[ast-developers] Array bugs with "${a[@]}, a=() and bool -A

[Wendy Lin]

I do not know why, but the following three postings with bug reports
about array bugs with "${a[@]}, a=() and bool -A never made it to the
list. Can someone check the spam filter of the lists for the emails?

Wendy

---------- Forwarded message ----------
From: Wendy Lin <wendlin1974 at gmail.com>
Date: 10 July 2013 08:06
Subject: "${a[@]}" returns random numbers for Boolean associative
array (bool -A)
To: ast-developers at research.att.com, ast-users at research.att.com
Cc: David Korn <dgk at research.att.com>


"${a[@]}" returns random numbers for Boolean associative array (bool -A):

ksh -c 'bool -A a=( [2]=true [4]=false ) ; printf "1=%q\n" "${a[@]}" ;
a=( ) ; printf "2=%q\n" "${a[@]}"'
1=37792
1=37824
2=

AFAIK it should print:
1=true
1=false
2=


This happens with the latest 2013-06-28.

If you have a patch please share it before the next release. Thank you!

Wendy


---------- Forwarded message ----------
From: Wendy Lin <wendlin1974 at gmail.com>
Date: 10 July 2013 07:44
Subject: ksh93 returns "()" for "${a[@]}" if I clear an array with a=( )
To: ast-users at research.att.com
Cc: David Korn <dgk at research.att.com>


ksh93 returns "()" for "${a[@]}" if I clear an array with a=( )

ksh -c 'typeset -a a=( 4 6 ) ; printf "1=%q\n" "${a[@]}" ; a=( ) ;
printf "2=%q\n" "${a[@]}"'
1=4
1=6
2=$'(\n)'

This happens with 2012-08-01 and the latest 2013-06-28
/bin/ksh --version
version sh (AT&T Research) 93u+ 2012-08-01

If you have a patch please share it before the next release. Thank you!

Wendy


---------- Forwarded message ----------
From: Wendy Lin <wendlin1974 at gmail.com>
Date: 10 July 2013 07:31
Subject: Clearing an array with a=() doesn't reset the index counter
for the next array element to add
To: ast-users at research.att.com
Cc: David Korn <dgk at research.att.com>


Clearing an array with a=() doesn't reset the index counter for the
next array element to add.

Example:
ksh -c 'typeset -a a=( [0]=first ) ; a=() ; a+=( another_first ) ; print -v a'
(
[1]=another_first
)

I would expect that after a a=() the position where an add operation
inserts an element is index 0 and not index 1. Or is this intended?

This happens with 2012-08-01 and the latest 2013-06-28
/bin/ksh --version
version sh (AT&T Research) 93u+ 2012-08-01

If you have a patch please share it before the next release. Thank you!

Wendy