What is the process order of redirection and process substitution? One line command to process pre error works Ok.
{ echo hello;jiji jio; } 2> >(while read line;do echo "____$line";done) >right.log
While this order does not.
{ echo hello;jiji jio; } >right.log 2> >(while read line;do echo "____$line";done)
$ { echo hello;jiji jio; } 2> >(while read line;do echo "____$line";done) >right.log
____jiji: command not found
$ cat right.log
hello
In the above, it is like jiji jio 2>&1 | while read line;do echo "____$line";done and echo hello >right.log.
$ { echo hello;jiji jio; } >right.log 2> >(while read line;do echo "____$line";done)
$ cat right.log
hello
____jiji: command not found
In the second, it is like { echo hello; jiji jio 2>&1 | while read line;do echo "____$line";done } >right.log. Remember that echo "____$line" portion triggers it to be redirected to >right.log (because it writes to stdout).
So yes, order does matter. In the second case >right.log redirection is remembered when processing 2> >(...) and whenever ... writes to stdout it is redirected to >right.log.