Style example output and errors in the docs.

doc/site/classes.markdown

@@ -83,7 +83,7 @@ Many methods on a class exist to expose or compute some property of the object.
 For example:
 
     :::wren
-    System.print("string".count) // "6".
+    System.print("string".count) //> 6
 
 These *getters* are just another kind of method—one without a parameter
 list. You can define them like so:
@@ -326,13 +326,13 @@ not the instance. They can be used in *both* instance and static methods.
 Just like instance fields, static fields are initially `null`:
 
     :::wren
-    System.print(Foo.bar) // null.
+    System.print(Foo.bar) //> null
 
 They can be used from static methods:
 
     :::wren
     Foo.set("foo")
-    System.print(Foo.bar) // foo.
+    System.print(Foo.bar) //> foo
 
 And also instance methods. When you do so, there is still only one static field
 shared among all instances of the class:
@@ -342,7 +342,7 @@ shared among all instances of the class:
     var foo2 = Foo.new()
 
     foo1.setFromInstance("updated")
-    System.print(foo2.baz) // updated.
+    System.print(foo2.baz) //> updated
 
 ## Inheritance
 
@@ -374,20 +374,20 @@ are not inherited.
 
     class Pegasus is Unicorn {}
 
-    Pegasus.canFly // ERROR: Static methods are not inherited.
+    Pegasus.canFly //! Static methods are not inherited.
 
 This also means constructors are not inherited:
 
     :::wren
     class Unicorn {
-      this new(name) {
+      construct new(name) {
         System.print("My name is " + name + ".")
       }
     }
 
     class Pegasus is Unicorn {}
 
-    Pegasus.new("Fred") // Error!
+    Pegasus.new("Fred") //! Pegasus does not define new().
 
 Each class gets to control how it may be constructed independently of its base
 classes. However, constructor *initializers* are inherited since those are
@@ -397,15 +397,15 @@ This means you can do `super` calls inside a constructor:
 
     :::wren
     class Unicorn {
-      this new(name) {
+      construct new(name) {
         System.print("My name is " + name + ".")
       }
     }
 
     class Pegasus is Unicorn {
-      this new(name) {
+      construct new(name) {
         super(name)
       }
     }
 
-    Pegasus.new("Fred") // Prints "My name is Fred.".
+    Pegasus.new("Fred") //> My name is Fred

doc/site/core/bool.markdown

@@ -10,8 +10,8 @@ Boolean values. There are two instances, `true` and `false`.
 Returns the logical complement of the value.
 
     :::wren
-    System.print(!true) // "false".
-    System.print(!false) // "true".
+    System.print(!true) //> false
+    System.print(!false) //> true
 
 ### toString
 

doc/site/core/class.markdown

@@ -15,14 +15,14 @@ The superclass of this class.
     class Crustacean {}
     class Crab is Crustacean {}
 
-    System.print(Crab.supertype) // "Crustacean".
+    System.print(Crab.supertype) //> Crustacean
 
 A class with no explicit superclass implicitly inherits Object:
 
     :::wren
-    System.print(Crustacean.supertype) // "Object".
+    System.print(Crustacean.supertype) //> Object
 
 Object forms the root of the class hierarchy and has no supertype:
 
     :::wren
-    System.print(Object.supertype) // "null".
+    System.print(Object.supertype) //> null

doc/site/core/fiber.markdown

@@ -39,9 +39,9 @@ here means the last fiber that was started using `call` and not `run`.
       System.print("After yield")
     }
 
-    fiber.call()                // "Before yield"
-    System.print("After call")  // "After call"
-    fiber.call()                // "After yield"
+    fiber.call()                //> Before yield
+    System.print("After call")  //> After call
+    fiber.call()                //> After yield
 
 When resumed, the parent fiber's `call()` method returns `null`.
 
@@ -50,7 +50,7 @@ If a yielded fiber is resumed by calling `call()` or `run()` with an argument,
 
     :::wren
     var fiber = Fiber.new {
-      System.print(Fiber.yield()) // "value"
+      System.print(Fiber.yield()) //> value
     }
 
     fiber.call()        // Run until the first yield.
@@ -77,7 +77,7 @@ Similar to `Fiber.yield` but provides a value to return to the parent fiber's
       Fiber.yield("value")
     }
 
-    System.print(fiber.call()) // "value"
+    System.print(fiber.call()) //> value
 
 ## Methods
 
@@ -107,7 +107,7 @@ If the called fiber is resuming from a yield, the `yield()` method returns
     }
 
     fiber.call()
-    fiber.call() // Prints "null".
+    fiber.call() //> null
 
 ### **call**(value)
 
@@ -120,7 +120,7 @@ Invokes the fiber or resumes the fiber if it is in a paused state and sets
     }
 
     fiber.call()
-    fiber.call("value") // Prints "value".
+    fiber.call("value") //> value
 
 ### **isDone**
 

doc/site/core/fn.markdown

@@ -25,8 +25,8 @@ It is a runtime error if `function` is not a function.
 The number of arguments the function requires.
 
     :::wren
-    System.print(Fn.new {}.arity)             // 0.
-    System.print(Fn.new {|a, b, c| a }.arity) // 3.
+    System.print(Fn.new {}.arity)             //> 0
+    System.print(Fn.new {|a, b, c| a }.arity) //> 3
 
 ### **call**(args...)
 
@@ -34,10 +34,10 @@ Invokes the function with the given arguments.
 
     :::wren
     var fn = Fn.new { |arg|
-      System.print(arg)
+      System.print(arg)     //> Hello world
     }
 
-    fn.call("Hello world") // Prints "Hello world".
+    fn.call("Hello world")
 
 It is a runtime error if the number of arguments given is less than the arity
 of the function. If more arguments are given than the function's arity they are

doc/site/core/list.markdown

@@ -26,14 +26,14 @@ Inserts the `item` at `index` in the list.
     :::wren
     var list = ["a", "b", "c", "d"]
     list.insert(1, "e")
-    System.print(list) // "[a, e, b, c, d]".
+    System.print(list) //> [a, e, b, c, d]
 
 The `index` may be one past the last index in the list to append an element.
 
     :::wren
     var list = ["a", "b", "c"]
     list.insert(3, "d")
-    System.print(list) // "[a, b, c, d]".
+    System.print(list) //> [a, b, c, d]
 
 If `index` is negative, it counts backwards from the end of the list. It bases this on the length of the list *after* inserted the element, so that `-1` will append the element, not insert it before the last element.
 
@@ -41,12 +41,12 @@ If `index` is negative, it counts backwards from the end of the list. It bases t
     var list = ["a", "b"]
     list.insert(-1, "d")
     list.insert(-2, "c")
-    System.print(list) // "[a, b, c, d]".
+    System.print(list) //> [a, b, c, d]
 
 Returns the inserted item.
 
     :::wren
-    System.print(["a", "c"].insert(1, "b")) // "b".
+    System.print(["a", "c"].insert(1, "b")) //> b
 
 It is a runtime error if the index is not an integer or is out of bounds.
 
@@ -64,11 +64,11 @@ are shifted up to fill in where the removed element was.
     :::wren
     var list = ["a", "b", "c", "d"]
     list.removeAt(1)
-    System.print(list) // "[a, c, d]".
+    System.print(list) //> [a, c, d]
 
 Returns the removed item.
 
-    System.print(["a", "b", "c"].removeAt(1)) // "b".
+    System.print(["a", "b", "c"].removeAt(1)) //> b
 
 It is a runtime error if the index is not an integer or is out of bounds.
 
@@ -79,7 +79,7 @@ the end of the list where `-1` is the last element.
 
     :::wren
     var list = ["a", "b", "c"]
-    System.print(list[1]) // "b".
+    System.print(list[1]) //> b
 
 It is a runtime error if the index is not an integer or is out of bounds.
 
@@ -91,6 +91,6 @@ backwards from the end of the list where `-1` is the last element.
     :::wren
     var list = ["a", "b", "c"]
     list[1] = "new"
-    System.print(list) // "[a, new, c]".
+    System.print(list) //> [a, new, c]
 
 It is a runtime error if the index is not an integer or is out of bounds.

doc/site/core/map.markdown

@@ -50,8 +50,8 @@ map, returns `null`.
 
     :::wren
     var map = {"george": "harrison", "ringo": "starr"}
-    System.print(map["ringo"]) // "starr".
-    System.print(map["pete"])  // "null".
+    System.print(map["ringo"]) //> starr
+    System.print(map["pete"])  //> null
 
 ### **[**key**]=**(value) operator
 

doc/site/core/null.markdown

@@ -8,4 +8,4 @@
 Returns `true`, since `null` is considered [false](../control-flow.html#truth).
 
     :::wren
-    System.print(!null) // "true".
+    System.print(!null) //> true

doc/site/core/num.markdown

@@ -12,7 +12,7 @@ It is a runtime error if `value` is not a string.
 
 ### Num.**pi**
 
-The value of π.
+The value of π.
 
 ## Methods
 
@@ -21,7 +21,7 @@ The value of π.
 The absolute value of the number.
 
     :::wren
-    -123.abs // 123
+    System.print(-123.abs) //> 123
 
 ### **acos**
 
@@ -45,8 +45,8 @@ numbers to determine the quadrant of the result.
 Rounds the number up to the nearest integer.
 
     :::wren
-    1.5.ceil // 2
-    (-3.2).ceil // -3
+    System.print(1.5.ceil)    //> 2
+    System.print((-3.2).ceil) //> -3
 
 ### **cos**
 
@@ -57,24 +57,24 @@ The cosine of the number.
 Rounds the number down to the nearest integer.
 
     :::wren
-    1.5.floor    // 1
-    (-3.2).floor // -4
+    System.print(1.5.floor)    //> 1
+    System.print((-3.2).floor) //> -4
 
 ### **isInfinity**
 
 Whether the number is positive or negative infinity or not.
 
     :::wren
-    99999.isInfinity  // false
-    (1/0).isInfinity  // true
+    System.print(99999.isInfinity)  //> false
+    System.print((1/0).isInfinity)  //> true
 
 ### **isInteger**
 
 Whether the number is an integer or has some fractional component.
 
     :::wren
-    2.isInteger   // true
-    2.3.isInteger // false
+    System.print(2.isInteger)   //> true
+    System.print(2.3.isInteger) //> false
 
 ### **isNan**
 
@@ -100,7 +100,7 @@ Negates the number.
 
     :::wren
     var a = 123
-    -a // -123
+    System.print(-a) //> -123
 
 ### **-**(other), **+**(other), **/**(other), **\***(other) operators
 
@@ -148,9 +148,9 @@ from the beginning number to the ending number.
 
     :::wren
     var range = 1.2..3.4
-    System.print(range.min)         // 1.2
-    System.print(range.max)         // 3.4
-    System.print(range.isInclusive) // true
+    System.print(range.min)         //> 1.2
+    System.print(range.max)         //> 3.4
+    System.print(range.isInclusive) //> true
 
 ### **...**(other) operator
 
@@ -159,6 +159,6 @@ from the beginning number to the ending number not including the ending number.
 
     :::wren
     var range = 1.2...3.4
-    System.print(range.min)         // 1.2
-    System.print(range.max)         // 3.4
-    System.print(range.isInclusive) // false
+    System.print(range.min)         //> 1.2
+    System.print(range.max)         //> 3.4
+    System.print(range.isInclusive) //> false

doc/site/core/range.markdown

@@ -14,8 +14,8 @@ The starting point of the range. A range may be backwards, so this can be
 greater than [to].
 
     :::wren
-    (3..5).min // 3.
-    (4..2).min // 4.
+    System.print((3..5).min) //> 3
+    System.print((4..2).min) //> 4
 
 ### **to**
 
@@ -23,8 +23,8 @@ The endpoint of the range. If the range is inclusive, this value is included,
 otherwise it is not.
 
     :::wren
-    (3..5).min // 5.
-    (4..2).min // 2.
+    System.print((3..5).min) //> 5
+    System.print((4..2).min) //> 2
 
 ### **min**
 
@@ -32,8 +32,8 @@ The minimum bound of the range. Returns either `from`, or `to`, whichever is
 lower.
 
     :::wren
-    (3..5).min // 3.
-    (4..2).min // 2.
+    System.print((3..5).min) //> 3
+    System.print((4..2).min) //> 2
 
 ### **max**
 
@@ -41,16 +41,16 @@ The maximum bound of the range. Returns either `from`, or `to`, whichever is
 greater.
 
     :::wren
-    (3..5).min // 5.
-    (4..2).min // 4.
+    System.print((3..5).min) //> 5
+    System.print((4..2).min) //> 4
 
 ### **isInclusive**
 
 Whether or not the range includes `to`. (`from` is always included.)
 
     :::wren
-    (3..5).isInclusive   // true.
-    (3...5).isInclusive  // false.
+    System.print((3..5).isInclusive)   //> true
+    System.print((3...5).isInclusive)  //> false
 
 ### **iterate**(iterator), **iteratorValue**(iterator)
 

doc/site/core/sequence.markdown

@@ -17,8 +17,8 @@ If it returns something [false](../control-flow.html#truth), stops iterating
 and returns the value. Otherwise, returns `true`.
 
     :::wren
-    [1, 2, 3].all {|n| n > 2} // False.
-    [1, 2, 3].all {|n| n < 4} // True.
+    System.print([1, 2, 3].all {|n| n > 2}) //> false
+    System.print([1, 2, 3].all {|n| n < 4}) //> true
 
 ### **any**(predicate)
 
@@ -29,8 +29,8 @@ If it returns something [true](../control-flow.html#truth), stops iterating and
 returns that value. Otherwise, returns `false`.
 
     :::wren
-    [1, 2, 3].any {|n| n < 1} // False.
-    [1, 2, 3].any {|n| n > 2} // True.
+    System.print([1, 2, 3].any {|n| n < 1}) //> false
+    System.print([1, 2, 3].any {|n| n > 2}) //> true
 
 ### **contains**(element)
 
@@ -51,8 +51,8 @@ Iterates over the sequence, passing each element to the function `predicate`
 and counting the number of times the returned value evaluates to `true`.
 
     :::wren
-    [1, 2, 3].count {|n| n > 2} // 1.
-    [1, 2, 3].count {|n| n < 4} // 3.
+    System.print([1, 2, 3].count {|n| n > 2}) //> 1
+    System.print([1, 2, 3].count {|n| n < 4}) //> 3
 
 ### **each**(function)
 
@@ -88,7 +88,9 @@ original sequence while it is iterated.
     :::wren
     var doubles = [1, 2, 3].map {|n| n * 2 }
     for (n in doubles) {
-      System.print(n) // "2", "4", "6".
+      System.print(n) //> 2
+                      //> 4
+                      //> 6
     }
 
 The returned sequence is *lazy*. It only applies the mapping when you iterate
@@ -105,7 +107,7 @@ To force eager evaluation, just call `.toList` on the result.
     var numbers = [1, 2, 3]
     var doubles = numbers.map {|n| n * 2 }.toList
     numbers.add(4)
-    System.print(doubles) // [2, 4, 6].
+    System.print(doubles) //> [2, 4, 6]
 
 ### **reduce**(function)
 
@@ -127,7 +129,7 @@ the sequence is empty, returns `seed`.
 Creates a [list](list.html) containing all the elements in the sequence.
 
     :::wren
-    (1..3).toList  // [1, 2, 3].
+    System.print((1..3).toList)  //> [1, 2, 3]
 
 If the sequence is already a list, this creates a copy of it.
 
@@ -140,9 +142,11 @@ During iteration, each element in the original sequence is passed to the
 function `predicate`. If it returns `false`, the element is skipped.
 
     :::wren
-    var odds = (1..10).where {|n| n % 2 == 1 }
+    var odds = (1..6).where {|n| n % 2 == 1 }
     for (n in odds) {
-      System.print(n) // "1", "3", "5", "7", "9".
+        System.print(n) //> 1
+                        //> 3
+                        //> 5
     }
 
 The returned sequence is *lazy*. It only applies the filtering when you iterate
@@ -160,4 +164,4 @@ To force eager evaluation, just call `.toList` on the result.
     var numbers = [1, 2, 3, 4, 5, 6]
     var odds = numbers.where {|n| n % 2 == 1 }.toList
     numbers.add(7)
-    System.print(odds) // [1, 3, 5].
+    System.print(odds) //> [1, 3, 5]

doc/site/core/string.markdown

@@ -37,7 +37,7 @@ on strings *return* byte indexes too. So, for example, this does what you want:
     :::wren
     var metalBand = "Fäcëhämmër"
     var hPosition = metalBand.indexOf("h")
-    System.print(metalBand[hPosition]) // "h"
+    System.print(metalBand[hPosition]) //> h
 
 If you want to work with a string as a sequence numeric code points, call the
 `codePoints` getter. It returns a [Sequence](sequence.html) that decodes UTF-8
@@ -53,7 +53,7 @@ ignores any UTF-8 encoding and works directly at the byte level.
 Creates a new string containing the UTF-8 encoding of `codePoint`.
 
     :::wren
-    String.fromCodePoint(8225) // "‡"
+    String.fromCodePoint(8225) //> ‡
 
 It is a runtime error if `codePoint` is not an integer between `0` and
 `0x10ffff`, inclusive.
@@ -68,7 +68,7 @@ methods, the returned object also has a subscript operator that can be used to
 directly index bytes.
 
     :::wren
-    System.print("hello".bytes[1]) // 101, for "e".
+    System.print("hello".bytes[1]) //> 101 (for "e")
 
 The `count` method on the returned sequence returns the number of bytes in the
 string. Unlike `count` on the string itself, it does not have to iterate over
@@ -83,15 +83,15 @@ single-character strings, this returns the numeric code point values.
 
     :::wren
     var string = "(ᵔᴥᵔ)"
-    System.print(string.codePoints[0]) // 40, for "(".
-    System.print(string.codePoints[4]) // 7461, for "ᴥ".
+    System.print(string.codePoints[0]) //> 40 (for "(")
+    System.print(string.codePoints[4]) //> 7461 (for "ᴥ")
 
 If the byte at `index` does not begin a valid UTF-8 sequence, or the end of the
 string is reached before the sequence is complete, returns `-1`.
 
     :::wren
     var string = "(ᵔᴥᵔ)"
-    System.print(string.codePoints[2]) // -1, in the middle of "ᵔ".
+    System.print(string.codePoints[2]) //> -1 (in the middle of "ᵔ")
 
 ### **contains**(other)
 
@@ -132,7 +132,7 @@ for iterating over the *code points* in the string:
       codePoints.add(c)
     }
 
-    System.print(codePoints) // ["(", "ᵔ", "ᴥ", "ᵔ", ")"].
+    System.print(codePoints) //> [(, ᵔ, ᴥ, ᵔ, )]
 
 If the string contains any bytes that are not valid UTF-8, this iterates over
 those too, one byte at a time.
@@ -162,7 +162,7 @@ Check if the string is not equal to `other`.
 Returns a string containing the code point starting at byte `index`.
 
     :::wren
-    System.print("ʕ•ᴥ•ʔ"[5]) // "ᴥ".
+    System.print("ʕ•ᴥ•ʔ"[5]) //> ᴥ
 
 Since `ʕ` is two bytes in UTF-8 and `•` is three, the fifth byte points to the
 bear's nose.
@@ -171,7 +171,7 @@ If `index` points into the middle of a UTF-8 sequence or at otherwise invalid
 UTF-8, this returns a one-byte string containing the byte at that index:
 
     :::wren
-    System.print("I ♥ NY"[3]) // One-byte string whose value is 153.
+    System.print("I ♥ NY"[3]) //> (one-byte string [153])
 
 It is a runtime error if `index` is greater than the number of bytes in the
 string.

doc/site/core/system.markdown

@@ -16,7 +16,7 @@ Prints [object] to the console followed by a newline. If not already a string,
 the object is converted to a string by calling `toString` on it.
 
     :::wren
-    System.print("I like bananas") // Prints "I like bananas".
+    System.print("I like bananas") //> I like bananas
 
 ### System.**printAll**(sequence)
 
@@ -24,7 +24,7 @@ Iterates over [sequence] and prints each element, then prints a single newline
 at the end. Each element is converted to a string by calling `toString` on it.
 
     :::wren
-    System.printAll([1, [2, 3], 4]) // Prints "1[2, 3]4".
+    System.printAll([1, [2, 3], 4]) //> 1[2, 3]4
 
 ### System.**write**(object)
 
@@ -32,7 +32,7 @@ Prints a single value to the console, but does not print a newline character
 afterwards. Converts the value to a string by calling `toString` on it.
 
     :::wren
-    System.write(4 + 5) // Prints "9".
+    System.write(4 + 5) //> 9
 
 In the above example, the result of `4 + 5` is printed, and then the prompt is
 printed on the same line because no newline character was printed afterwards.

doc/site/error-handling.markdown

@@ -11,11 +11,11 @@ simple bugs where your code doesn't follow the language's grammar, like:
     :::wren
     1 + * 2
 
-Wren will detect these errors as soon as it tries to read your code. When it
-hits one, you'll get a friendly error message, like:
+Wren detects these errors as soon as it tries to read your code. When it hits
+one, you get a friendly error message, like:
 
     :::text
-    [script.wren line 1] Error on '*': Unexpected token for expression.
+    [main line 1] Error on '*': Unexpected token for expression.
 
 Some slightly more "semantic" errors fall into this bucket too. Things like
 using a variable that hasn't been defined, or declaring two variables with the
@@ -28,7 +28,7 @@ same name in the same scope. So if you do:
 Wren tells you:
 
     :::text
-    [script.wren line 2] Error on 'a': Top-level variable is already defined.
+    [main line 2] Error on 'a': Top-level variable is already defined.
 
 Note that it does this before it executes *any* code. Unlike some other
 scripting languages, Wren tries to help you find your errors as soon as
@@ -78,7 +78,7 @@ This exits with:
 
     :::text
     Subscript must be a number or a range.
-    [script.wren line 2] in (script)
+    [main line 2] in (script)
 
 These are the most two common kinds of runtime errors, but there are others.
 Stuff like out of bounds errors on lists, calling a function with the wrong

doc/site/expressions.markdown

@@ -97,14 +97,14 @@ call:
 
     :::wren
     class Base {
-      method {
+      method() {
         System.print("base method")
       }
     }
 
     class Derived is Base {
-      method {
-        super.method // Prints "base method".
+      method() {
+        super.method() //> base method
       }
     }
 
@@ -113,19 +113,17 @@ base class constructor:
 
     :::wren
     class Base {
-      this new(arg) {
-        System.print("base constructor got " + arg)
+      construct new(arg) {
+        System.print("base got " + arg)
       }
     }
 
     class Derived is Base {
-      this new() {
-        super("value") // Prints "base constructor got value".
+      construct new() {
+        super("value") //> base got value
       }
     }
 
-**TODO: constructors**
-
 ## Operators
 
 Wren has most of the same operators you know and love with the same precedence
@@ -213,21 +211,20 @@ value of the left-hand side, the right-hand operand expression may or may not
 be evaluated. Because of this, they cannot be overloaded and their behavior is
 fixed.
 
-A `&&` ("logical and") expression evaluates the left-hand argument. If it's
-[false](control-flow.html#truth), it returns that value. Otherwise it evaluates
-and returns the right-hand argument.
+A `&&` ("logical and") expression evaluates the left-hand argument. If
+it's [false](control-flow.html#truth), it returns that value. Otherwise it evaluates and returns the right-hand argument.
 
     :::wren
-    System.print(false && 1)  // false
-    System.print(1 && 2)      // 2
+    System.print(false && 1)  //> false
+    System.print(1 && 2)      //> 2
 
 An `||` ("logical or") expression is reversed. If the left-hand argument is
 [true](control-flow.html#truth), it's returned, otherwise the right-hand
 argument is evaluated and returned:
 
     :::wren
-    System.print(false || 1)  // 1
-    System.print(1 || 2)      // 1
+    System.print(false || 1)  //> 1
+    System.print(1 || 2)      //> 1
 
 ## The conditional operator `?:`
 
@@ -251,11 +248,11 @@ operand is a class. It evaluates to `true` if the object is an instance of the
 class (or one of its subclasses).
 
     :::wren
-    123 is Num     // true
-    "s" is Num     // false
-    null is String // false
-    [] is List     // true
-    [] is Sequence // true
+    System.print(123 is Num)     //> true
+    System.print("s" is Num)     //> false
+    System.print(null is String) //> false
+    System.print([] is List)     //> true
+    System.print([] is Sequence) //> true
 
 ## Precedence
 

doc/site/fibers.markdown

@@ -32,8 +32,8 @@ using the `Fiber` class's constructor:
     }
 
 Creating a fiber does not immediately run it. It's just a first class bundle of
-code sitting there waiting to be activated, a bit like a
-[function](functions.html).
+code sitting there waiting to be activated, a bit like
+a [function](functions.html).
 
 ## Invoking fibers
 
@@ -49,9 +49,9 @@ it's considered *done*:
 
     :::wren
     var fiber = Fiber.new { System.print("Hi") }
-    fiber.isDone // false
+    System.print(fiber.isDone) //> false
     fiber.call()
-    fiber.isDone // true
+    System.print(fiber.isDone) //> true
 
 When it finishes, it automatically resumes the fiber that called it. It's a
 runtime error to try to call a fiber that is already done.
@@ -91,9 +91,10 @@ This program prints:
     fiber 2
     main 3
 
-Note that even though this program has *concurrency*, it's still
-*deterministic*. You can reason precisely about what it's doing and aren't at
-the mercy of a thread scheduler playing Russian roulette with your code.
+Note that even though this program has *concurrency*, it's
+still *deterministic*. You can reason precisely about what it's doing and
+aren't at the mercy of a thread scheduler playing Russian roulette with your
+code.
 
 ## Passing values
 
@@ -116,7 +117,7 @@ ignored. That's because the fiber isn't waiting on a `yield()` call, so there's
 nowhere for the sent value to go.
 
 Fibers can also pass values *back* when they yield. If you pass an argument to
-`yield()`, that will become the return value of the `call` that was used to
+`yield()`, that will become the return value of the `call()` that was used to
 invoke the fiber:
 
     :::wren
@@ -135,8 +136,8 @@ Python and C# that have *generators*. Those let you define a function call that
 you can suspend and resume. When using the function, it appears like a sequence
 you can iterate over.
 
-Wren's fibers can do that, but they can do much more. Like Lua, they are full
-*coroutines*—they can suspend from anywhere in the callstack. For
+Wren's fibers can do that, but they can do much more. Like Lua, they are
+full *coroutines*—they can suspend from anywhere in the callstack. For
 example:
 
     :::wren

doc/site/functions.markdown

@@ -26,8 +26,9 @@ a method. At its simplest, it looks like this:
     }
 
 Here we're invoking the `callMe` method on `blondie`. We're passing one
-argument, a function whose body is the following
-[block](syntax.html#blocks)—everything between that pair of curly braces.
+argument, a function whose body is the
+following [block](syntax.html#blocks)—everything between that pair of
+curly braces.
 
 Methods that take a block argument receive it as a normal parameter. `callMe`
 could be defined like so:
@@ -144,7 +145,7 @@ leaving the scope where the function is defined:
 
     :::wren
     class Counter {
-      static create {
+      static create() {
         var i = 0
         return Fn.new { i = i + 1 }
       }
@@ -156,7 +157,7 @@ the function is returned from `create`, it is still able to read and assign
 to`i`:
 
     :::wren
-    var counter = Counter.create
-    System.print(counter.call()) // Prints "1".
-    System.print(counter.call()) // Prints "2".
-    System.print(counter.call()) // Prints "3".
+    var counter = Counter.create()
+    System.print(counter.call()) //> 1
+    System.print(counter.call()) //> 2
+    System.print(counter.call()) //> 3

doc/site/lists.markdown

@@ -19,20 +19,20 @@ element you want. Like most languages, indexes start at zero:
 
     :::wren
     var hirsute = ["sideburns", "porkchops", "'stache", "goatee"]
-    hirsute[0] // "sideburns".
-    hirsute[1] // "porkchops".
+    hirsute[0] //> sideburns
+    hirsute[1] //> porkchops
 
 Negative indices counts backwards from the end:
 
     :::wren
-    hirsute[-1] // "goatee".
-    hirsute[-2] // "'stache".
+    hirsute[-1] //> goatee
+    hirsute[-2] //> 'stache
 
 It's a runtime error to pass an index outside of the bounds of the list. If you
 don't know what those bounds are, you can find out using count:
 
     :::wren
-    hirsute.count // 4.
+    hirsute.count //> 4
 
 ## Slices and ranges
 
@@ -40,7 +40,7 @@ Sometimes you want to copy a chunk of elements from a list. You can do that by
 passing a [range](values.html#ranges) to the subscript operator, like so:
 
     :::wren
-    hirsute[1..2] // ["porkchops", "'stache"].
+    hirsute[1..2] //> [porkchops, 'stache]
 
 This returns a new list containing the elements of the original list whose
 indices are within the given range. Both inclusive and exclusive ranges work
@@ -59,14 +59,14 @@ existing element in the list using the subscript setter:
 
     :::wren
     hirsute[1] = "muttonchops"
-    System.print(hirsute[1]) // muttonchops.
+    System.print(hirsute[1]) //> muttonchops
 
 It's an error to set an element that's out of bounds. To grow a list, you can
 use `add` to append a single item to the end:
 
     :::wren
     hirsute.add("goatee")
-    System.print(hirsute.count) // 4.
+    System.print(hirsute.count) //> 4
 
 You can insert a new element at a specific position using `insert`:
 
@@ -84,9 +84,9 @@ back. Doing so counts back from the size of the list *after* it's grown by one:
     :::wren
     var letters = ["a", "b", "c"]
     letters.insert(3, "d")   // OK: inserts at end.
-    System.print(letters)        // ["a", "b", "c", "d"]
+    System.print(letters)        //> [a, b, c, d]
     letters.insert(-2, "e")  // Counts back from size after insert.
-    System.print(letters)        // ["a", "b", "c", "e", "d"]
+    System.print(letters)        //> [a, b, c, e, d]
 
 ## Removing elements
 
@@ -97,15 +97,15 @@ gap:
     :::wren
     var letters = ["a", "b", "c", "d"]
     letters.removeAt(1)
-    System.print(letters) // ["a", "c", "d"]
+    System.print(letters) //> [a, c, d]
 
 The `removeAt` method returns the removed item:
 
     :::wren
-    System.print(letters.removeAt(1)) // "c"
+    System.print(letters.removeAt(1)) //> c
 
 If you want to remove everything from the list, you can clear it:
 
     :::wren
     hirsute.clear()
-    System.print(hirsute) // []
+    System.print(hirsute) //> []

doc/site/maps.markdown

@@ -58,7 +58,7 @@ To find the value associated with some key, again you use your friend the
 subscript operator:
 
     :::wren
-    System.print(capitals["Idaho"]) // "Boise".
+    System.print(capitals["Idaho"]) //> Boise
 
 If the key is present, this returns its value. Otherwise, it returns `null`. Of
 course, `null` itself can also be used as a value, so seeing `null` here
@@ -69,15 +69,15 @@ To tell definitively if a key exists, you can call `containsKey()`:
     :::wren
     var belief = {"nihilism": null}
 
-    System.print(belief["nihilism"]) // "null" though key exists.
-    System.print(belief["solipsism"]) // Also "null".
-    System.print(belief.containsKey("nihilism")) // "true".
-    System.print(belief.containsKey("solipsism")) // "false".
+    System.print(belief["nihilism"]) //> null (though key exists)
+    System.print(belief["solipsism"]) //> null
+    System.print(belief.containsKey("nihilism")) //> true
+    System.print(belief.containsKey("solipsism")) //> false
 
 You can see how many entries a map contains using `count`:
 
     :::wren
-    System.print(capitals.count) // "3".
+    System.print(capitals.count) //> 3
 
 ## Removing entries
 
@@ -86,12 +86,12 @@ entry you want to delete:
 
     :::wren
     capitals.remove("Maine")
-    System.print(capitals.containsKey("Maine")) // "false".
+    System.print(capitals.containsKey("Maine")) //> false
 
 If the key was found, this returns the value that was associated with it:
 
     :::wren
-    System.print(capitals.remove("Georgia")) // "Atlanta".
+    System.print(capitals.remove("Georgia")) //> Atlanta
 
 If the key wasn't in the map to begin with, `remove()` just returns `null`.
 
@@ -100,7 +100,7 @@ can just call `clear()`:
 
     :::wren
     capitals.clear()
-    System.print(capitals.count) // "0".
+    System.print(capitals.count) //> 0
 
 [lists]: lists.html
 

doc/site/style.scss

@@ -138,6 +138,13 @@ h2 {
   font: 500 24px $header;
   margin: 24px 0 0 0;
   color: $link;
+
+  code {
+    border: none;
+    background: inherit;
+    color: inherit;
+    font-size: 24px;
+  }
 }
 
 h3 {
@@ -245,7 +252,7 @@ footer {
   span.vc { color: hsl(130, 60%, 40%); }
 
   // Numbers.
-  span.m, span.mi, span.mf { color: hsl(30, 80%, 45%); }// hsl(90, 80%, 35%)
+  span.m, span.mi, span.mf { color: hsl(30, 80%, 45%); }
 
   // Strings.
   span.s, span.s2  { color: hsl(40, 80%, 45%); }
@@ -257,6 +264,29 @@ footer {
 
   // Preprocessor directives.
   span.cp          { color: hsl(270, 40%, 60%); }
+
+  span.output {
+    float: right;
+    color: hsl(150, 50%, 45%);
+
+    &::before, &::after {
+      color: hsl(150, 50%, 80%);
+    }
+
+    // Tall angle brackets.
+    &::before { content: "\276c  "; }
+    &::after { content: ' \276d'; }
+  }
+
+  span.error {
+    float: right;
+    color: hsl(350, 50%, 60%);
+
+    &::before {
+      color: hsl(350, 50%, 75%);
+      content: "Error: ";
+    }
+  }
 }
 
 // Have a different primary color for the core library docs.

doc/site/syntax.markdown

@@ -91,7 +91,7 @@ statement for the else:
 
     :::wren
     if (happy && knowIt) {
-      hands.clap
+      hands.clap()
     } else System.print("sad")
 
 Blocks have two similar but not identical forms. Typically, blocks contain a

doc/site/values.markdown

@@ -57,12 +57,12 @@ A handful of escape characters are supported:
 A `\u` followed by four hex digits can be used to specify a Unicode code point:
 
     :::wren
-    System.print("\u0041\u0b83\u00DE") // "AஃÞ"
+    System.print("\u0041\u0b83\u00DE") //> AஃÞ
 
 A `\x` followed by two hex digits specifies a single unencoded byte:
 
     :::wren
-    System.print("\x48\x69\x2e") // "Hi."
+    System.print("\x48\x69\x2e") //> Hi.
 
 Strings are instances of class [String](core/string.html).
 
@@ -91,7 +91,7 @@ example:
     :::wren
     var list = ["a", "b", "c", "d", "e"]
     var slice = list[1..3]
-    System.print(slice) // ["b", "c", "d"]
+    System.print(slice) //> [b, c, d]
 
 Their class is [Range](core/range.html).
 

doc/site/variables.markdown

@@ -13,7 +13,7 @@ defined, it can be accessed by name as you would expect.
 
     :::wren
     var animal = "Slow Loris"
-    System.print(animal) // Prints "Slow Loris".
+    System.print(animal) //> Slow Loris
 
 ## Scope
 
@@ -22,11 +22,11 @@ until the end of the [block](syntax.html#blocks) where that definition appears.
 
     :::wren
     {
-      System.print(a) // ERROR! a doesn't exist yet.
+      System.print(a) //! "a" doesn't exist yet.
       var a = 123
-      System.print(a) // "123"
+      System.print(a) //> 123
     }
-    System.print(a) // ERROR! a doesn't exist anymore.
+    System.print(a) //! "a" doesn't exist anymore.
 
 Variables defined at the top level of a script are *top-level* and are visible
 to the [module](modules.html) system. All other variables are *local*.
@@ -38,15 +38,15 @@ intend to do much).
     var a = "outer"
     {
       var a = "inner"
-      System.print(a) // Prints "inner".
+      System.print(a) //> inner
     }
-    System.print(a) // Prints "outer".
+    System.print(a) //> outer
 
 Declaring a variable with the same name in the *same* scope *is* an error.
 
     :::wren
     var a = "hi"
-    var a = "again" // ERROR!
+    var a = "again" //! "a" is already declared.
 
 ## Assignment
 
@@ -65,6 +65,6 @@ assigned value.
 
     :::wren
     var a = "before"
-    System.print(a = "after") // Prints "after".
+    System.print(a = "after") //> after
 
 **TODO: Top-level names.**

util/generate_docs.py

@@ -93,6 +93,10 @@ def format_file(path, skip_up_to_date):
 
   html = markdown.markdown(contents, ['def_list', 'codehilite', 'smarty'])
 
+  # Use special formatting for example output and errors.
+  html = html.replace('<span class="c1">//&gt; ', '<span class="output">')
+  html = html.replace('<span class="c1">//! ', '<span class="error">')
+
   modified = datetime.fromtimestamp(os.path.getmtime(in_path))
   mod_str = modified.strftime('%B %d, %Y')