good to have typedef?

D

doublemaster007

Is it good to typedef the types?? What is the actual use of it? I am
working on project where tooo many typedefs are used..If many people
are working on a project..only the person who defined will be
remembering it..finally too many typedefs will be there and code
becomes hard to use...

So i wanna know..is typedefs are really useful in full fledged
project??
 
J

Johannes Schaub (litb)

Is it good to typedef the types?? What is the actual use of it? I am
working on project where tooo many typedefs are used..If many people
are working on a project..only the person who defined will be
remembering it..finally too many typedefs will be there and code
becomes hard to use...
Sometimes, you can make better use of identity than of typedef. For example
to make a pointer to an array, and you want to use a typedef, you can
instead use identity

identity<int[5]>::type *ptr;

I've seen that identity has lots of nice uses, where otherwise a typedef
would have to be used to make things easy to read. The above would look like
the below with typedef

typedef int array_type[5];
array_type *ptr;

And i think using identity, this looks and feels better.
 
R

Richard

[Please do not mail me a copy of your followup]

"(e-mail address removed)" <[email protected]> spake the secret code
So i wanna know..is typedefs are really useful in full fledged
project??

Like anything in your code, you should use them if they add value and
avoid them if they are not carrying their weight.

Its possible to overuse typedefs to the point where they fail to
provide value.

For instance, I personally don't find value in typedefs like this:

typedef std::string *stringPtr;

Particularly since using that typedef leads you down the problem of
improperly specifying pointers to const strings. People think its
this:

const stringPtr unchangedStringThroughPointer;

Unfortunately you didn't specify the string was const, but that the
pointer was const. That's why Windows headers have the typedefs LPSTR
and LPCSTR, with the former being pointer to char and the latter being
pointer to const char. (The LP business is just an anachronism from
Win16 days.) Is LPSTR really more clear than char *? Is LPCSTR
really more clear than char const *? I don't think so, but I'll use
those typedefs because they are pervasive in Windows and Windows
programmers are familiar enough with them to know what they mean. I
stop short of using things like LPDIRECT3DDEVICE9 instead of
IDirect3DDevice9 *. The former just feels like screaming to me, while
the latter is easier to read and is just as clear (in context) that
we're talking about a COM interface pointer.

I'm currently working on a code base that is replete with typedefs
like these:

typedef unsigned char uint8;
typedef int int32;
typedef unsigned int uint32;
typedef float real;
typedef double lreal;

and so-on.

Personally I don't like declaring my types as having a bit size unless
it is truly relevant to their role. Once *everything* is declared as
an int32, how do you know which ones *really* need 32-bits and which
ones are comfortable in the native size of int? If you really need to
guarantee that an int is at least 32-bits you can do this in other
ways than polluting the entire source base with spurious int32's
everywhere.

The real/lreal thing is particularly annoying. It just obfuscates the
code in an attemp to abstract away a coupling to an implementation
dependency, yet it fails to do so in any meaningful way because it is
just an alias for an implementation dependency instead of an
abstraction like a new type (i.e. a class) would do. Because the
built-in types like int, float and double can't be extended or
customized, the typedef doesn't carry its own weight. It obscures the
fact that you *are* coupled to the implementation details of a float
or a double, while at the same time pretending to decouple you from
those details. It adds no value and reduces the clarity of the code.

The one area where typedefs are most useful is in consuming template
classes. std::string::find returns std::string::size_type; if I'm
using that type repeatedly in a function its easier to create a local
typedef for it:

void wipeOutSlashes(std::string &text)
{
typedef std::string::size_type size_type;

// declare locals of type size_type
}

That's just a simple example. It gets more useful when declaring
iterators and other types related to a template class. If you are
writing template libraries, creating nested typedefs that refer to
related types in your base class or in the template argument are very
useful for making the template expansion more readable.
 
S

Saeed Amrollahi

Is it good to typedef the types?? What is the actual use of it? I am
working on project where tooo many typedefs are used..If many people
are working on a project..only the person who defined will be
remembering it..finally too many typedefs will be there and code
becomes hard to use...

So i wanna know..is typedefs are really useful in full fledged
project??

Hi

Just my two cents ...
In addition to what Richard wrote, sometimes typedef declaration is
almost necessary.
For example, in C++ standard library, there are a lot of typedef
declarations
inside declaration/defintion of containers. A typical typedef is a
another
name for template parameter:
template<class T> // the bare bones
class std::list {
// ...
typedef T value_type;
// ...
};
So, we can use the template parameter:
void f(list<T>& L)
{
typename list<T>::value_type v = *v.begin();
// ...
}
without such typedef we can't use template parameter in generic
manner.
Also the keyword "typename" in above line is necessary.

Regards,
-- Saeed Amrollahi
 
J

Jorgen Grahn

Is it good to typedef the types?? What is the actual use of it? I am
working on project where tooo many typedefs are used..If many people
are working on a project..only the person who defined will be
remembering it..finally too many typedefs will be there and code
becomes hard to use...

I have to say that doesn't make sense to me. It's hard to see what
the problem is with your code.
So i wanna know..is typedefs are really useful in full fledged
project??

I agree with "Richard" elsewhere in this thread. And also:

- Most people, I think, dislike when a typedef is used to try to hide
that something is a struct or a pointer. Like, having to type FooPtr
instead of Foo*. I think Richard mentioned why that is not only
annoying but also harmful.

- Projects often overuse CHAR, UCHAR, uint32, etc ... typedefs.
Not a plain char or int in sight. I hate that.

- Typedefs are just aliases; they don't buy you type safety.
You might have
typedef std::string PersonName;
typedef std::string Gender;
and still mix up the arguments to
void DatingService::add(const PersonName&, const Gender&);
In other words, that prototype adds *no* value compared to
one where the arguments are const std::string&.

Is your problem one of these?

/Jorgen
 
J

James Kanze

Jorgen Grahn wrote:
We've got an awful lot of uint32_t in our project. That's
because it represents a data item which, because of published
specifications, is 32 bits, no more, no less.

I'm curious. What published specifications impose anything
concerning the internal representation? (Externally, of course,
the published specifications probably impose more than just 32
bits. Like byte order, for example.)
int is a word size which is convenient for the machine - good
for loopcounts, but not for precise data mappings.

That's not it's role. When you need to input or output data in
a precise format, you write code to do so: int32_t and others
aren't directly applicable here. (Indirectly... if you're
willing to restrict your portability to machines where uint32_t
is available---and that's still a lot of them---, then using it
for certain intermediate representations can make your code
significantly simpler.)
 
N

Nick Keighley

get a decent source browser. I agree typedefs can be over-used but the
they can be under-used as well.

- Most people, I think, dislike when a typedef is used to try to hide
  that something is a struct or a pointer.

I've no objection to hiding a struct behind a typedef (or at least in
C becasue I hate scattering the "struct" keyword all over the place
(I'm aware C++ doesn't have this problem)). I'm almost with you on
pointers except I justify hungarian for smart pointers. This is one of
the few places I'd advocate hungarian because I /want/ my nose rubbed
in what-type-of-pointer is this to make sure I cleanup properly.

Like, having to type FooPtr
  instead of Foo*. I think Richard mentioned why that is not only
  annoying but also harmful.

oh, I agree. Then there's hiding arrays (but C++ programmers don't use
arrays :) )

- Projects often overuse CHAR, UCHAR, uint32, etc ... typedefs.
  Not a plain char or int in sight. I hate that.

yuk. I've worked on projects where the programmming standards said you
must not use raw types. And the configuration control system
checked. :-(

Lovely bug:-

Station_id read_station_id (FILE *in)
{
Station_id station_id;
fscanf (in, "%d", &station_id);
return station_id;
}

Station_id was a short.

- Typedefs are just aliases; they don't buy you type safety.

too true. Possibly one of C's design mistakes.
  You might have
     typedef std::string PersonName;
     typedef std::string Gender;
  and still mix up the arguments to
     void DatingService::add(const PersonName&, const Gender&);
  In other words, that prototype adds *no* value compared to
  one where the arguments are const std::string&.

it documents intent. I'd maybe write

void DatingService::add(const std::string &personName&, std::const
string& gender);
 
J

James Kanze

James Kanze wrote:
You've answered your own question - it's a data transfer
specification.

In which case, uint32_t doesn't necessarily buy you that much.
(It does restrict portability, however, since it is guaranteed
not to be available on some machines.)
 
I

Ian Collins

Andy said:
Luckily for _me_ I only need to worry about Windoze boxes - albeit 64
bit is likely to come up soon. And I'm also lucky that my byte order
matches the spec. - not everyone does the ISO 9660 trick of storing a
number in both common orders, it's not space efficient!

But what should I use if I want to represent a guaranteed-to-be-32-bit
quantity?

uint32_t if the system has it. If not, define your own.
 
J

James Kanze

uint32_t if the system has it. If not, define your own.

If he only has to worry about Windows, the system has it. On
systems which don't have it, however, defining your own might be
a bit difficult, and the type will probably be significantly
slower than one of the built-in types. (On at least one system,
the built-in unsigned types are significantly slower than the
the signed ones, because the hardware doesn't directly support
them.)
 
G

Gerhard Fiedler

James said:
If he only has to worry about Windows, the system has it. On systems
which don't have it, however, defining your own might be a bit
difficult, and the type will probably be significantly slower than
one of the built-in types. (On at least one system, the built-in
unsigned types are significantly slower than the the signed ones,
because the hardware doesn't directly support them.)

I'm curious... what system doesn't "have" a uint32_t? I work also on
small 8bit micros (in C, but that's not really the issue here), and if I
really need it, I can use a uint32_t. It's not a native word size on
those systems, but the compiler couldn't care less about that. (Well,
the compiler writer of course does, and I do, too, where performance and
resources are an issue -- but the system "has it".)

Gerhard
 
G

Gerhard Fiedler

Nick said:
Lovely bug:-

Station_id read_station_id (FILE *in)
{
Station_id station_id;
fscanf (in, "%d", &station_id);
return station_id;
}

Station_id was a short.

You should get a compiler or linter that checks for such things -- or
use C++ streams in the first place :)

I found a few of these while porting a project from VC++ to gcc. gcc has
an option to check printf-type format strings.

Gerhard
 
B

Bo Persson

Gerhard said:
I'm curious... what system doesn't "have" a uint32_t? I work also on
small 8bit micros (in C, but that's not really the issue here), and
if I really need it, I can use a uint32_t. It's not a native word
size on those systems, but the compiler couldn't care less about
that. (Well, the compiler writer of course does, and I do, too,
where performance and resources are an issue -- but the system "has
it".)

Systems like these, that are based on 36-bit hardware

http://www.unisys.com/clearpath


Requiring them to have a uint32_t would be REALLY expensive. To top it
off, these machines also have 9-bit bytes, one's complement
arithmetic, and 72-bit non-IEEE floating point. :)

Unlike Java, C++ could actually be implemented.


Bo Persson
 
J

James Kanze

I'm curious... what system doesn't "have" a uint32_t?

Neither of the Unisys mainframe architectures: one's 36 bit ones
complement, the other 48 bit signed magnitude (with 8 bits
reserved).
I work also on small 8bit micros (in C, but that's not really
the issue here), and if I really need it, I can use a
uint32_t. It's not a native word size on those systems, but
the compiler couldn't care less about that. (Well, the
compiler writer of course does, and I do, too, where
performance and resources are an issue -- but the system "has
it".)

The compiler has to provide 32 bit arithmetic somehow, and if
the machine is 8 bits, using 32 bits for long, rather than more,
is the preferred solution. The problems are more with larger
machines, where the machine word size is 36 or 48 bits (or in
the past, 60 bits)---at one time, 36 bits was quite common.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

Forum statistics

Threads
473,744
Messages
2,569,484
Members
44,904
Latest member
HealthyVisionsCBDPrice

Latest Threads

Top