A question,

shouldn't be documented on the 'std.string.toString(char*)', that the d-string returned shares the same buffer as the input c-string?

I've spent some (too much) time debugging to find out why the heck all my strings were messing up... now I know.

Mentioning this on the docs could have been of so much help, it's not that obvious!

Regards,
--
Tom;

Tom wrote:
> A question,
> 
> shouldn't be documented on the 'std.string.toString(char*)', that the d-string returned shares the same buffer as the input c-string?
> 
> I've spent some (too much) time debugging to find out why the heck all my strings were messing up... now I know.
> 
> Mentioning this on the docs could have been of so much help, it's not that obvious!

I think it's pretty obvious if you read the first paragraph of http://www.digitalmars.com/d/phobos/std_string.html, and consider that a char* c-string is also a 'string' as described in that bit of text. It's not modified, so a slice of the original is returned (as a D char[]).

Remember: don't just read the function description, the module description may contain information that applies to all (or a lot of) the functions in a module.

Frits van Bommel escribió:
> Tom wrote:
>> A question,
>>
>> shouldn't be documented on the 'std.string.toString(char*)', that the d-string returned shares the same buffer as the input c-string?
>>
>> I've spent some (too much) time debugging to find out why the heck all my strings were messing up... now I know.
>>
>> Mentioning this on the docs could have been of so much help, it's not that obvious!
> 
> I think it's pretty obvious if you read the first paragraph of http://www.digitalmars.com/d/phobos/std_string.html, and consider that a char* c-string is also a 'string' as described in that bit of text. It's not modified, so a slice of the original is returned (as a D char[]).

It wasn't so obvious for me. I knew the COW thing for char[] but didn't guess that a char* was considered a string in this matters.

> Remember: don't just read the function description, the module description may contain information that applies to all (or a lot of) the functions in a module.

Uhmm, I know :). Not reading wasn't the issue. It was disregarding the fact that in D, a char* is also considered a string.

Regards,
--
Tom;

Tom wrote:
> Frits van Bommel escribió:
>> Tom wrote:
>>> shouldn't be documented on the 'std.string.toString(char*)', that the d-string returned shares the same buffer as the input c-string?
[snip]
>>
>> I think it's pretty obvious if you read the first paragraph of http://www.digitalmars.com/d/phobos/std_string.html, and consider that a char* c-string is also a 'string' as described in that bit of text. It's not modified, so a slice of the original is returned (as a D char[]).
> 
> It wasn't so obvious for me. I knew the COW thing for char[] but didn't guess that a char* was considered a string in this matters.
[snip]
> Not reading wasn't the issue. It was disregarding the fact that in D, a char* is also considered a string.

Well, char* isn't typically considered a string in D, but in this case[1] it's treated as one[2].
Also, you clearly referred to it as a (c-)string in your original post ;).


[1]: And AFAIK only in this case, outside of *.c.* modules. Well, perhaps you could also count the return value of toStringz...

[2]: The description of toString(char*):  "Convert C-style 0 terminated _string_ s to char[] string." (emphasis added)