implicit conversion of bool to int

[tim]

We've started to use a coding standards checker at work. The following
code results in a warning:

38: CPlainText::CPlainText(const char *szPath,bool bTimeStamp,bool
bSaveLast)
39: :CLog()//call base class constructor to initialize members
40: {
41:
42: m_bTimeStamp=bTimeStamp;
43: try
44: {
45:
46:
47: if( true == bSaveLast )
^
Msg(2:3051) Be aware that an implicit conversion from 'bool' to 'int'
takes place.
48: {
49: CLog::SaveLastRunLog( szPath );
50: }
51:
The warning is flagged at line 47. Since the parameter is type bool,
and the 'true' is a bool, is there an implicit conversion taking place?

Yes, I've just spotted that the sub-class attributes arent in the
initialization list :(

===

I'm also interested in how others have found coding standard checkers.
Have they improved productivity, maintainability?

Thanks

Tim

[Rolf Magnus]

ti*@DELETEMErobotcrazy.com wrote:

We've started to use a coding standards checker at work. The following
code results in a warning:

38: CPlainText::CPlainText(const char *szPath,bool bTimeStamp,bool
bSaveLast)
39: :CLog()//call base class constructor to initialize members
40: {
41:
42: m_bTimeStamp=bTimeStamp;
43: try
44: {
45:
46:
47: if( true == bSaveLast )
^
Msg(2:3051) Be aware that an implicit conversion from 'bool' to 'int'
takes place.
48: {
49: CLog::SaveLastRunLog( szPath );
50: }
51:
The warning is flagged at line 47. Since the parameter is type bool,
and the 'true' is a bool, is there an implicit conversion taking place?
Yes. bool is an integral type (3.9.1p7), which means it undergoes integral
promotions (3.9.1p6). This in turn means that bool values get converted to
int when using comparison operators on them (13.6p12).
So true and bSaveLast are first both converted to int, and then the results
are compared.
I'm also interested in how others have found coding standard checkers.
Have they improved productivity, maintainability?

Sorry, I have no experience with any of those.

[Alf P. Steinbach]

* ti*@DELETEMErobotcrazy.com:

We've started to use a coding standards checker at work. The following
code results in a warning:

38: CPlainText::CPlainText(const char *szPath,bool bTimeStamp,bool
bSaveLast)
39: :CLog()//call base class constructor to initialize members
40: {
41:
42: m_bTimeStamp=bTimeStamp;
43: try
44: {
45:
46:
47: if( true == bSaveLast )
^
Msg(2:3051) Be aware that an implicit conversion from 'bool' to 'int'
takes place.
48: {
49: CLog::SaveLastRunLog( szPath );
50: }
51:
The warning is flagged at line 47. Since the parameter is type bool,
and the 'true' is a bool, is there an implicit conversion taking place?
Yes, AFAICT. bool is an integral type, an integral type is an
arithmetic type, and '==' applies the "usual arithmetic conversions" on
operands of arithmetic and enumeration type. The usual arithmetic
conversions include integral promotion, which includes the conversion
bool -> int.

Instead of 'true == something', write just 'something'.

Btw., I wouldn't name a boolean "timestamp", because that is just an
association, that it's somehow related to timestamping. And I would
certainly not use Hungarian notation! E.g, name it 'm_addTimestamp'.

Yes, I've just spotted that the sub-class attributes arent in the
initialization list :(

--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?