www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.learn - next!T

reply Andrew Edwards <ridimz yahoo.com> writes:
Hello All,

I wrote the following convenience functions to aid in my studies. 
Unfortunately, I'm using Java books (Ali I will get to yours soon 
enough) so the need was necessitated by the frequency of use in the 
examples. Would appreciate a sanity check. Is there something that I 
should be thinking about am obviously not? Comments/criticisms appreciated.

------------ io.d ------------
module io;

public import std.stdio;

private string buffer;

auto next(T)()
{
	import std.traits;
	import std.conv;
	import std.string;

	if(buffer.length == 0)
		buffer = stdin.readln;

	static if (isSomeString!T) {
		scope (exit) buffer = null;
		return buffer.strip;
	}
	else {
		scope (exit) buffer = buffer.strip;
		return parse!T(buffer);
	}
}

auto next(T)(string msg)
{
	if (msg != null)
		msg.write;

	return next!T;
}
------------ End ------------

Thanks,
Andrew
May 25 2014
parent reply =?UTF-8?B?QWxpIMOHZWhyZWxp?= <acehreli yahoo.com> writes:
On 05/25/2014 05:21 PM, Andrew Edwards wrote:

 Hello All,

 I wrote the following convenience functions to aid in my studies.
 Unfortunately, I'm using Java books (Ali I will get to yours soon
 enough) so the need was necessitated by the frequency of use in the
 examples. Would appreciate a sanity check. Is there something that I
 should be thinking about am obviously not? Comments/criticisms appreciated.

 ------------ io.d ------------
 module io;

 public import std.stdio;

 private string buffer;

 auto next(T)()
 {
      import std.traits;
      import std.conv;
      import std.string;

      if(buffer.length == 0)
          buffer = stdin.readln;

      static if (isSomeString!T) {
          scope (exit) buffer = null;
          return buffer.strip;
      }
      else {
          scope (exit) buffer = buffer.strip;
          return parse!T(buffer);
      }
 }

 auto next(T)(string msg)
 {
      if (msg != null)
          msg.write;

      return next!T;
 }
 ------------ End ------------

 Thanks,
 Andrew


Works like a charm here! :)

void main()
{
     auto s = next!string("What is your name? ");
     auto i = next!uint("Your age? ");
     auto a = next!string("Your address? ");
     auto arr = next!(string[])("Names of your children? ");

     writefln("%s year old %s lives in %s. Children: %s", i, s, a, arr);
}

Ali
May 25 2014
next sibling parent Andrew Edwards <ridimz yahoo.com> writes:
On 5/25/14, 10:12 PM, Ali Çehreli wrote:

 On 05/25/2014 05:21 PM, Andrew Edwards wrote:

 Hello All,

 I wrote the following convenience functions to aid in my studies.
 Unfortunately, I'm using Java books (Ali I will get to yours soon
 enough) so the need was necessitated by the frequency of use in the
 examples. Would appreciate a sanity check. Is there something that I
 should be thinking about am obviously not? Comments/criticisms
 appreciated.





 Works like a charm here! :)


Thanks Ali.
May 26 2014
prev sibling parent reply Andrew Edwards <ridimz yahoo.com> writes:
On 5/25/14, 10:12 PM, Ali Çehreli wrote:

 On 05/25/2014 05:21 PM, Andrew Edwards wrote:

 Hello All,

 I wrote the following convenience functions to aid in my studies.
 Unfortunately, I'm using Java books (Ali I will get to yours soon
 enough) so the need was necessitated by the frequency of use in the
 examples. Would appreciate a sanity check. Is there something that I
 should be thinking about am obviously not? Comments/criticisms
 appreciated.

 ------------ io.d ------------
 module io;

 public import std.stdio;

 private string buffer;

 auto next(T)()
 {
      import std.traits;
      import std.conv;
      import std.string;

      if(buffer.length == 0)
          buffer = stdin.readln;

      static if (isSomeString!T) {
          scope (exit) buffer = null;
          return buffer.strip;
      }
      else {
          scope (exit) buffer = buffer.strip;
          return parse!T(buffer);
      }
 }

 auto next(T)(string msg)
 {
      if (msg != null)
          msg.write;

      return next!T;
 }
 ------------ End ------------

 Thanks,
 Andrew


 Works like a charm here! :)

 void main()
 {
      auto s = next!string("What is your name? ");
      auto i = next!uint("Your age? ");
      auto a = next!string("Your address? ");
      auto arr = next!(string[])("Names of your children? ");

      writefln("%s year old %s lives in %s. Children: %s", i, s, a, arr);
 }

 Ali


So one of the things I really wanted to do is Unicode IO (namely 
Japanese) from the console. I tried using this and quickly realized that 
no considerations were taken. The modified version next(T)() now looks 
like this:

auto next(T)()
{
	import std.traits;
	import std.conv;
	import std.string;
	import std.range;

	if(buffer.length == 0)
		buffer = stdin.readln;

	static if (isSomeString!T) {                     // [1]
		scope (exit) buffer = null;
		return buffer.strip.to!T;
	}
	static if (isSomeChar!T) {
		scope (exit) { buffer.popFront; buffer = buffer.strip; }
		return stride(buffer, 1).array[0].to!T;
	}
	else {
		scope (exit) buffer = buffer.strip;
		return parse!T(buffer);                   // [2] Line 64
	}
}

This function now works for all types except dstring. This remains a 
problem I cannot figure out. The error code is as follows:

$ rdmd -unittest textnext
/usr/share/dmd/src/phobos/std/conv.d(3293): Error: cannot modify 
immutable expression c
/usr/share/dmd/src/phobos/std/conv.d(3297): Error: cannot modify 
immutable expression c
/usr/share/dmd/src/phobos/std/conv.d(2904): Error: template instance 
std.conv.parseElement!(immutable(dchar), string) error instantiating
io.d(64):        instantiated from here: parse!(immutable(dchar)[], string)
textnext.d(12):        instantiated from here: next!(immutable(dchar)[])
io.d(64): Error: template instance std.conv.parse!(immutable(dchar)[], 
string) error instantiating
textnext.d(12):        instantiated from here: next!(immutable(dchar)[])
textnext.d(12): Error: template instance io.next!(immutable(dchar)[]) 
error instantiating
Failed: ["dmd", "-unittest", "-v", "-o-", "textnext.d", "-I."]

This error is report failure on dstring at [2] but I fully expected them 
at [1].

I've looked at implementation of isSomeString() 
(https://github.com/D-Programming-Language/phobos/blob/master
std/traits.d#L5178) 
and clearly dstrings are addressed in the implementation there. What 
exactly am I missing in my understanding?

Any assistance/advice is appreciated.

Thanks,
Andrew
Jun 06 2014
parent reply "Chris Nicholson-Sauls" <ibisbasenji gmail.com> writes:
On Saturday, 7 June 2014 at 02:23:18 UTC, Andrew Edwards wrote:

 This function now works for all types except dstring. This 
 remains a problem I cannot figure out. The error code is as 
 follows:

 $ rdmd -unittest textnext
 /usr/share/dmd/src/phobos/std/conv.d(3293): Error: cannot 
 modify immutable expression c
 /usr/share/dmd/src/phobos/std/conv.d(3297): Error: cannot 
 modify immutable expression c
 /usr/share/dmd/src/phobos/std/conv.d(2904): Error: template 
 instance std.conv.parseElement!(immutable(dchar), string) error 
 instantiating
 io.d(64):        instantiated from here: 
 parse!(immutable(dchar)[], string)
 textnext.d(12):        instantiated from here: 
 next!(immutable(dchar)[])
 io.d(64): Error: template instance 
 std.conv.parse!(immutable(dchar)[], string) error instantiating
 textnext.d(12):        instantiated from here: 
 next!(immutable(dchar)[])
 textnext.d(12): Error: template instance 
 io.next!(immutable(dchar)[]) error instantiating
 Failed: ["dmd", "-unittest", "-v", "-o-", "textnext.d", "-I."]

 This error is report failure on dstring at [2] but I fully 
 expected them at [1].

 I've looked at implementation of isSomeString() 
 (https://github.com/D-Programming-Language/phobos/blob/master
std/traits.d#L5178) 
 and clearly dstrings are addressed in the implementation there. 
 What exactly am I missing in my understanding?

 Any assistance/advice is appreciated.

 Thanks,
 Andrew



I got it working here by putting an else before the second 
static-if.
Jun 06 2014
parent reply Andrew Edwards <ridimz yahoo.com> writes:
On 6/6/14, 10:57 PM, Chris Nicholson-Sauls wrote:

 On Saturday, 7 June 2014 at 02:23:18 UTC, Andrew Edwards wrote:

 Any assistance/advice is appreciated.

 Thanks,
 Andrew



 I got it working here by putting an else before the second static-if.


Wow. Sometimes you really cannot see the things you type no matter how 
long you stare at it. Thank you soo much.
Jun 06 2014
parent "Chris Nicholson-Sauls" <ibisbasenji gmail.com> writes:
On Saturday, 7 June 2014 at 03:21:49 UTC, Andrew Edwards wrote:

 Wow. Sometimes you really cannot see the things you type no 
 matter how long you stare at it. Thank you soo much.



No problem.  I only noticed when I re-typed it by hand to study 
the flow, and instinctively added the else out of habit and 
wondered why I was getting no errors.  :)  Started thinking it 
was one of those "this is an error, but only on Thursdays" kind 
of things.

Anyway, since the bugged code was freaking out in the second, 
unreachable, return statement, it leads me to thinking whether we 
could provide some sane way for the compiler to point out such 
things (unreachable returns).  It would have made this case, and 
ones like it, more obvious.
Jun 06 2014