Re: [boost] a lexical_cast<> bug?

From: Corwin Joy (cjoy_at_[hidden])
Date: 2001-07-20 19:12:18

----- Original Message -----
From: "Jeff Garland" <jeff_at_[hidden]>
To: <boost_at_[hidden]>
Sent: Friday, July 20, 2001 7:00 PM
Subject: RE: [boost] a lexical_cast<> bug?

> I believe this is the documented (and reasonable) behavior.

I'm not necessarily saying it's unreasonable.  My issue was that if this is
seen to be the 'correct' behavior, I think it would be good for the docs to
mention this explicitly since it was fairly non-obvious from reading them
that the int cast would fail.

> You are really
> asking for 2 conversion from string->double and then from double to int
with
> truncation.

Well, I'm not sure I agree with that.  Really, what  I want is for it to
stream into an integer & tell me if it cannot do so correctly.

>I would note, however, that the docs speak of "lexical_context"
> that allows control of this behavior.  AFAIK the context isn't
implemented.
> Anyway the doc sez -->
>
> "Returns the result of streaming arg into a std::stringstream and then out
as a
> Target object. The conversion is parameterized by the current
lexical_context,
> if set. If the conversion is unsuccessful, a bad_lexical_cast exception is
> thrown if the current
>               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> lexical_context is set for throwing or if there is no current
lexical_context
> set, otherwise a Target() is returned. "
>

I saw that, but I didn't think it made it obvious that the int cast would
fail.

> The more subtle issue is when the double is truncated due to the default
> precision setting of the stream.  This doesn't throw an exception, but
doesn't
> give the right correct result either.  I think this has been discussed
> before....
>

Another good point!  And perhaps also worth mentioning in the doc!

> Jeff
>
>
> > -----Original Message-----
> > From: Corwin Joy [mailto:cjoy_at_[hidden]]
> > Sent: Friday, July 20, 2001 7:38 PM
> > To: boost_at_[hidden]
> > Subject: [boost] a lexical_cast<> bug?
> >
> >
> > While I was working with the lexical_cast program under VC6 I noticed
the
> > following (suprising) behaviour:
> >
> > string s("123.456");
> > double d = lexical_cast<double>(s);  // this will succeed
> > int i = lexical_cast<int>(s); // this will throw
> >
> > It turns out that the lexical_cast to int will throw because the stream
> > operator into int stops at the decimal point.  Therefore lexical cast
says
> > that the int operator did not "succeed" because it did not read the full
> > string -- even though it reads the integer part of the string and
> > initializes the local result to 123 just fine.  It's a bit of a surprise
> > from a user point of view that this fails.  So the question is:
> > A.  Is it intended that the class should throw in this case?
> > B.  If it is intended / this kind of cast is not allowed it seems to me
like
> > a strong warning in the docs should be added.
> >
> > IMHO - given the documentation, my preference would be to modify
> > lexical_cast to support casting of a string to an int even when the
string
> > contains a fractional part.
> >
> >
> >
> >
> > Info: http://www.boost.org  Unsubscribe:
> > <mailto:boost-unsubscribe_at_[hidden]>
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
>
>
> Info: http://www.boost.org  Unsubscribe:
<mailto:boost-unsubscribe_at_[hidden]>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>