Differences

This shows you the differences between two versions of the page.

--- projects:fsharp_workshop [2015/03/14 18:11] – mkucia
+++ projects:fsharp_workshop [2015/05/02 12:45] (current) – [hello(x)] mkucia
@@ Line 1: / Line 1: @@
 ====== F# workshop ======
+{{:projects:fsharp_logo.png?nolink|}}
+===== Prerequisites =====
+    * https://www.linqpad.net/ or anything that can execute a REPL((Read, Evaluate, Print Loop))
+<hidden ↘ alternatives>
+    * Try F# in browser: http://www.tryfsharp.org/Create
+    * Use Visual Studio https://www.visualstudio.com/
+    * Get a full distribution: http://fsharp.org/use/windows/ OR http://fsharp.org/use/linux/
+</hidden>
+===== Agenda =====
-  * f(x) = Hello_Word
+  * hello(x)
-  * On Testing and
+  * Type providers (strongly typed)
+  * Matching
+  * Options
-===== How to use? Where to get? =====
-Try F# in browser:
-http://www.tryfsharp.org/Create
-Want to get a full distribution?
+===== hello(x) =====
-http://fsharp.org/ → Use F#
+<code fsharp>
-===== [Hello] =====
-<code>
 let hello = printf "Hello World"
+hello
 </code>
-===== Why? =====
+F# (pronounced “F Sharp”) is
 <WRAP group>
 <WRAP half column box>
-Pros:
   * Open Source
-  * Multi platform (Mono, .NET core)
+  * Multi platform
   * Modern, rising popularity
-  * Lazy evaluation
+  * Lazy evaluated
   * Reusable
-  * Reduced code
+  * Concise (See [[http://theburningmonk.com/2014/12/seven-ineffective-coding-habits-many-f-programmers-dont-have/|signal-to-noise ratio]])
   * Strongly typed
+  * Functional first
+    * Use functional to write easy to test and understand code
+    * Use imperative for performance and compatibility
+    * Use objective for organization and encapsulation
+  * Compiled
 </WRAP>
 <WRAP half column box>
-Cons:
   * Not suitable for real time (unpredictable performance)
-  * Different thinking process (you need to re-learn)
+  * Hard to optimize (Functional languages fundamentally don't model how your computer works)
+  * Functional first (different thinking process, you will need to re-learn)
 </WRAP>
 </WRAP>
+programming language for writing simple code to solve complex problems.
-<hidden ↘ C# equivalent>Since you want to know more, here is an explanation with more details […]</hidden>
+<code fsharp>
+let hello x = printfn "Hello %s" x
+hello "world"
+</code>
-===== Pitfalls and drawbacks =====
+Introducing pipeline operator:
-Functional languages fundamentally don't model how your computer works.
+<code fsharp>
-Harder to optimize.
+let hello x = printfn "Hello %s" x
+"World" |> hello
+</code>
+<hidden pipeline operator definition>
+How to define the operator?
+<code fsharp>
+let inline (|>) x f = f x
+</code>
+</hidden>
+<code fsharp>
+let square x = x * x
+let subtract x y = x - y
+</code>
+<code fsharp>
+let ``complicated stuff`` x =
+    printfn "%d" (add 5 (square x))
+</code>
+<code fsharp>
+let ``another complicated stuff`` x =
+    x |> square |> subtract 5 |> printfn "%d"
+</code>
+<code fsharp>
+let equation x =
+	x
+	|> (fun y -> y * y )
+	|> (fun y z -> y - z ) 5
+	|> printfn "%d"
+</code>
+Note that F# uses indentation to indicate block structure. Find out more [[http://fsharpforfunandprofit.com/posts/fsharp-syntax/|here]].
+===== Getting data =====
+The following example loads data from webpage and stores it in CSV file.
+<hidden>
+<code fsarp>
+open System.Net
+open System
+open System.IO
+let myURL = @"..."
+let fetchUrl callback url =
+    let req    = WebRequest.Create(Uri(url))
+    use resp   = req.GetResponse()
+    use stream = resp.GetResponseStream()
+    use reader = new IO.StreamReader(stream)
+    callback reader url
+let myCallback (reader:IO.StreamReader) url =
+    let htmlSource = reader.ReadToEnd()
+    htmlSource
+let getWebpageHTML = fetchUrl myCallback myURL
+type FirstOrLast =
+	| GetFirst
+	| GetLast
+let getDataFromHTML (rawHTML:String) =
+	let split (split_point:String) (firstOrLast:FirstOrLast) (text:String) =
+		let array = text.Split([|split_point|], 2, StringSplitOptions.RemoveEmptyEntries)
+		match firstOrLast with
+		| GetFirst -> array.First()
+		| GetLast  -> array.Last()
+	rawHTML
+	|> split @"<table id=""t2"">" GetLast
+	|> split @"</table>"          GetFirst
+let removeAllTags text =
+    Regex.Replace(text, "<.*?>", String.Empty);
+let getCSVfromHTML rawHTML =
+	let mutable (text:String) = String.Empty
+	text <- rawHTML
+	text <- text.Replace(@"</th>", ";");
+	text <- text.Replace(@"</td>", ";");
+	text <- text.Replace(@"</tr>", Environment.NewLine);
+	text <- text.Replace(',', '|');
+	text <- text.Replace('.', ',');
+	text <- removeAllTags text
+	text
+let CSVresult = getCSVfromHTML (getDataFromHTML getWebpageHTML)
+CSVresult.Dump()
+</code>
+</hidden>
+> An F# **type provider** is a component that provides types, properties, and methods for use in your program.
+> The key to information-rich programming is to eliminate barriers to working with diverse information sources. One significant barrier to including a source of information into a program is the need to represent that information as types, properties, and methods for use in a programming language environment.
+> Writing these types manually is very time-consuming and difficult to maintain. A common alternative is to use a code generator which adds files to your project; however, the conventional types of code generation do not integrate well into exploratory modes of programming supported by F# because the generated code must be replaced each time a service reference is adjusted.
 ===== References =====
   * http://fsharp.org/
+  * http://fsharp3sample.codeplex.com/wikipage?Title=MicroSamples
   * http://en.wikibooks.org/wiki/F_Sharp_Programming
   * http://fsharp.org/specs/language-spec/3.0/FSharpSpec-3.0-final.pdf
@@ Line 57: / Line 164: @@
   * http://fsharpforfunandprofit.com/site-contents/
   * http://research.microsoft.com/en-us/projects/kiwi/
+  * [[https://www.youtube.com/watch?v=hMjv8m--W28|Ten things F# can do, and C# can't - Liam McLennan @ DDD Brisbane 2013]]