You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@openoffice.apache.org by Mathias Röllig <mr...@gmx.net> on 2015/10/12 21:40:13 UTC
[API] Priority Problem with AND and OR
Hello!
I stumbled into a priority problem with the boolean operators AND and
OR. I cannot find any documentation for this.
First mathematical examples.
With
MsgBox( 3 * 2 ^ 2 )
you will see, that ^ has a higher priority than *.
(2 ^ 2) = 4
4 * 3 = 12
With
MsgBox( (3 * 2) ^ 2 )
you will get the right result 6 ^ 2 = 36.
With
MsgBox( 3 + 4 * 3 )
you will see, that * has a higher priority than +.
(4 * 3) = 12
3 * 12 = 36
With
MsgBox( (3 + 4) * 3 )
you will get the right result 7 * 3 = 21.
Now looking at the same logic with logical operators.
Dim bResult As Boolean
bResult = TRUE Or FALSE And TRUE
MsgBox( bResult )
bResult = TRUE Or TRUE And FALSE
MsgBox( bResult )
What do you expect?
For logical operations AND is equivalent to * and OR is equivalent to +.
AND should have a higher priority than OR, so I would expect in both
cases (because TRUE Or (<anything>) = TRUE):
bResult = TRUE
But you will get
TRUE Or FALSE And TRUE = (TRUE Or FALSE) And TRUE = TRUE
TRUE Or TRUE And FALSE = (TRUE Or TRUE) And FALSE = FALSE
Is there any explanation that AND and OR have (and should have) the same
priority?
Regards, Mathias
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by Andrew Pitonyak <an...@pitonyak.org>.
I was tempted to open a bug against this some years back, but, this is
the type of change that I can't help but wonder if it is more dangerous
to affect existing code, or to use rules by new users who are unaware
that ^ does not follow expected rules, and neither does AND and OR. Same
is true for bitwise operations.
On 12.10.2015 15:58, Rory O'Farrell wrote:
> On Mon, 12 Oct 2015 15:45:16 -0400
> Andrew Pitonyak <an...@pitonyak.org> wrote:
>
>> Yes, I note this travesty in OOME. There is also a difference in the
>> way that it handles exponentiation. Standard rules indicate that
>> 2^3^4
>> is evaluated as 2^(3^4) rather than (2^3)^4, wihch is what OOo does.
>
> This problem came up recently several times on the en-Forum; we
> advised that it is best to use the brackets to define order of
> calculation (particularly as so many Calc users are mathematically
> inexperienced). It would nevertheless be good to have the matter
> correct in some revision.
>
> Rory O'Farrell
>> On 12.10.2015 15:40, Mathias Röllig wrote:
>> > Hello!
>> >
>> > I stumbled into a priority problem with the boolean operators AND
>> and
>> > OR. I cannot find any documentation for this.
>> >
>> > First mathematical examples.
>> >
>> > With
>> > MsgBox( 3 * 2 ^ 2 )
>> > you will see, that ^ has a higher priority than *.
>> > (2 ^ 2) = 4
>> > 4 * 3 = 12
>> > With
>> > MsgBox( (3 * 2) ^ 2 )
>> > you will get the right result 6 ^ 2 = 36.
>> >
>> > With
>> > MsgBox( 3 + 4 * 3 )
>> > you will see, that * has a higher priority than +.
>> > (4 * 3) = 12
>> > 3 * 12 = 36
>> > With
>> > MsgBox( (3 + 4) * 3 )
>> > you will get the right result 7 * 3 = 21.
>> >
>> > Now looking at the same logic with logical operators.
>> >
>> > Dim bResult As Boolean
>> > bResult = TRUE Or FALSE And TRUE
>> > MsgBox( bResult )
>> > bResult = TRUE Or TRUE And FALSE
>> > MsgBox( bResult )
>> >
>> > What do you expect?
>> > For logical operations AND is equivalent to * and OR is equivalent
>> to
>> > +.
>> > AND should have a higher priority than OR, so I would expect in
>> both
>> > cases (because TRUE Or (<anything>) = TRUE):
>> > bResult = TRUE
>> > But you will get
>> > TRUE Or FALSE And TRUE = (TRUE Or FALSE) And TRUE = TRUE
>> > TRUE Or TRUE And FALSE = (TRUE Or TRUE) And FALSE = FALSE
>> >
>> > Is there any explanation that AND and OR have (and should have)
>> the
>> > same priority?
>> >
>> > Regards, Mathias
>> >
>> >
>> ---------------------------------------------------------------------
>> > To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
>> > For additional commands, e-mail: dev-help@openoffice.apache.org
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
>> For additional commands, e-mail: dev-help@openoffice.apache.org
>>
>>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by Rory O'Farrell <of...@iol.ie>.
On Mon, 12 Oct 2015 15:45:16 -0400
Andrew Pitonyak <an...@pitonyak.org> wrote:
> Yes, I note this travesty in OOME. There is also a difference in the
> way that it handles exponentiation. Standard rules indicate that 2^3^4
> is evaluated as 2^(3^4) rather than (2^3)^4, wihch is what OOo does.
This problem came up recently several times on the en-Forum; we advised that it is best to use the brackets to define order of calculation (particularly as so many Calc users are mathematically inexperienced). It would nevertheless be good to have the matter correct in some revision.
Rory O'Farrell
> On 12.10.2015 15:40, Mathias Röllig wrote:
> > Hello!
> >
> > I stumbled into a priority problem with the boolean operators AND and
> > OR. I cannot find any documentation for this.
> >
> > First mathematical examples.
> >
> > With
> > MsgBox( 3 * 2 ^ 2 )
> > you will see, that ^ has a higher priority than *.
> > (2 ^ 2) = 4
> > 4 * 3 = 12
> > With
> > MsgBox( (3 * 2) ^ 2 )
> > you will get the right result 6 ^ 2 = 36.
> >
> > With
> > MsgBox( 3 + 4 * 3 )
> > you will see, that * has a higher priority than +.
> > (4 * 3) = 12
> > 3 * 12 = 36
> > With
> > MsgBox( (3 + 4) * 3 )
> > you will get the right result 7 * 3 = 21.
> >
> > Now looking at the same logic with logical operators.
> >
> > Dim bResult As Boolean
> > bResult = TRUE Or FALSE And TRUE
> > MsgBox( bResult )
> > bResult = TRUE Or TRUE And FALSE
> > MsgBox( bResult )
> >
> > What do you expect?
> > For logical operations AND is equivalent to * and OR is equivalent to
> > +.
> > AND should have a higher priority than OR, so I would expect in both
> > cases (because TRUE Or (<anything>) = TRUE):
> > bResult = TRUE
> > But you will get
> > TRUE Or FALSE And TRUE = (TRUE Or FALSE) And TRUE = TRUE
> > TRUE Or TRUE And FALSE = (TRUE Or TRUE) And FALSE = FALSE
> >
> > Is there any explanation that AND and OR have (and should have) the
> > same priority?
> >
> > Regards, Mathias
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
> > For additional commands, e-mail: dev-help@openoffice.apache.org
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
> For additional commands, e-mail: dev-help@openoffice.apache.org
>
>
--
Rory O'Farrell <of...@iol.ie>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by Andrew Pitonyak <an...@pitonyak.org>.
Yes, I note this travesty in OOME. There is also a difference in the
way that it handles exponentiation. Standard rules indicate that 2^3^4
is evaluated as 2^(3^4) rather than (2^3)^4, wihch is what OOo does.
On 12.10.2015 15:40, Mathias Röllig wrote:
> Hello!
>
> I stumbled into a priority problem with the boolean operators AND and
> OR. I cannot find any documentation for this.
>
> First mathematical examples.
>
> With
> MsgBox( 3 * 2 ^ 2 )
> you will see, that ^ has a higher priority than *.
> (2 ^ 2) = 4
> 4 * 3 = 12
> With
> MsgBox( (3 * 2) ^ 2 )
> you will get the right result 6 ^ 2 = 36.
>
> With
> MsgBox( 3 + 4 * 3 )
> you will see, that * has a higher priority than +.
> (4 * 3) = 12
> 3 * 12 = 36
> With
> MsgBox( (3 + 4) * 3 )
> you will get the right result 7 * 3 = 21.
>
> Now looking at the same logic with logical operators.
>
> Dim bResult As Boolean
> bResult = TRUE Or FALSE And TRUE
> MsgBox( bResult )
> bResult = TRUE Or TRUE And FALSE
> MsgBox( bResult )
>
> What do you expect?
> For logical operations AND is equivalent to * and OR is equivalent to
> +.
> AND should have a higher priority than OR, so I would expect in both
> cases (because TRUE Or (<anything>) = TRUE):
> bResult = TRUE
> But you will get
> TRUE Or FALSE And TRUE = (TRUE Or FALSE) And TRUE = TRUE
> TRUE Or TRUE And FALSE = (TRUE Or TRUE) And FALSE = FALSE
>
> Is there any explanation that AND and OR have (and should have) the
> same priority?
>
> Regards, Mathias
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
> For additional commands, e-mail: dev-help@openoffice.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by "Howard Morris (aka Col Boogie)" <Ho...@hotmail.com>.
Not true or misstated, see https://technet.microsoft.com/en-us/library/ms186992%28v=sql.105%29.aspx for correct priority of operators.
Without parentheses AND is evaluated before OR
Attached is a little browser program I use to test Boolean arithmetic.
It is written in html && stands for AND, || stands for OR.
Use == to test if two things are equal, like (3 == 2+1)
Howard
Hi,
Am 12.10.2015 um 21:40 schrieb Mathias Röllig:
> ...
> What do you expect?
> For logical operations AND is equivalent to * and OR is equivalent to +.
no, that's not true. AND and OR have the same priority, just as * and /
(division). (* and / have a higher priority than + and -.)
> AND should have a higher priority than OR, so I would expect in both
Expressions have to be evaluated from left to right with respect to
priority and parenthesis.
"and" and "or" are binary operators. Therfore your examples can be
written as "A or B" where "B = (C and D)". If you write your example as
"A or (C and D)" then it is the same and everyone sees immediatly, how
it is evaluated.
In both cases, the result is TRUE, because
"TRUE or anything" is TRUE, just as "TRUE or (anything)".
"(A or B) and C" is totally different.
On http://www.p-roocks.de/truthtable2.php you can create "tables of
truth" (Wahrheitstabellen) for logical expressions like "A or B and C"
and more complex ones.
> cases (because TRUE Or (<anything>) = TRUE):
> bResult = TRUE
> But you will get
> TRUE Or FALSE And TRUE = (TRUE Or FALSE) And TRUE = TRUE
> TRUE Or TRUE And FALSE = (TRUE Or TRUE) And FALSE = FALSE
>
> Is there any explanation that AND and OR have (and should have) the same
> priority?
It is as it is!
--
Grüße
Günter Marxen
---
Diese E-Mail wurde von Avast Antivirus-Software auf Viren geprüft.
https://www.avast.com/antivirus
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by Andrew Douglas Pitonyak <an...@pitonyak.org>.
On 10/13/2015 06:25 AM, Regina Henschel wrote:
>
> Other example: Basic has no shortcut evaluation of boolean expressions.
>
A major annoyance to me..... It causes a bunch or nested if statements.
--
Andrew Pitonyak
My Macro Document: http://www.pitonyak.org/AndrewMacro.odt
Info: http://www.pitonyak.org/oo.php
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by Regina Henschel <rb...@t-online.de>.
Hi,
Andrew Douglas Pitonyak schrieb:
[..]
>
> The real question becomes.... should it be changed to follow expected
> mathematical norms (breaking all existing correct code and fixing all
> broken code where people assumed it was done in the generally accepted
> way). I am glad I need not make that call.
No, it should not be changed. You cannot break existing macros after so
many years, and the behavior is not really "wrong", because it follows
the mathematical definition.
The problem is not, that Basic makes it different than some other
languages. The problem is, that it is so badly documented.
Other example: Basic has no shortcut evaluation of boolean expressions.
Kind regards
Regina
>
> On 10/12/2015 08:15 PM, Guenter Marxen wrote:
>> Hi,
>>
>> Am 12.10.2015 um 21:40 schrieb Mathias Röllig:
>>> ...
>>> What do you expect?
>>> For logical operations AND is equivalent to * and OR is equivalent to +.
>>
>> no, that's not true. AND and OR have the same priority, just as * and
>> / (division). (* and / have a higher priority than + and -.)
>>
>>> AND should have a higher priority than OR, so I would expect in both
>>
>> Expressions have to be evaluated from left to right with respect to
>> priority and parenthesis.
>
> Only for those specific operators. And the complaint was that your
> statement is unexpectedly true.
>
>>
>> "and" and "or" are binary operators. Therfore your examples can be
>> written as "A or B" where "B = (C and D)". If you write your example as
>> "A or (C and D)" then it is the same and everyone sees immediatly, how
>> it is evaluated.
>
> I wrote an implementation for a proposed international standard in
> Fortran 90 many years back.... It was suggested that I assume that
> people would not know precedence rules so that I should use parenthesis
> for all statements.... and that I should avoid things like ternary
> operators since beginning programmers may not know what they were.
>
>>
>> In both cases, the result is TRUE, because
>> "TRUE or anything" is TRUE, just as "TRUE or (anything)".
>>
>> "(A or B) and C" is totally different.
>>
>> On http://www.p-roocks.de/truthtable2.php you can create "tables of
>> truth" (Wahrheitstabellen) for logical expressions like "A or B and C"
>> and more complex ones.
>>
>>> cases (because TRUE Or (<anything>) = TRUE):
>>> bResult = TRUE
>>> But you will get
>>> TRUE Or FALSE And TRUE = (TRUE Or FALSE) And TRUE = TRUE
>>> TRUE Or TRUE And FALSE = (TRUE Or TRUE) And FALSE = FALSE
>>>
>>> Is there any explanation that AND and OR have (and should have) the same
>>> priority?
>>
>> It is as it is!
>>
> You are correct in that!
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by Andrea Pescetti <pe...@apache.org>.
On 13/10/2015 Andrew Douglas Pitonyak wrote:
> The real question becomes.... should it be changed to follow expected
> mathematical norms (breaking all existing correct code and fixing all
> broken code where people assumed it was done in the generally accepted
> way).
No, but we should advise people to use parentheses to ensure
disambiguation when the operator precedence it is not totally obvious.
Regards,
Andrea.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by Andrew Douglas Pitonyak <an...@pitonyak.org>.
Guenter, although you are correct with respect to Basic as implemented
in AOO (and you stated it very well... I enjoyed your post), I am not
aware of any other computer language where this is true. For example:
C and C++
http://en.cppreference.com/w/c/language/operator_precedence
Java is the same
https://docs.oracle.com/javase/tutorial/java/nutsandbolts/operators.html
Even VBA gives AND higher precedence than OR
https://msdn.microsoft.com/en-us/library/fw84t893.aspx
General comments on what is expected is listed here:
https://en.wikipedia.org/wiki/Order_of_operations
It is noted that APL, Smalltalk, and Occam do not have operator
precedence. Can't say I have used APL in years (like 25 to 30 years). I
have not used the other two.
The real question becomes.... should it be changed to follow expected
mathematical norms (breaking all existing correct code and fixing all
broken code where people assumed it was done in the generally accepted
way). I am glad I need not make that call.
On 10/12/2015 08:15 PM, Guenter Marxen wrote:
> Hi,
>
> Am 12.10.2015 um 21:40 schrieb Mathias Röllig:
>> ...
>> What do you expect?
>> For logical operations AND is equivalent to * and OR is equivalent to +.
>
> no, that's not true. AND and OR have the same priority, just as * and
> / (division). (* and / have a higher priority than + and -.)
>
>> AND should have a higher priority than OR, so I would expect in both
>
> Expressions have to be evaluated from left to right with respect to
> priority and parenthesis.
Only for those specific operators. And the complaint was that your
statement is unexpectedly true.
>
> "and" and "or" are binary operators. Therfore your examples can be
> written as "A or B" where "B = (C and D)". If you write your example as
> "A or (C and D)" then it is the same and everyone sees immediatly, how
> it is evaluated.
I wrote an implementation for a proposed international standard in
Fortran 90 many years back.... It was suggested that I assume that
people would not know precedence rules so that I should use parenthesis
for all statements.... and that I should avoid things like ternary
operators since beginning programmers may not know what they were.
>
> In both cases, the result is TRUE, because
> "TRUE or anything" is TRUE, just as "TRUE or (anything)".
>
> "(A or B) and C" is totally different.
>
> On http://www.p-roocks.de/truthtable2.php you can create "tables of
> truth" (Wahrheitstabellen) for logical expressions like "A or B and C"
> and more complex ones.
>
>> cases (because TRUE Or (<anything>) = TRUE):
>> bResult = TRUE
>> But you will get
>> TRUE Or FALSE And TRUE = (TRUE Or FALSE) And TRUE = TRUE
>> TRUE Or TRUE And FALSE = (TRUE Or TRUE) And FALSE = FALSE
>>
>> Is there any explanation that AND and OR have (and should have) the same
>> priority?
>
> It is as it is!
>
You are correct in that!
--
Andrew Pitonyak
My Macro Document: http://www.pitonyak.org/AndrewMacro.odt
Info: http://www.pitonyak.org/oo.php
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org
Re: [API] Priority Problem with AND and OR
Posted by Guenter Marxen <gu...@gmail.com>.
Hi,
Am 12.10.2015 um 21:40 schrieb Mathias Röllig:
> ...
> What do you expect?
> For logical operations AND is equivalent to * and OR is equivalent to +.
no, that's not true. AND and OR have the same priority, just as * and /
(division). (* and / have a higher priority than + and -.)
> AND should have a higher priority than OR, so I would expect in both
Expressions have to be evaluated from left to right with respect to
priority and parenthesis.
"and" and "or" are binary operators. Therfore your examples can be
written as "A or B" where "B = (C and D)". If you write your example as
"A or (C and D)" then it is the same and everyone sees immediatly, how
it is evaluated.
In both cases, the result is TRUE, because
"TRUE or anything" is TRUE, just as "TRUE or (anything)".
"(A or B) and C" is totally different.
On http://www.p-roocks.de/truthtable2.php you can create "tables of
truth" (Wahrheitstabellen) for logical expressions like "A or B and C"
and more complex ones.
> cases (because TRUE Or (<anything>) = TRUE):
> bResult = TRUE
> But you will get
> TRUE Or FALSE And TRUE = (TRUE Or FALSE) And TRUE = TRUE
> TRUE Or TRUE And FALSE = (TRUE Or TRUE) And FALSE = FALSE
>
> Is there any explanation that AND and OR have (and should have) the same
> priority?
It is as it is!
--
Grüße
Günter Marxen
---
Diese E-Mail wurde von Avast Antivirus-Software auf Viren geprüft.
https://www.avast.com/antivirus
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org
For additional commands, e-mail: dev-help@openoffice.apache.org