Quick way to convert int to array of numbers

[John B]

Hi all,
Say I have the int 123456789
What would be the quickest/best way to convert it to
int[]{1,2,3,4,5,6,7,8,9}

What I came up with was to determine the largest factor of 10 (100M)
divide by that, truncate decimals and that'd be the 1st number, subtract
that number multiplied by the current factor, next smallest factor etc..

But it seems very roundabout, surely there'd be a better way :)

TIA

JB

[Paul E Collins]

"John B" <jb******@yahoo.comwrote:

Say I have the int 123456789
What would be the quickest/best way to convert it to
int[]{1,2,3,4,5,6,7,8,9}

I think this is the most readable way:

int[] x = new int[s.Length];
for (int i = 0; i < s.Length; i++)
{
x[i] = Convert.ToInt32(s[i]);
}

Unless this is really a performance bottleneck, there's no need to do
anything cleverer.

Eq.

[John B]

Paul E Collins wrote:

"John B" <jb******@yahoo.comwrote:

>Say I have the int 123456789
What would be the quickest/best way to convert it to
int[]{1,2,3,4,5,6,7,8,9}

I think this is the most readable way:

int[] x = new int[s.Length];
for (int i = 0; i < s.Length; i++)
{
x[i] = Convert.ToInt32(s[i]);
}

Unless this is really a performance bottleneck, there's no need to do
anything cleverer.

Yep, except its an actual long so I just convert with ToString to begin
with.
Its not at all a bottleneck but its very strange as I have had one
instance where I got an "Input string was not in a correct format" on
int.Parse

The actual code is

//value is a long
string sVal = value.ToString();
for (int i = sVal.Length - 1; i >= 0; i--)
{
int iVal = int.Parse(sVal[i].ToString()); //Here is where the
exception was raised.
//...Do some validation
}

So I thought it'd just be easier to not convert them to a string and
just treat them as int all the time.
Of course this happened at a client site and I cannot reproduce it here :)

Cheers

JB

[=?ISO-8859-1?Q?Arne_Vajh=F8j?=]

John B wrote:

Say I have the int 123456789
What would be the quickest/best way to convert it to
int[]{1,2,3,4,5,6,7,8,9}

What I came up with was to determine the largest factor of 10 (100M)
divide by that, truncate decimals and that'd be the 1st number, subtract
that number multiplied by the current factor, next smallest factor etc..

But it seems very roundabout, surely there'd be a better way :)

I would say that it is easier/more readable to do:

for(int i = 0; i < n; i++)
{
digs[n-i-1] = tmp % 10;
tmp /= 10;
}

Arne

[Paul E Collins]

"John B" <jb******@yahoo.comwrote:

Yep, except its an actual long so I just convert with ToString to
begin with.
Its not at all a bottleneck but its very strange as I have had one
instance where I got an "Input string was not in a correct format" on
int.Parse

You probably had "NaN" (zero divided by zero) or "Infinity" (anything
else divided by zero). You won't be able to convert those strings to an
integer.

Eq.

[Peter Duniho]

On Wed, 30 Apr 2008 19:31:26 -0700, Paul E Collins
<fi******************@CL4.orgwrote:

"John B" <jb******@yahoo.comwrote:

>Yep, except its an actual long so I just convert with ToString to
begin with.
Its not at all a bottleneck but its very strange as I have had one
instance where I got an "Input string was not in a correct format" on
int.Parse

You probably had "NaN" (zero divided by zero) or "Infinity" (anything
else divided by zero). You won't be able to convert those strings to an
integer.

Those should only apply to floating point values. Assuming he started
with an int or a long, that shouldn't have happened.

Taking as granted that it did happen, I guess that suggests that the
actual code was wrong. For sure, in theory code that starts with an int,
converts to a string, and parses each character individually to convert
back to an array of ints, should never fail.

Of course, without seeing his actual code that failed, it's not possible
to say what was actually wrong with it. Assuming the code he posted is in
fact the actual code, or at least reasonably close to it, I think it's
more likely that he had a negative number and the conversion failed on the
minus sign. That'd be easy enough to deal with, but of course it begs the
question as to what the actual result in his code should be.

Pete

[John B]

Peter Duniho wrote:

On Wed, 30 Apr 2008 19:31:26 -0700, Paul E Collins
<fi******************@CL4.orgwrote:

>"John B" <jb******@yahoo.comwrote:

>>Yep, except its an actual long so I just convert with ToString to
begin with.
Its not at all a bottleneck but its very strange as I have had one
instance where I got an "Input string was not in a correct format" on
int.Parse

You probably had "NaN" (zero divided by zero) or "Infinity" (anything
else divided by zero). You won't be able to convert those strings to an
integer.

Those should only apply to floating point values. Assuming he started
with an int or a long, that shouldn't have happened.

Taking as granted that it did happen, I guess that suggests that the
actual code was wrong. For sure, in theory code that starts with an
int, converts to a string, and parses each character individually to
convert back to an array of ints, should never fail.

Of course, without seeing his actual code that failed, it's not possible
to say what was actually wrong with it. Assuming the code he posted is
in fact the actual code, or at least reasonably close to it, I think

*****

it's more likely that he had a negative number and the conversion failed
on the minus sign. That'd be easy enough to deal with, but of course it

******
<....>
Thanks Peter, that must've been it.
I have now added a check to ensure that negative numbers are flagged as
invalid to prevent this from happening again.

The actual code is just for validating the IMEI of a phone.

Thanks,

JB

[Ben Voigt [C++ MVP]]

John B wrote:

Peter Duniho wrote:

>On Wed, 30 Apr 2008 19:31:26 -0700, Paul E Collins
<fi******************@CL4.orgwrote:

>>"John B" <jb******@yahoo.comwrote:

Yep, except its an actual long so I just convert with ToString to
begin with.
Its not at all a bottleneck but its very strange as I have had one
instance where I got an "Input string was not in a correct format"
on int.Parse

You probably had "NaN" (zero divided by zero) or "Infinity"
(anything else divided by zero). You won't be able to convert those
strings to an integer.

Those should only apply to floating point values. Assuming he
started with an int or a long, that shouldn't have happened.

Taking as granted that it did happen, I guess that suggests that the
actual code was wrong. For sure, in theory code that starts with an
int, converts to a string, and parses each character individually to
convert back to an array of ints, should never fail.

Of course, without seeing his actual code that failed, it's not
possible to say what was actually wrong with it. Assuming the code
he posted is in fact the actual code, or at least reasonably close
to it, I think ***** it's more likely that he had a negative number
and the conversion failed on the minus sign. That'd be easy enough
to deal with, but of course it

******
<....>
Thanks Peter, that must've been it.
I have now added a check to ensure that negative numbers are flagged
as invalid to prevent this from happening again.

The actual code is just for validating the IMEI of a phone.

That's probably long enough to overflow int.

>
Thanks,

JB

[Ben Voigt [C++ MVP]]

John B wrote:

Paul E Collins wrote:

>"John B" <jb******@yahoo.comwrote:

>>Say I have the int 123456789
What would be the quickest/best way to convert it to
int[]{1,2,3,4,5,6,7,8,9}

I think this is the most readable way:

int[] x = new int[s.Length];
for (int i = 0; i < s.Length; i++)
{
x[i] = Convert.ToInt32(s[i]);
}

Unless this is really a performance bottleneck, there's no need to do
anything cleverer.

Yep, except its an actual long so I just convert with ToString to
begin with.
Its not at all a bottleneck but its very strange as I have had one
instance where I got an "Input string was not in a correct format" on
int.Parse

Skip the Parse step, it's actually easier to convert the raw characters than
the integer.

int[] x = new int[s.Length];
int i = 0;
foreach (char c in s)
x[i++] = c - '0';

[Peter Duniho]

On Thu, 01 May 2008 06:37:31 -0700, Ben Voigt [C++ MVP]
<rb*@nospam.nospamwrote:

[...]

>I have now added a check to ensure that negative numbers are flagged
as invalid to prevent this from happening again.

The actual code is just for validating the IMEI of a phone.

That's probably long enough to overflow int.

He did write, in a more recent post, that his input was typed as long.
Though, he could be mistaken and somewhere along the line it's actually an
int. That would certainly explain why the number wound up negative at
some point.

Pete