Inexplicable error with std::string

[Dave]

Can anybody tell me why the code below should produce the error below? I'm
going nuts here! Am I missing something incredibly obvious and simple?

std::string::size_type idx;
std::string one_line;
..
..
..
idx = one_line.find('\t', 0);
error C2059: syntax error : 'constant'

[John Carson]

"Dave" <be***********@yahoo.com> wrote in message
news:vv************@news.supernews.com

Can anybody tell me why the code below should produce the error
below? I'm going nuts here! Am I missing something incredibly obvious
and simple?

std::string::size_type idx;
std::string one_line;
.
.
.
idx = one_line.find('\t', 0);
error C2059: syntax error : 'constant'

Try supplying a complete compileable sample that produces the error. The
three lines you have supplied compile fine with both VC++ 2002 and Comeau
online.
--
John Carson
1. To reply to email address, remove donald
2. Don't reply to email address (post here instead)

[Dave]

"John Carson" <do***********@datafast.net.au> wrote in message
news:3f******@usenet.per.paradox.net.au...

"Dave" <be***********@yahoo.com> wrote in message
news:vv************@news.supernews.com

Can anybody tell me why the code below should produce the error
below? I'm going nuts here! Am I missing something incredibly obvious
and simple?

std::string::size_type idx;
std::string one_line;
.
.
.
idx = one_line.find('\t', 0);
error C2059: syntax error : 'constant'

Try supplying a complete compileable sample that produces the error. The
three lines you have supplied compile fine with both VC++ 2002 and Comeau
online.
--
John Carson
1. To reply to email address, remove donald
2. Don't reply to email address (post here instead)

I don't know if this will help, but here's the full function. However, to
provide enough context to make it fully compileable would draw in thousands
of lines of code. So, this is the best I can do, but I understand if it's
not enough context and doesn't help... (BTW my platform is VC++ 7.1, but I
suppose that shouldn't matter; famous last words...)

void Sell::ChangeSale()
{
std::string::size_type first_char;
std::string::size_type idx;
std::string::size_type last_char;
int num_tabs;
std::string one_line;
std::string percent;
int total(0);

for (int i = 0; i < mVisibleMatchUps_psxListBox->count(); ++i)
{
one_line = std::string(mVisibleMatchUps_psxListBox->text(i));
num_tabs = std::count(one_line.begin(), one_line.end(), '\t');

if (num_tabs < 7)
continue;

idx = one_line.find('\t', 0);
first_char = idx + 1;
last_char = one_line.find('\t', first_char) - 1;

if (last_char < first_char)
continue;

percent = one_line.substr(first_char, last_char - first_char + 1);
total += atoi(percent.c_str());

}

mVisibleMatchUps_psxListBox->storeData() ;
DeleteSale() ;
AddSale() ;
}//ChangeSale(...

[Ron Natalie]

"Dave" <be***********@yahoo.com> wrote in message news:vv************@news.supernews.com...
]> .

idx = one_line.find('\t', 0);

Try sticking an #undef find before this line. The line looks fine, could be some joker
redefined find.

[Dave]

"Ron Natalie" <ro*@sensor.com> wrote in message
news:3f***********************@news.newshosting.co m...

"Dave" <be***********@yahoo.com> wrote in message news:vv************@news.supernews.com... ]> .

idx = one_line.find('\t', 0);
Try sticking an #undef find before this line. The line looks fine,

could be some joker redefined find.

$%&#^!*#$#*!$& (repeat 100 times)!!!!!

That actually was the problem! Man, I could just scream trying to maintain
the code base for the project I'm on. This is the most God-awful piece of
programming I've ever seen, and there are 300,000 lines of this nonsense!
The programmers that did this and the managers that allowed this and
everything else that's wrong with this app. to happen should all be
summarily executed...

Thanks, I think you just saved my sanity! (But it's doomed eventually anyway
since I'll have to continue to work on this project!)

[Cy Edmunds]

"Dave" <be***********@yahoo.com> wrote in message
news:vv************@news.supernews.com...

"Ron Natalie" <ro*@sensor.com> wrote in message
news:3f***********************@news.newshosting.co m...

"Dave" <be***********@yahoo.com> wrote in message news:vv************@news.supernews.com...

]> .

idx = one_line.find('\t', 0);

Try sticking an #undef find before this line. The line looks fine,

could be some joker

redefined find.

$%&#^!*#$#*!$& (repeat 100 times)!!!!!

That actually was the problem! Man, I could just scream trying to maintain
the code base for the project I'm on. This is the most God-awful piece of
programming I've ever seen, and there are 300,000 lines of this nonsense!
The programmers that did this and the managers that allowed this and
everything else that's wrong with this app. to happen should all be
summarily executed...

Thanks, I think you just saved my sanity! (But it's doomed eventually

anyway since I'll have to continue to work on this project!)

Ron made a great call here. It goes to show you why I hate C macros so much.
As BS once commented, they make what the compiler sees different from what
you see. How can that be good?

--
Cy
http://home.rochester.rr.com/cyhome/

[Jeff Schwab]

Dave wrote:

"Ron Natalie" <ro*@sensor.com> wrote

"Dave" <be***********@yahoo.com> wrote

idx = one_line.find('\t', 0);
Try sticking an #undef find before this line. The line looks fine,
could be some joker redefined find.

$%&#^!*#$#*!$& (repeat 100 times)!!!!!

That actually was the problem!

Okay, *that's* impressive.

Ron, are you actually the author of the dubious code in question?

;),
Jeff

[P.J. Plauger]

"Cy Edmunds" <ce******@spamless.rochester.rr.com> wrote in message
news:DJ*******************@twister.nyroc.rr.com...

Ron made a great call here. It goes to show you why I hate C macros so much. As BS once commented, they make what the compiler sees different from what
you see. How can that be good?

Maybe the compiler doesn't recognize the same dialect that you want
to speak. I agree that macros can be abused, and the OP has presented
an extreme case of such abuse. But the irony is that C++ has spawned
so many dialects over the past decade that implementors *must* make
extensive use of the preprocessor to cope.

P.J. Plauger
Dinkumware, Ltd.
http://www.dinkumware.com

[Ron Natalie]

"Jeff Schwab" <je******@comcast.net> wrote in message news:6d********************@comcast.com...

Dave wrote:

"Ron Natalie" <ro*@sensor.com> wrote

"Dave" <be***********@yahoo.com> wrote
idx = one_line.find('\t', 0);

Try sticking an #undef find before this line. The line looks fine,
could be some joker redefined find.

$%&#^!*#$#*!$& (repeat 100 times)!!!!!

That actually was the problem!

Okay, *that's* impressive.

Ron, are you actually the author of the dubious code in question?

Nope, I looked up the error in the Microsoft docs. Then I thought about how the
above syntax could gag on the '\t' or 0 literals and the only way I could think of was
if find() were NOT a function.

[Cy Edmunds]

"P.J. Plauger" <pj*@dinkumware.com> wrote in message
news:UH*******************@nwrddc02.gnilink.net...

"Cy Edmunds" <ce******@spamless.rochester.rr.com> wrote in message
news:DJ*******************@twister.nyroc.rr.com...

Ron made a great call here. It goes to show you why I hate C macros so

much.

As BS once commented, they make what the compiler sees different from what you see. How can that be good?

Maybe the compiler doesn't recognize the same dialect that you want
to speak. I agree that macros can be abused, and the OP has presented
an extreme case of such abuse. But the irony is that C++ has spawned
so many dialects over the past decade that implementors *must* make
extensive use of the preprocessor to cope.

P.J. Plauger
Dinkumware, Ltd.
http://www.dinkumware.com

Absolutely no argument here.

--
Cy
http://home.rochester.rr.com/cyhome/

[Jeff Schwab]

Cy Edmunds wrote:

"P.J. Plauger" <pj*@dinkumware.com> wrote in message
news:UH*******************@nwrddc02.gnilink.net...

"Cy Edmunds" <ce******@spamless.rochester.rr.com> wrote in message
news:DJ*******************@twister.nyroc.rr.com. ..

Ron made a great call here. It goes to show you why I hate C macros so
much.

As BS once commented, they make what the compiler sees different from
what you see. How can that be good?

Maybe the compiler doesn't recognize the same dialect that you want
to speak. I agree that macros can be abused, and the OP has presented
an extreme case of such abuse. But the irony is that C++ has spawned
so many dialects over the past decade that implementors *must* make
extensive use of the preprocessor to cope.

P.J. Plauger
Dinkumware, Ltd.
http://www.dinkumware.com

Absolutely no argument here.

I didn't realize the situation was that bad. I'm sure this is a
hopelessly naive question, but... What's wrong with complying (as
closely as possible) with the most recent standard, and providing flags
for compatibility with the most common dialects?

[Cy Edmunds]

"Jeff Schwab" <je******@comcast.net> wrote in message
news:jL********************@comcast.com...

Cy Edmunds wrote:

"P.J. Plauger" <pj*@dinkumware.com> wrote in message
news:UH*******************@nwrddc02.gnilink.net...

"Cy Edmunds" <ce******@spamless.rochester.rr.com> wrote in message
news:DJ*******************@twister.nyroc.rr.com. ..
Ron made a great call here. It goes to show you why I hate C macros so
much.

As BS once commented, they make what the compiler sees different from
what you see. How can that be good?

Maybe the compiler doesn't recognize the same dialect that you want
to speak. I agree that macros can be abused, and the OP has presented
an extreme case of such abuse. But the irony is that C++ has spawned
so many dialects over the past decade that implementors *must* make
extensive use of the preprocessor to cope.

P.J. Plauger
Dinkumware, Ltd.
http://www.dinkumware.com

Absolutely no argument here.

I didn't realize the situation was that bad. I'm sure this is a
hopelessly naive question, but... What's wrong with complying (as
closely as possible) with the most recent standard, and providing flags
for compatibility with the most common dialects?

Nothing, if that's all there was to it. Check out the macro spaghetti at the
top of various library code (standard library, boost, etc.) and see what
these poor guys have to put up with in order to get their code to compile
across platforms and compilers. It's brutal.

I'm opposed to the preprocessor if there is any reasonable alternative but
you do what you gotta do...

--
Cy
http://home.rochester.rr.com/cyhome/

[Jeff Schwab]

Cy Edmunds wrote:

"Jeff Schwab" <je******@comcast.net> wrote in message
news:jL********************@comcast.com...

Cy Edmunds wrote:

"P.J. Plauger" <pj*@dinkumware.com> wrote in message
news:UH*******************@nwrddc02.gnilink.net ...
"Cy Edmunds" <ce******@spamless.rochester.rr.com> wrote in message
news:DJ*******************@twister.nyroc.rr.co m...

>Ron made a great call here. It goes to show you why I hate C macros so
>much.

>As BS once commented, they make what the compiler sees different from
>what you see. How can that be good?

Maybe the compiler doesn't recognize the same dialect that you want
to speak. I agree that macros can be abused, and the OP has presented
an extreme case of such abuse. But the irony is that C++ has spawned
so many dialects over the past decade that implementors *must* make
extensive use of the preprocessor to cope.

P.J. Plauger
Dinkumware, Ltd.
http://www.dinkumware.com


Absolutely no argument here.

I didn't realize the situation was that bad. I'm sure this is a
hopelessly naive question, but... What's wrong with complying (as
closely as possible) with the most recent standard, and providing flags
for compatibility with the most common dialects?

Nothing, if that's all there was to it. Check out the macro spaghetti at the
top of various library code (standard library, boost, etc.) and see what
these poor guys have to put up with in order to get their code to compile
across platforms and compilers. It's brutal.

I'm opposed to the preprocessor if there is any reasonable alternative but
you do what you gotta do...

I see what you mean. So, the problem isn't so much with dialects of
C++, but with portability across platforms?

[Cy Edmunds]

"Jeff Schwab" <je******@comcast.net> wrote in message
news:lN********************@comcast.com...

Cy Edmunds wrote:

"Jeff Schwab" <je******@comcast.net> wrote in message
news:jL********************@comcast.com...

Cy Edmunds wrote:

"P.J. Plauger" <pj*@dinkumware.com> wrote in message
news:UH*******************@nwrddc02.gnilink.net ...
>"Cy Edmunds" <ce******@spamless.rochester.rr.com> wrote in message
>news:DJ*******************@twister.nyroc.rr.co m...
>
>
>
>>Ron made a great call here. It goes to show you why I hate C macros so>>much.
>
>>As BS once commented, they make what the compiler sees different from
>>what you see. How can that be good?
>
>Maybe the compiler doesn't recognize the same dialect that you want
>to speak. I agree that macros can be abused, and the OP has presented
>an extreme case of such abuse. But the irony is that C++ has spawned
>so many dialects over the past decade that implementors *must* make
>extensive use of the preprocessor to cope.
>
>P.J. Plauger
>Dinkumware, Ltd.
>http://www.dinkumware.com
>
>
Absolutely no argument here.
I didn't realize the situation was that bad. I'm sure this is a
hopelessly naive question, but... What's wrong with complying (as
closely as possible) with the most recent standard, and providing flags
for compatibility with the most common dialects?

Nothing, if that's all there was to it. Check out the macro spaghetti at the top of various library code (standard library, boost, etc.) and see what
these poor guys have to put up with in order to get their code to compile across platforms and compilers. It's brutal.

I'm opposed to the preprocessor if there is any reasonable alternative but you do what you gotta do...

I see what you mean. So, the problem isn't so much with dialects of
C++, but with portability across platforms?

That and compilers, which have far more "personality" than the standard
allows. :) I guess that counts as dialects. There are other problems too,
like calling sequences, various non-standard declarations required to
compile DLL's, mumble mumble ...

I'm the wrong guy to ask though. My job doesn't really require me to get my
code to compile across a lot of platforms and compilers. It's guys like PJ
who bear the brunt of these problems.

--
Cy
http://home.rochester.rr.com/cyhome/