39-2.x86_64/src/bash-4.3/include 
-I/usr/src/bash-4.3.39-2.x86_64/src/bash-4.3/lib  -DWORDEXP_OPTION -ggdb -O2 
-pipe -Wimplicit-function-declaration 
-fdebug-prefix-map=/usr/src/bash-4.3.39-2.x86_64/build=/usr/src/debug/bash-4.3.39-2
 
-fdebug-prefix-map=/usr/src/bash-4.3.39-2.x86_64/src/bash-4.3=/usr/src/debug/bash-4.3.39-2
uname output: CYGWIN_NT-6.3 VAIO-lynn 2.0.4(0.287/5/3) 2015-06-09 12:22 x86_64 
Cygwin
Machine Type: x86_64-unknown-cygwin

Description:
        With the current version of bash source, when `make tests` or `cd 
tests; sh run-all` is invoked, bash reports several differences in locale 
messages which can all be
        reduced by adding a line like `export LC_ALL=c` in `tests/run-all`.

        The `run-intl` test also, asks the user to manually ignore whitespace 
differences, but it can be ignored with `diff -w`.

Repeat-By:
        1. Get a non-C-locale environment, say `LC_ALL=de_DE.UTF-8` or 
`LC_ALL=zh_CN.UTF-8`.
        2. Type `make tests`. On Cygwin where ld fails with all those missing 
`setsigjmp` symbols, I forced it using `ln -s /bin/bash.exe .`.
        3. Sit back and see.

        Here I am using the `de` locale instead of `zh` which I use, so most of 
the characters used would be in the ASCII set and it can display properly in 
more places.

        $ LC_ALL=de_DE.UTF-8 THIS_SH=$BASH sh run-alias
        3c3
        < ./alias.tests: Zeile 25: qfoo: Kommando nicht gefunden.
        ---
        > ./alias.tests: line 25: qfoo: command not found

Fix:
        Already mentioned in description.

Notes:
        `grep -Er '(LC_ALL|LANG)' .  | cut -d: -f 1 | cut -c 3- | sed -e 
's/\..*//g' -e 's/[0-9]$//g' | uniq | tr '\n' ' '` tells me that the following 
test files handles the locale stuffs somehow:

        dstack errors execscript globstar histexp intl new-exp posixpat printf 
redir source unicode

        But there is still some way to go.