Hello! Steffen Nurpmeso wrote in <8629f1bb0471e00284eac7c61b06f9eea999ec1a.1659730455.git.steffen@sdaoden\ .eu>: ... |You will not believe it, but now that i was reading over the mail |i sent to busybox@ i saw a comma related bug in _op_apply(), and |that lead me to another test, and that then to another fix even!
So please wait with that one, shall you already look at it. On Sunday i thought about it and have to say it still has issues for some of those conditions, for example 2 and 3 of these four, s I1=I2=10 I3=I2+=1;p "<$((I1))>";e "<$I1><$I2><$I3>" s I1=I2=10 I3=I2+=1;p "<$((I1,I3))>";e "<$I1><$I2><$I3>" s I1=I2=10 I3=I2+=1;p "<$((I1+I3))>";e "<$I1><$I2><$I3>" s I1=I2=10 I3=I2+=1;p "<$((I1?I1:I3))>";e "<$I1><$I2><$I3>" and compared to how bash evaluates these constructs ( == WILD RECUR <10><I2=10><10><I2+=1> -<11><I2=10><11><I2+=1> -<21><I2=10><11><I2+=1> +<11><I2=10><10><I2+=1> +<21><I2=10><10><I2+=1> <10><I2=10><10><I2+=1> ): i will add more such recursive-evaluation-that-modifies- following-constructs tests (of course: POSIX makes it unspecified if a veriable contains anything else but solely a number, but still), and once i have it going i will post a single changeset again, ok? Sorry for the noise, the problem just came to my mind when i was reading the email that came back from busybox@, i had not seen it before. --steffen | |Der Kragenbaer, The moon bear, |der holt sich munter he cheerfully and one by one |einen nach dem anderen runter wa.ks himself off |(By Robert Gernhardt) _______________________________________________ busybox mailing list busybox@busybox.net http://lists.busybox.net/mailman/listinfo/busybox