[ruby-talk:444418] logic in Ruby
Hi p(1 and 2 or 3 and 4) 1. how many brackets do we need? 2. in my opinion the result should be 2, not 4. What do you think? Andreas
There are orders of precedence at work here, but also some subtle ones that might not be clear. `and` and `&&` have different precedence. Likewise `or` and `||`. And I cannot count the number of times that I have tried to be all minimal and introduced a logical error in my code. Only staring at it long enough made it clear to me that a whole branch of my comparison was being ignored. I personally tend to put lots more parentheses in than might strictly be needed, and I treat them as if they were comments -- they explain visually what I am imagining the grouping to be. And sometimes that turns out to be necessary in order for it work the way I expect it to. A good question to ask yourself: would this be clearer as two lines? What would those two lines be? Could you use `return foo if 1 and 2` to short-circuit it? Not an answer, but maybe a way for you to interrogate the question from a different angle. Walter
On Feb 17, 2024, at 2:04 PM, Information via ruby-talk <ruby-talk@ml.ruby-lang.org> wrote:
Hi
p(1 and 2 or 3 and 4)
1. how many brackets do we need?
2. in my opinion the result should be 2, not 4.
What do you think?
Andreas
______________________________________________ ruby-talk mailing list -- ruby-talk@ml.ruby-lang.org To unsubscribe send an email to ruby-talk-leave@ml.ruby-lang.org ruby-talk info -- https://ml.ruby-lang.org/mailman3/postorius/lists/ruby-talk.ml.ruby-lang.org...
In general, you should always use `||` and `&&`. Except for expressions like follows: system "false" or raise "False return value" Note how you don't need any parentheses in this case, but you would if you used `||` or `&&`. Think of `or` and `and` more like how you think about `if` or `unless` (as in, all 4 are for control flow). In fact, all `||`, `&&`, `or`, `and` under the hood work exactly the same (except for the priority). It's rather that the verbose ones were designed for the usecase I wrote about. On Sat, 2024-02-17 at 20:04 +0100, Information via ruby-talk wrote:
Hi
p(1 and 2 or 3 and 4)
1. how many brackets do we need?
2. in my opinion the result should be 2, not 4.
What do you think?
Andreas
______________________________________________ ruby-talk mailing list -- ruby-talk@ml.ruby-lang.org To unsubscribe send an email to ruby-talk-leave@ml.ruby-lang.org ruby-talk info -- https://ml.ruby-lang.org/mailman3/postorius/lists/ruby-talk.ml.ruby-lang.org...
On Sun, Feb 18, 2024 at 3:05 AM Information via ruby-talk < ruby-talk@ml.ruby-lang.org> wrote:
Hi p(1 and 2 or 3 and 4) 1. how many brackets do we need?
none if you know the order :)
2. in my opinion the result should be 2, not 4. What do you think?
1 and 2 or 3 and 4 => 4
evaluation left to right
1 && 2 || 3 && 4 => 2
evaluation 1st: 1 && 2 => 2 evaluation 2nd: 3 && 4 => 4 evaluation last: 2 || 4 => 2 "or" and "and" have same priority. "&&" precedes "||". check: https://docs.ruby-lang.org/en/master/syntax/precedence_rdoc.html kind regards --botp Andreas
Am 19.02.24 um 11:22 schrieb botp:
On Sun, Feb 18, 2024 at 3:05 AM Information via ruby-talk <ruby-talk@ml.ruby-lang.org <mailto:ruby-talk@ml.ruby-lang.org>> wrote:
Hi p(1 and 2 or 3 and 4) 1. how many brackets do we need?
none if you know the order :)
Maybe there is a misunderstanding: p(--"--) doesnt work at all (Syntax error) p((--"--) does the job. (Maybe for this the subject was misleading_
2. in my opinion the result should be 2, not 4. What do you think?
1 and 2 or 3 and 4 => 4
evaluation left to right
1 && 2 || 3 && 4 => 2
evaluation 1st: 1 && 2 => 2 evaluation 2nd: 3 && 4 => 4 evaluation last: 2 || 4
So why should be there another result if using and/or instead of &&/|| ? (1 and 2 or 3 and 4) != (1 && 2 || 3 && 4) - That shouldn't be the case! Andreas
Unlike Python, the "word" operators in Ruby aren't logic operators, they're flow control. They way they work is basically a carry-over from the way that Perl handled them. They serve different purposes, and are not equivalent. On Sat, Feb 24, 2024, 03:24 Information via ruby-talk < ruby-talk@ml.ruby-lang.org> wrote:
Am 19.02.24 um 11:22 schrieb botp:
On Sun, Feb 18, 2024 at 3:05 AM Information via ruby-talk < ruby-talk@ml.ruby-lang.org> wrote:
Hi p(1 and 2 or 3 and 4) 1. how many brackets do we need?
none if you know the order :)
Maybe there is a misunderstanding: p(--"--) doesnt work at all (Syntax error) p((--"--) does the job. (Maybe for this the subject was misleading_
2. in my opinion the result should be 2, not 4.
What do you think?
1 and 2 or 3 and 4 => 4
evaluation left to right
1 && 2 || 3 && 4 => 2
evaluation 1st: 1 && 2 => 2 evaluation 2nd: 3 && 4 => 4 evaluation last: 2 || 4
So why should be there another result if using and/or instead of &&/|| ?
(1 and 2 or 3 and 4) != (1 && 2 || 3 && 4)
- That shouldn't be the case!
Andreas ______________________________________________ ruby-talk mailing list -- ruby-talk@ml.ruby-lang.org To unsubscribe send an email to ruby-talk-leave@ml.ruby-lang.org ruby-talk info -- https://ml.ruby-lang.org/mailman3/postorius/lists/ruby-talk.ml.ruby-lang.org...
On Sat, Feb 24, 2024, at 6:22 AM, Information via ruby-talk wrote:
So why should be there another result if using and/or instead of &&/||?
https://www.prestonlee.com/2010/08/04/ruby-on-the-perl-origins-of-and-versus...
(1 and 2 or 3 and 4) != (1 && 2 || 3 && 4) - That shouldn't be the case!
Perhaps not. I don't know why matz gave the "and"/"or" keywords the same precedence rather than making "and" higher precedence than "or", but that's been the case in ruby forever. Perhaps it was a mistake since perl does maintain the relative precedence: $ perl -e 'print(1 && 2 || 3 && 4);' 2 $ perl -e 'print(1 and 2 or 3 and 4);' 2 Or, perhaps matz felt that since the keyword form is intended for flow-control, the flat precedence was appropriate. $ git log -p parse.y [...] commit 3db12e8b236ac8f88db8eb4690d10e4a3b8dbcd4 (tag: v1_0_r2) Author: matz <matz@b2dd03c8-39d4-4d8f-98ff-823fe69b080e> Date: Fri Jan 16 12:13:05 1998 +0000 Initial revision git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@2 b2dd03c8-39d4-4d8f-98ff-823fe69b080e diff --git a/parse.y b/parse.y new file mode 100644 index 0000000000..02ff720bb3 --- /dev/null +++ b/parse.y @@ -0,0 +1,3717 @@ +/************************************************ + + parse.y - + + $Author$ + $Date$ + created at: Fri May 28 18:02:42 JST 1993 + + Copyright (C) 1993-1996 Yukihiro Matsumoto + +************************************************/ [...] +%left OR AND [...] +%left OROP +%left ANDOP [...]
On Feb 19, 2024, at 02:22, botp via ruby-talk <ruby-talk@ml.ruby-lang.org> wrote:
1 and 2 or 3 and 4 => 4
evaluation left to right
1 && 2 || 3 && 4 => 2
evaluation 1st: 1 && 2 => 2 evaluation 2nd: 3 && 4 => 4 evaluation last: 2 || 4 => 2
not quite. the LHS isn't evaluated: #>>> 1 && 2 || raise("NO!") && 4 # => 2
"or" and "and" have same priority. "&&" precedes "||".
put another way, slightly more visual:
% rake debug R="1 && 2 || 3 && 4" s(:or, s(:and, s(:lit, 1), s(:lit, 2)), s(:and, s(:lit, 3), s(:lit, 4)))
which is what everyone expects, versus:
% rake debug R="1 and 2 or 3 and 4" s(:and, s(:or, s(:and, s(:lit, 1), s(:lit, 2)), s(:lit, 3)), s(:lit, 4))
which is the same as:
% rake debug R="((1 and 2) or 3) and 4" s(:and, s(:or, s(:and, s(:lit, 1), s(:lit, 2)), s(:lit, 3)), s(:lit, 4))
participants (7)
-
botp -
Frank J. Cameron -
hmdne -
Information -
James Pacheco -
Ryan Davis -
Walter Lee Davis