diff options
author | Ron Yorston <rmy@frippery.org> | 2015-07-01 16:45:40 +0100 |
---|---|---|
committer | Denys Vlasenko <vda.linux@googlemail.com> | 2015-07-13 03:50:27 +0200 |
commit | 0e056f7e9efcebbbb85444221e141b37d3ab79e6 (patch) | |
tree | 99836b8ecfbf80cea732c709eebc1304a762d185 /shell/README | |
parent | 97f2f7ca7fe10783a592087df989a7f394492fa1 (diff) | |
download | busybox-0e056f7e9efcebbbb85444221e141b37d3ab79e6.tar.gz |
ash: remove parsebackquote flag
Commit 503a0b8 from git://git.kernel.org/pub/scm/utils/dash/dash.git
by Herbert Xu says:
>The parsebackquote flag is only used in a test where it always has the
>value zero. So we can remove it altogether.
The first statement is incorrect: parsebackquote is non-zero when
backquotes (as opposed to $(...)) are used for command substitution.
It is possible for the test to be executed with parsebackquote != 0 in
that case.
The test is question checks whether quotes have been closed, raising
the error "unterminated quoted string" if they haven't. There seems
to be no good reason to allow unclosed quotes within backquotes. Bash,
hush and dash (after commit 503a0b8) all treat the following as an error:
XX=`"pwd`
whereas BusyBox ash doesn't. It just ignores the unclosed quote and
executes pwd.
So, parsebackquote should be removed but not for the reason stated.
function old new delta
parsebackquote 1 - -1
readtoken1 3222 3182 -40
------------------------------------------------------------------------------
(add/remove: 0/1 grow/shrink: 0/1 up/down: 0/-41) Total: -41 bytes
Signed-off-by: Ron Yorston <rmy@frippery.org>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Diffstat (limited to 'shell/README')
0 files changed, 0 insertions, 0 deletions