Problem with string.whitespace and newline
The following code cannot be compiled
string clean(string x)
{
return join(split(x),whitespace);
}
The compile error is
Error 2 Error: template std.array.join(RoR,R) if (isInputRange!(RoR)
isInputRange!(ElementType!(RoR)) isForwardRange!(R)) cannot
deduce template function from argument types
!()(string[],immutable(char[6u]))
However this line is fine
join(split(x), );
The same problem happens when using newline. Is this a bug?
Best,
Mike
Re: Problem with string.whitespace and newline
try return join(split(x),whitespace[]);
It seems whitespace is a static array.
On 6/12/11, Michael Chen sth4...@gmail.com wrote:
The following code cannot be compiled
string clean(string x)
{
return join(split(x),whitespace);
}
The compile error is
Error 2 Error: template std.array.join(RoR,R) if (isInputRange!(RoR)
isInputRange!(ElementType!(RoR)) isForwardRange!(R)) cannot
deduce template function from argument types
!()(string[],immutable(char[6u]))
However this line is fine
join(split(x), );
The same problem happens when using newline. Is this a bug?
Best,
Mike
Re: Problem with string.whitespace and newline
Thanks Andrej, it works.
On Sun, Jun 12, 2011 at 9:14 AM, Andrej Mitrovic
andrej.mitrov...@gmail.com wrote:
try return join(split(x),whitespace[]);
It seems whitespace is a static array.
On 6/12/11, Michael Chen sth4...@gmail.com wrote:
The following code cannot be compiled
string clean(string x)
{
return join(split(x),whitespace);
}
The compile error is
Error 2 Error: template std.array.join(RoR,R) if (isInputRange!(RoR)
isInputRange!(ElementType!(RoR)) isForwardRange!(R)) cannot
deduce template function from argument types
!()(string[],immutable(char[6u]))
However this line is fine
join(split(x), );
The same problem happens when using newline. Is this a bug?
Best,
Mike
Re: Problem with string.whitespace and newline
I've said it before, but I'll say it again: this is such a common error that it needs a better error message. Something as simple as this will do: foo can't take bar of type char[16]. Try using a slice: bar[].
Re: Problem with string.whitespace and newline
On 2011-06-11 18:46, Andrej Mitrovic wrote: I've said it before, but I'll say it again: this is such a common error that it needs a better error message. Something as simple as this will do: foo can't take bar of type char[16]. Try using a slice: bar[]. Have you created an enhancement request for it? If not, it's not likely to happen. Though honestly, this is a bit of a tough one, because all that the compiler really knows is that the argument that you gave to the template function didn't match any of that template function's overloads. It has no idea why. And maybe there's an overload which is _supposed_ to take a static array for some reason, and it's another argument to the function that you screwed up? There's probably some sort of special-casing that the compiler could do when trying to instantiating a template function and one of the arguments it was given was a static array, but the compiler is generally designed to tell that what you did doesn't work rather than tell you what you're supposed to do. So, getting error messages to give suggestions (particularly when templates are involved) can be tricky. It's been done before with the spellchecker, but I don't know how straightforward it would be to do that sort of thing with static arrays given that it has a much broader scope. Template errors are just generally bad. They're definitely better than in C++, but I still don't think that they're anywhere near good enough for a lot of the more average programmers. If you start messing with them a lot, then the template constraints are actually quite good at giving you a clue as to what's wrong, but for newbies in particular, they're pretty incomprehensible. - Jonathan M Davis
Re: Problem with string.whitespace and newline
I can't seem to find anything specific like an enhancement request. Here's a similar report by you: http://d.puremagic.com/issues/show_bug.cgi?id=4971 Maybe we should have a 'common (newbie) mistakes' section on the website and add this bit of information there.
Re: Problem with string.whitespace and newline
On 2011-06-11 19:29, Andrej Mitrovic wrote: I can't seem to find anything specific like an enhancement request. Here's a similar report by you: http://d.puremagic.com/issues/show_bug.cgi?id=4971 Maybe we should have a 'common (newbie) mistakes' section on the website and add this bit of information there. It certainly sounds like a good idea. And thanks to the fact that we're switching to d-programming-language.org, and its source is up on github, it's even possible for anyone to make the changes necessary and create a pull request for it. What I'd like to see templates be able to do though is better recognize a type that _will_ work in at least a few common cases. Static arrays are one (if all of the template overloads fail, perhaps it should try instiating them with the dynamic version of the same type). Another is immutable dynamic arrays. const(T)[] would work if you declared the template with those explicit parameters, but as long as it's taking R or Range instead of const(T)[] for arrays, a range-based function won't work with immutable arrays even though it should. But that sort of change could be too big a change to how templates work. I don't know. I should probably write up an enhancement request on the subject. - Jonathan M Davis
Re: Problem with string.whitespace and newline
On 2011-06-11 19:41, Jonathan M Davis wrote: On 2011-06-11 19:29, Andrej Mitrovic wrote: I can't seem to find anything specific like an enhancement request. Here's a similar report by you: http://d.puremagic.com/issues/show_bug.cgi?id=4971 Maybe we should have a 'common (newbie) mistakes' section on the website and add this bit of information there. It certainly sounds like a good idea. And thanks to the fact that we're switching to d-programming-language.org, and its source is up on github, it's even possible for anyone to make the changes necessary and create a pull request for it. What I'd like to see templates be able to do though is better recognize a type that _will_ work in at least a few common cases. Static arrays are one (if all of the template overloads fail, perhaps it should try instiating them with the dynamic version of the same type). Another is immutable dynamic arrays. const(T)[] would work if you declared the template with those explicit parameters, but as long as it's taking R or Range instead of const(T)[] for arrays, a range-based function won't work with immutable arrays even though it should. But that sort of change could be too big a change to how templates work. I don't know. I should probably write up an enhancement request on the subject. http://d.puremagic.com/issues/show_bug.cgi?id=6148
