www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Re: Biggest problems w/ D - strings

reply C. Dunn <cdunn2001 gmail.com> writes:
Derek Parnell Wrote:
 You could try this simpler method ...

Oh. Slices! I didn't think of that. It should be just as efficient too, since these slices will not copy the contents. As for your follow-up, I thought that char abc[32]; would initialize abc to 32 zeroes automatically. Am I wrong?
Aug 13 2007
parent reply Oskar Linde <oskar.lindeREM OVEgmail.com> writes:
C. Dunn wrote:

 As for your follow-up, I thought that
   char abc[32];
 would initialize abc to 32 zeroes automatically.  Am I wrong?

chars are initialized to 0xff -- Oskar
Aug 13 2007
parent reply "Jarrett Billingsley" <kb3ctd2 yahoo.com> writes:
"Oskar Linde" <oskar.lindeREM OVEgmail.com> wrote in message 
news:f9qcug$2sgj$1 digitalmars.com...
 C. Dunn wrote:

 As for your follow-up, I thought that
   char abc[32];
 would initialize abc to 32 zeroes automatically.  Am I wrong?

chars are initialized to 0xff

There are a few ways to get that initialized to 0 in the struct, tho: 1) Just write the initializer in the struct. struct S { char[32] abc = [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0]; } Kind of ugly, but straightforward. 2) Typedef a zero-init char type. typedef char zchar = 0; struct S { zchar[32] abc; } You then might run into annoyances when converting between char and zchar.
Aug 13 2007
parent reply Deewiant <deewiant.doesnotlike.spam gmail.com> writes:
Jarrett Billingsley wrote:
 "Oskar Linde" <oskar.lindeREM OVEgmail.com> wrote in message 
 news:f9qcug$2sgj$1 digitalmars.com...
 C. Dunn wrote:

 As for your follow-up, I thought that
   char abc[32];
 would initialize abc to 32 zeroes automatically.  Am I wrong?


There are a few ways to get that initialized to 0 in the struct, tho: 1) Just write the initializer in the struct. struct S { char[32] abc = [0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0]; } Kind of ugly, but straightforward.

Ever since Issue 1268 was fixed in 1.017, you can just write: char[32] abc = 0; -- Remove ".doesnotlike.spam" from the mail address.
Aug 13 2007
parent reply "Jarrett Billingsley" <kb3ctd2 yahoo.com> writes:
"Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
news:f9rjbi$1sav$1 digitalmars.com...
 Ever since Issue 1268 was fixed in 1.017, you can just write:

 char[32] abc = 0;

I always wished that that were possible, but never knew if it was legal :D
Aug 14 2007
parent reply Chris Nicholson-Sauls <ibisbasenji gmail.com> writes:
Jarrett Billingsley wrote:
 "Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
 news:f9rjbi$1sav$1 digitalmars.com...
 Ever since Issue 1268 was fixed in 1.017, you can just write:

 char[32] abc = 0;

I always wished that that were possible, but never knew if it was legal :D

Oh it is, but there's no equivelant one-liner for dynamic arrays. Instead you have to: auto var = new int[len]; var[] = 42; Or using Cashew: import cashew.utils.Array; // ... auto var = repeat(42, len); Which is actually just a wrapper around the above example. It'd be nice if there were some way to specify an initializor for a dynamic array's elements in place, that didn't rely on using a typedef. typedef int _MyInt42 = 42; auto var = cast(int[]) new _MyInt42[len]; o_O -- Chris Nicholson-Sauls
Aug 14 2007
next sibling parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Chris Nicholson-Sauls wrote:
 Jarrett Billingsley wrote:
 "Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
 news:f9rjbi$1sav$1 digitalmars.com...
 Ever since Issue 1268 was fixed in 1.017, you can just write:

 char[32] abc = 0;

I always wished that that were possible, but never knew if it was legal :D

Oh it is, but there's no equivelant one-liner for dynamic arrays. Instead you have to: auto var = new int[len]; var[] = 42; Or using Cashew: import cashew.utils.Array; // ... auto var = repeat(42, len); Which is actually just a wrapper around the above example. It'd be nice if there were some way to specify an initializor for a dynamic array's elements in place, that didn't rely on using a typedef. typedef int _MyInt42 = 42; auto var = cast(int[]) new _MyInt42[len]; o_O

You gonna make that cashew available via dsss net any time soon? If not you mind if I do it? --bb
Aug 14 2007
parent reply Chris Nicholson-Sauls <ibisbasenji gmail.com> writes:
Bill Baxter wrote:
 Chris Nicholson-Sauls wrote:
 Jarrett Billingsley wrote:
 "Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
 news:f9rjbi$1sav$1 digitalmars.com...
 Ever since Issue 1268 was fixed in 1.017, you can just write:

 char[32] abc = 0;

I always wished that that were possible, but never knew if it was legal :D

Oh it is, but there's no equivelant one-liner for dynamic arrays. Instead you have to: auto var = new int[len]; var[] = 42; Or using Cashew: import cashew.utils.Array; // ... auto var = repeat(42, len); Which is actually just a wrapper around the above example. It'd be nice if there were some way to specify an initializor for a dynamic array's elements in place, that didn't rely on using a typedef. typedef int _MyInt42 = 42; auto var = cast(int[]) new _MyInt42[len]; o_O

You gonna make that cashew available via dsss net any time soon? If not you mind if I do it? --bb

Well... I admit I don't actually use DSSS myself. (The shame, I know.) So if you want to do so, go ahead. I have a long list of to-do items for Cashew... should really get to work on it. (Like adding those nifty binutils.) -- Chris Nicholson-Sauls
Aug 14 2007
parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Chris Nicholson-Sauls wrote:
 Bill Baxter wrote:
 Chris Nicholson-Sauls wrote:
 Jarrett Billingsley wrote:
 "Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
 news:f9rjbi$1sav$1 digitalmars.com...
 Ever since Issue 1268 was fixed in 1.017, you can just write:

 char[32] abc = 0;

I always wished that that were possible, but never knew if it was legal :D

Oh it is, but there's no equivelant one-liner for dynamic arrays. Instead you have to: auto var = new int[len]; var[] = 42; Or using Cashew: import cashew.utils.Array; // ... auto var = repeat(42, len); Which is actually just a wrapper around the above example. It'd be nice if there were some way to specify an initializor for a dynamic array's elements in place, that didn't rely on using a typedef. typedef int _MyInt42 = 42; auto var = cast(int[]) new _MyInt42[len]; o_O

You gonna make that cashew available via dsss net any time soon? If not you mind if I do it? --bb

Well... I admit I don't actually use DSSS myself. (The shame, I know.) So if you want to do so, go ahead. I have a long list of to-do items for Cashew... should really get to work on it. (Like adding those nifty binutils.) -- Chris Nicholson-Sauls

Well, I gave it a go, but cashew, seems to depend on mango. And "dsss net install mango" is failing, so I guess I give up for now. Here's the error: mango\net\util\cache\model\ICache.d(15): module IMessage cannot read file 'mango\net\cluster\model\IMessage.d' ICache.d has this: private import mango.net.cluster.model.IMessage; but there is no mango/net/cluster directory in the sources that dsss net is grabbing. --bb
Aug 14 2007
next sibling parent reply kris <foo bar.com> writes:
Bill Baxter wrote:
[snip]
 Well, I gave it a go, but cashew, seems to depend on mango.  And "dsss 
 net install mango" is failing, so I guess I give up for now.
 
 Here's the error:
 
 mango\net\util\cache\model\ICache.d(15): module IMessage cannot read 
 file 'mango\net\cluster\model\IMessage.d'
 
 ICache.d has this:
 private import mango.net.cluster.model.IMessage;

fixed (thank you)
Aug 14 2007
parent Bill Baxter <dnewsgroup billbaxter.com> writes:
kris wrote:
 Bill Baxter wrote:
 [snip]
 Well, I gave it a go, but cashew, seems to depend on mango.  And "dsss 
 net install mango" is failing, so I guess I give up for now.

 Here's the error:

 mango\net\util\cache\model\ICache.d(15): module IMessage cannot read 
 file 'mango\net\cluster\model\IMessage.d'

 ICache.d has this:
 private import mango.net.cluster.model.IMessage;

fixed (thank you)

Did you check it in too? dsss net install mango still fails with the same error. Here's the svn url its using: svn export http://svn.dsource.org/projects/mango/trunk --bb
Aug 14 2007
prev sibling parent reply Chris Nicholson-Sauls <ibisbasenji gmail.com> writes:
Bill Baxter wrote:
 Chris Nicholson-Sauls wrote:
 Bill Baxter wrote:
 Chris Nicholson-Sauls wrote:
 Jarrett Billingsley wrote:
 "Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
 news:f9rjbi$1sav$1 digitalmars.com...
 Ever since Issue 1268 was fixed in 1.017, you can just write:

 char[32] abc = 0;

I always wished that that were possible, but never knew if it was legal :D

Oh it is, but there's no equivelant one-liner for dynamic arrays. Instead you have to: auto var = new int[len]; var[] = 42; Or using Cashew: import cashew.utils.Array; // ... auto var = repeat(42, len); Which is actually just a wrapper around the above example. It'd be nice if there were some way to specify an initializor for a dynamic array's elements in place, that didn't rely on using a typedef. typedef int _MyInt42 = 42; auto var = cast(int[]) new _MyInt42[len]; o_O

You gonna make that cashew available via dsss net any time soon? If not you mind if I do it? --bb

Well... I admit I don't actually use DSSS myself. (The shame, I know.) So if you want to do so, go ahead. I have a long list of to-do items for Cashew... should really get to work on it. (Like adding those nifty binutils.) -- Chris Nicholson-Sauls

Well, I gave it a go, but cashew, seems to depend on mango. And "dsss net install mango" is failing, so I guess I give up for now. Here's the error: mango\net\util\cache\model\ICache.d(15): module IMessage cannot read file 'mango\net\cluster\model\IMessage.d' ICache.d has this: private import mango.net.cluster.model.IMessage; but there is no mango/net/cluster directory in the sources that dsss net is grabbing. --bb

Oh... my. (*dejected sigh*) No frets. The bits that are causing that problem are horridly out of date and mostly mildly incomplete. Much of Cashew will be getting re-written and/or updated soon. I'll let you know when the Mango dependencies have been cleaned out. A question: when requesting, for example, Tango from 'dsss net' does it grab the latest official release, or the current SVN snapshot? If the latter, I may just go ahead and try DSSS out and make Cashew ready myself. -- Chris Nicholson-Sauls
Aug 16 2007
parent reply Bill Baxter <dnewsgroup billbaxter.com> writes:
Chris Nicholson-Sauls wrote:
 Bill Baxter wrote:
 Chris Nicholson-Sauls wrote:
 Bill Baxter wrote:
 Chris Nicholson-Sauls wrote:
 Jarrett Billingsley wrote:
 "Deewiant" <deewiant.doesnotlike.spam gmail.com> wrote in message 
 news:f9rjbi$1sav$1 digitalmars.com...
 Ever since Issue 1268 was fixed in 1.017, you can just write:

 char[32] abc = 0;

I always wished that that were possible, but never knew if it was legal :D

Oh it is, but there's no equivelant one-liner for dynamic arrays. Instead you have to: auto var = new int[len]; var[] = 42; Or using Cashew: import cashew.utils.Array; // ... auto var = repeat(42, len); Which is actually just a wrapper around the above example. It'd be nice if there were some way to specify an initializor for a dynamic array's elements in place, that didn't rely on using a typedef. typedef int _MyInt42 = 42; auto var = cast(int[]) new _MyInt42[len]; o_O

You gonna make that cashew available via dsss net any time soon? If not you mind if I do it? --bb

Well... I admit I don't actually use DSSS myself. (The shame, I know.) So if you want to do so, go ahead. I have a long list of to-do items for Cashew... should really get to work on it. (Like adding those nifty binutils.) -- Chris Nicholson-Sauls

Well, I gave it a go, but cashew, seems to depend on mango. And "dsss net install mango" is failing, so I guess I give up for now. Here's the error: mango\net\util\cache\model\ICache.d(15): module IMessage cannot read file 'mango\net\cluster\model\IMessage.d' ICache.d has this: private import mango.net.cluster.model.IMessage; but there is no mango/net/cluster directory in the sources that dsss net is grabbing. --bb

Oh... my. (*dejected sigh*) No frets. The bits that are causing that problem are horridly out of date and mostly mildly incomplete. Much of Cashew will be getting re-written and/or updated soon. I'll let you know when the Mango dependencies have been cleaned out.

Sure. Well, currently my dsss.conf for cashew consists of exactly two lines: ---------dsss.conf--------- name = cashew [cashew] -------end dsss.conf------- And mango's is only slightly longer: ---------dsss.conf--------- name = mango [mango/icu] [mango/net] [mango/xml] ---------dsss.conf--------- If there are parts that are basically dead right now and shouldn't be included, then you can just add an 'exclude' line. I.e. if you just want to leave out the mango-depending parts just make it e.g.: ---------dsss.conf--------- name = cashew [cashew] exclude = cashew/utils/Benchmark.d -------end dsss.conf-------
 A question: when requesting, for example, Tango from 'dsss net' does it 
 grab the latest official release, or the current SVN snapshot?  If the 
 latter, I may just go ahead and try DSSS out and make Cashew ready myself.

I _think_ it grabs it from the latest SVN. And then optionally applies a dsss compatibility patch to it. If so then it should be pretty trivial to change dsss to grab a particular branch/tag from svn instead of the default branch. Maybe it already has this capability? I do think DSSS is going to have to confront the versioning issue soon, though. There needs to be a way to ask for a specific version when using net install. And maybe even a way to install more than one version of the same package. The other package manager I'm familiar with (Python setuptools/eggs) handle both. Basically instead of installing package foo into the "foo" directory, it installs it into a directory like "foo-{fooversion}-{pythonversion}.egg/foo", and then adds "foo-{fooversion}-{pythonversion}.egg" to a file that lists the current include path. I think a similar scheme could work for dsss. It lets you have more than one version of a package installed, but the order in that that master list of includes determines globally which version is active. --bb
Aug 16 2007
parent Chris Nicholson-Sauls <ibisbasenji gmail.com> writes:
Bill Baxter wrote:
 Chris Nicholson-Sauls wrote:

 Oh... my.  (*dejected sigh*)  No frets.  The bits that are causing 
 that problem are horridly out of date and mostly mildly incomplete.  
 Much of Cashew will be getting re-written and/or updated soon.  I'll 
 let you know when the Mango dependencies have been cleaned out.

Sure. Well, currently my dsss.conf for cashew consists of exactly two lines: ---------dsss.conf--------- name = cashew [cashew] -------end dsss.conf------- And mango's is only slightly longer: ---------dsss.conf--------- name = mango [mango/icu] [mango/net] [mango/xml] ---------dsss.conf--------- If there are parts that are basically dead right now and shouldn't be included, then you can just add an 'exclude' line. I.e. if you just want to leave out the mango-depending parts just make it e.g.: ---------dsss.conf--------- name = cashew [cashew] exclude = cashew/utils/Benchmark.d -------end dsss.conf-------
 A question: when requesting, for example, Tango from 'dsss net' does 
 it grab the latest official release, or the current SVN snapshot?  If 
 the latter, I may just go ahead and try DSSS out and make Cashew ready 
 myself.

I _think_ it grabs it from the latest SVN. And then optionally applies a dsss compatibility patch to it. If so then it should be pretty trivial to change dsss to grab a particular branch/tag from svn instead of the default branch. Maybe it already has this capability? I do think DSSS is going to have to confront the versioning issue soon, though. There needs to be a way to ask for a specific version when using net install. And maybe even a way to install more than one version of the same package. The other package manager I'm familiar with (Python setuptools/eggs) handle both. Basically instead of installing package foo into the "foo" directory, it installs it into a directory like "foo-{fooversion}-{pythonversion}.egg/foo", and then adds "foo-{fooversion}-{pythonversion}.egg" to a file that lists the current include path. I think a similar scheme could work for dsss. It lets you have more than one version of a package installed, but the order in that that master list of includes determines globally which version is active.

Ruby Gems also allows having multiple versions installed, although I've never looked into the details of how it keeps them. Gem also has 'update' for generically grabbing new versions of installed packages, and 'cleanup' for likewise uninstalling old versions. If and when DSSS adds support for multiple installations, something like these two commands would be doubtless useful. I've used them, for example, to test whether or not a new version of a library breaks anything before I commit to updating it. -- Chris Nicholson-Sauls
Aug 16 2007
prev sibling parent C. Dunn <cdunn2001 gmail.com> writes:
Chris Nicholson-Sauls Wrote:
 Ever since Issue 1268 was fixed in 1.017, you can just write:
 char[32] abc = 0;


auto var = new int[len]; var[] = 42;

If I cannot tell the "new int[]" expression what the initializer should be, then I would rather it not initialize at all. In the case above, it initializes twice!
Aug 14 2007