Hughmoris posted:

Since the OP hasn't been updated in 3 years, are there any new recommended books for people looking to pick up C# as a first language?

Ithaqua posted:

Check this out:
http://msdn.microsoft.com/en-us/library/hh304369(v=vs.100).aspx

Bozart posted:

Well that's sure cool as hell and answers my second question (use Array2d for matrix multiplication) but how do a take a list of lists as outlined before and get the item in the same position out of the inner list and put it into another list?

Knyteguy posted:

Ithaqua posted:

That book looks like it's terribly organized. How can you spend 10 chapters talking about Winforms GUI stuff before introducing the concept of variables?

I wish Jon Skeet would get around to writing his book for beginners, because then there would be a book that I could wholeheartedly recommend.

Ithaqua posted:

Yes. Do not submit this code to a potential employer as it is.

Here's my short list, without digging into the actual functionality:

• You're not separating your concerns. Presentation logic and application logic should not be mixed. This is a huge problem -- it shows that either you lack a fundamental understanding of object oriented programming (which is bad), or that you don't give a poo poo about doing it right (which is worse).
  
• Where are your unit tests? How do you know your application functions properly? How do you know that you're not breaking existing functionality when making changes? Having unit tests earns you points with a lot of employers. It certainly does with me!
  
• You do not follow Microsoft's naming conventions. I'm a stickler for this one, especially in sample code. Other people will tell you it doesn't matter, but my position is that Microsoft took the trouble to create a style guide, it makes a lot of sense, and there are lots of tools that can help enforce it. If you have a good reason for not following the guide and can articulate it and convince me, that's fine. If you don't have a good reason, then I chalk it up to sloppiness and/or not giving a poo poo.
  

Question: Does your application even update the progress bar visually? You're blocking the UI thread as far as I can tell.

Knyteguy posted:

So after re-reading this I decided I'm probably just going to rewrite most of my inventory program tomorrow. I've never actually done unit tests, so I'm going to try a test-driven design with multi-threading and see how it goes.

I'm still not very familiar with OOP principals (I'm not formally taught and usually just try different things until they work), do you or does anyone else have a good book I could read on the subject? I feel a bit over my head with best practices regarding this.

Anyway here's another sample of code that I'm hoping is good as-is.

It's not the most complicated thing, but if anyone would mind giving just a quick critique that would be great (I won't ask again for awhile)

E: had to remove html special codes

Ithaqua posted:

* The code itself is meaningless. You're providing no background on why this code exists. What problem were you solving? Why did you approach it that way? Did you try any other approaches?

* The method name sucks. It's named "parseLinks", but it's doing a hell of a lot more than that. It probably should be two (or more) methods, each descriptively named. Don't be afraid to make really long method names, as long as it's descriptive.

* Why are you creating a new variable here?

code:

if (!url.StartsWith("http"))
{
	url = "http://REMOVED FOR EMPLOYER PRIVACY.com" + url;
}
string formattedUrl = url;

Just keep using url.

* Why does the method always return null? That makes no sense. If it doesn't return anything, it should be void.

* That for loop is looking pretty suspect. First off, be careful about using the Count() method -- it has to iterate over the entire collection to get that count. List<T> has a Count property, which you should be using in this case.

* This code makes no sense whatsoever:

code:

// Make sure we don't follow this link later
hrefTags[i] = null;

* This code also makes no sense:

code:

if (!hrefTags[i].StartsWith("?") && hrefTags[i] != "/")
{
    // this is a link, add this to list so we can check it later for more images
    hrefTags[i] = formattedUrl;
}

What do you mean "check it later"? There is no later in this method. It's pointless.

Ithaqua posted:

* That for loop is looking pretty suspect. First off, be careful about using the Count() method -- it has to iterate over the entire collection to get that count. List<T> has a Count property, which you should be using in this case.

dwazegek posted:

This isn't entirely true. The IEnumerable.Count() extension method checks if the IEnumerable implements ICollection, and if it does, then it just returns the value of the ICollection.Count property. I'm not entirely sure, but I think it does the same check to see if it is an array type and then returns the Length. If it can't do either of those things, then it will enumerate the entire collection.

Ithaqua posted:

That makes sense. I checked the code for Count() and you're right, it does check if it's an ICollection and return the Count property. It doesn't do it for arrays, though.

mortarr posted:

Has anyone built custom reports over TFS / TFS Express - I'm looking to find out stuff like:
- How many solutions we have
- What version of VS was used to create the solution (eg. VS05, VS08 BIDS, VS10)
- How many projects the solutions have
- What .net frameworks those projects are addressing (eg. 2, 3.5, 4, 4.5)

I'm doing some investigation into upgrading my teams development environment/toolset to VS12 / some kind of source control that isn't sourcesafe, and trying to get a handle on the size of the problem... I've got a copy of the VSS db moved into TFS express to play around with, but I'm struggling to make any headway on the analysis. If TFS express is no good, I think there is a 90-day eval option for TFS full which I can use.

If there's an SQL query, custom reports, or code anyone has that answer those questions I'm keen to hear about it.

Knyteguy posted:

Anyway here's another sample of code that I'm hoping is good as-is.

Bozart posted:

I've got it stuck in my head to learn how to do matrix multiplication in a few different languages. I'm having some trouble with F# (which I am also learning):

code:

#light
module FSharpBlotter.MatrixMult
   // define two matrixes for multiplication
   let A = [[1; 2; 3]; 
            [3; 2; 1]]
   
   let B = [[1; 2]; 
            [3; 4]; 
            [5; 6]]
   
   // function to pull out a specific row and return as a list of int
   let Row (lst:_ list) (i: int) = 
      lst.[i]

   // function to pull out a specific column and return as a list of int
   let Col (lst:(int list) list) (i: int) = 
      [for R in lst do R.[i]]
      
   
   let A1 = A.[1]
   let A2 = Row A 1
   
   //error:
   //stdin(129,8): error FS0030: Value restriction. The value 'A3' has been inferred to have generic type
   let A3 = Col A 1
   
   // and then:
   // dot product function for two lists of ints
   
   // putting it together with a fold

Any ideas on how to get the Col function to work correctly? Also, while I want to implement this as a function of lists (and plan on completing this implementation) is there another way I should represent matrices in F#?

Munkeymon posted:

Unrelated to my response, I'm trying to make a (what I would think is) fairly simple XML transformation happen in a web.config for local debugging.

web.config

XML code:

<?xml version="1.0" encoding="UTF-8"?>
<configuration>
	<configSections>
		<section name="SomeConfigSource" blah blah blah/>
<!-- etc -->
	</configSections>
<!-- etc -->
	<SomeConfigSource application="bin\Moar.config" />
<configuration/>

web.debug.config

XML code:

<?xml version="1.0"?>
<configuration xmlns:xdt="http://schemas.microsoft.com/XML-Document-Transform">
	<SomeConfigSource xdt:Transform="Replace" application="C:\Long\Path\To\My\Local\bin\Moar.config"/>
</configuration>

Am I nuts or should this work?

Okita posted:

Here's a slightly more concise(but more esoteric and cheeky) way of essentially doing what you did. Just trying to give you some ideas about ways you can go about doing this kind of stuff:

code:

//...

url = url.Replace("&", "&");
if (!url.StartsWith("http"))
{
    url = "http://REMOVED FOR EMPLOYER PRIVACY.com" + url;
}

var validExtensions = new List<string> {".jpg", ".jpeg", ".tif"};
var links = GetLinks(url, link => validExtensions.Exists(link.EndsWith));

foreach (string link in links)
{
    string productCode = link.Substring(0, link.LastIndexOf('.'));
    skuTextBox.AppendText(string.Format("{0};{1}{2}", productCode, url + link, Environment.NewLine));
}

//...

private List<string> GetLinks(string url, Predicate<string> condition = null)
{
    var htmlWeb = new HtmlAgilityPack.HtmlWeb();
    var doc = htmlWeb.Load(url);

    var links = new List<string>();
    
    foreach (HtmlNode link in doc.DocumentNode.SelectNodes("//a[@href]"))
    {
        if (link.Attributes["href"] != null)
        {
            string value = link.Attributes["href"].Value;
            
            if ((condition != null) && (! condition.Invoke(value)))
            {
                continue;
            }

            links.Add(value);
        }
    }
       
    return links;
}

GrumpyDoctor posted:

I need someone to point me in the right direction regarding versioning. I don't even know the best practices for simple cases, and that's assuredly not what I'm working with.

My solution has nine projects: four native C++ projects, two C# projects, one F# project, one C++/CLI project, and a Setup & Deployment project. Some kind of automatic versioning would be great, if that's plausible, but I'd settle for keeping my version numbers in as few places as possible so that there aren't synchronization issues. I know that solutions for this sort of problem can go all the way to full-blown CI servers, but this is just me on a laptop and it's going to stay that way. Thus far it's been "remember to update all the disparate version entries with each release and hope you don't gently caress it up," and it's time to learn how to do things, if not right, at least better.

So: any suggestions for where to start?

Gilg posted:

A couple days old, but I didn't see anyone respond, so I think you should look into SlowCheetah, a Visual Studio extension. The built-in Web.config transformation is only done during publishing, i.e. not when local debugging. SlowCheeath will cover that case and also allow you to use transforms with non-ASP.NET projects.

wwb posted:

No idea what to do for the C++ stuff. Not exactly sure what the build system is for F# but this trick should work if it is using an assembly info file. Tricks were written for C# and I'm sure they work there as we use them across the board. At least across the board until teamcity added the assemblyinfo patcher.

What you want to do is create a central version info file someplace source controlled that everything can hit. In this file, at least for C#, you'd have the relevent parts of the AssemblyInfo.cs file -- eg:

code:

using System.Reflection;

// Version information for an assembly consists of the following four values:
//
//      Major Version
//      Minor Version 
//      Build Number
//      Revision
//
// You can specify all the values or you can default the Build and Revision Numbers 
// by using the '*' as shown below:
// [assembly: AssemblyVersion("1.0.*")]
[assembly: AssemblyVersion("1.9.0.*")]
[assembly: AssemblyFileVersion("1.9.0.0")]

[assembly: AssemblyCompany("Whateavs")]
[assembly: AssemblyProduct("My lovely project")]
[assembly: AssemblyCopyright("Copyright © Whomever 2012")]

You should also remove these lines from each assembly info.

Then go to each project and choose add existing item and choose to link the file in. MsBuild will pull it in every time.

No idea how to do this with C++, F# should probably work similarly but I've never got anything working in F# large enough to worry about versioning.

Another approach might be DropKick, not quite a full-blown CI server but designed to handle scenarios like this.

GrumpyDoctor posted:

Sure, but the F# project wants a .fs info file, and the C++/CLI project wants a .cpp info file. They can't all use the same file (because the syntaxes don't match), hence my question.

GrumpyDoctor posted:

Sure, but the F# project wants a .fs info file, and the C++/CLI project wants a .cpp info file. They can't all use the same file (because the syntaxes don't match), hence my question.

e: This also doesn't synchronize anything with the Setup & Deployment project version number. I might not want to do that anyway (I don't know best practices here) but if I don't how should I be handling that?

Munkeymon posted:

That's pretty nice in that it confirms that A) I'm not crazy and B) it should work, but, of course, it just doesn't and doesn't say why, though I am getting some message-level errors that our custom config section isn't in the schema, so maybe I'll spend several hours figuring that out and it'll magically work

Edit: wasted all morning making schemas and tweaking configs and schemas until everything built without any Warnings or Messages to indicate a problem with any XML, schema or config and it still simply doesn't happen even though SlowCheetah has no trouble figuring out the transformation. gently caress enterprisey XML bullshit forever.

Prefect Six posted:

I'm dumb and obviously didn't pay close attention to the beginning variables chapter but why does

code:

double IceLoad = .25;
Console.WriteLine(IceLoad);

Produce the result .25, but

code:

double IceLoad = 1 / 4;
Console.WriteLine(IceLoad);

produce 0?

i barely GNU her! posted:

The result of dividing an integer by an integer is an integer.

Prefect Six posted:

So how do I make C# perform mathematical equations? Put everything in a variable?

e: yes this would be easier and already done in a spreadsheet, but I want to do it just to learn to code.

e2: Answer: add f to the end of integers. Thanks so much thread!

GrumpyDoctor posted:

What development environment are you using? Because dropping this code into a .fs file in Visual Studio 2010 gives me the exact problem right away: "Value restriction. The value 'A3' has been inferred to have generic type. Either define 'A3' as a simple data term, make it a function with explicit arguments or, if you do not intend for it to be generic, add a type annotation." There's a specific MSDN page about this.

You can see the problem yourself by noticing the type that's been inferred for Col, which is int list list -> 'a list. (This is the number one way I debug F# type errors.) That's not right, as you definitely were expecting an int list. What's the problem? Well, again, Visual Studio has immediately noticed something wrong with this line:

code:

      [for R in lst do R.[i]]

which should actually be

code:

      [for R in lst -> R.[i]]

do is used for "throw away this result" operations. (I'm not actually sure why this didn't result in the type of Col being int list list -> unit list, but whatever.) Fixing your list comprehension fixed the type of Col which fixed your definition of A3.

Bozart posted:

Thanks for this - I'm still learning the F# syntax and I didn't / don't fully understand the difference between -> and do. I'll figure it out now.

GrumpyDoctor posted:

If there's an overarching rule about the difference between them, I'm unaware of it, because they're both used in various, unrelated places. It's just that in list comprehensions specifically they have the difference I described.

The Gripper posted:

It's pretty much what you said, any do block must return unit and in the list comprehension it will produce an empty list unless that expression yields a value, and I think the reason the type of Col was generic and not list unit was that unit isn't actually a type (in the same sense as int and list):

code:

> int;;
val it : (int -> int) = <fun:it@106-2>
> unit;;

  unit;;
  ^^^^

stdin(107,1): error FS0039: The value or constructor 'unit' is not defined

The list comprehension would work with a do if it were worded as [for R in list do yield R.[i]], since the block conforms to the requirements and the expression yields a result.

ninjeff posted:

I could be totally wrong and just embarrassing myself, though, in which case my bad.