On 12/12/16 5:11 AM, Laur Aliste wrote: > > Hi, > is it accepted behavior for a failing assignment of a readonly variable to > return successfully?
The readonly builtin did not fail; the variable `j' was set readonly, and, as a side effect, the shell assigned it the null string. The assignment to i fails because Posix says it should: "If there is no command name, but the command contained a command substitution, the command shall complete with the exit status of the last command substitution performed." http://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#tag_18_09_01 > > See following code: > > fail() { > return 1 > } > > o() { > local i j > > i="$(fail)" || echo "i init failed" > readonly j="$(fail)" || echo "j init failed" # this echo is not executed > } -- ``The lyf so short, the craft so long to lerne.'' - Chaucer ``Ars longa, vita brevis'' - Hippocrates Chet Ramey, UTech, CWRU c...@case.edu http://cnswww.cns.cwru.edu/~chet/
