Value Types, with their methods

Integer

An integer is a whole number, i.e. one that has no fractional part. It may be represented in decimal, hexadecimal or binary:

100 (decimal), 0x64 (hexadecimal) and 0b1100100 all have the same value.

Dot methods on an Int

Notes

• A literal integer or a named value of Type Int may always be passed into a function or procedure that is expecting a Float.
• The Int type is intended to represent whole numbers in the range:
• Maximum: 2⁵³ – 1 (which is 9007199254740991)
• Minimum value: – (2⁵³ – 1)
• If you go greater than the limit given above, the number will be accurate to approximately 17 decimal digits. Larger numbers will be rounded with zeros at the end, unless they are larger than around 2⁷⁰, when they will automatically be shown in exponential (scientific) format, as shown by the output from the following code:
+for ivariableName from 50expression to 100expression step 1expression call printTabprocedureName(0, i.asString()position, text) call printTabprocedureName(10, (2^i).asString()position, text) print expressionexpression end for

which produces:

Floating point

A floating point number is a decimal number that may have both integer and fractional parts.

It may be written using exponential (scientific) notation, e.g. 120.0 = 1.20e2, and 0.012 = 1.20e-2.

Dot methods on a Float

Notes

• The limits on floating point numbers are:
  • Maximum value: just over 1.0 × 10³⁰⁸
  • Minimum value: approximately 5.0 × 10^-324
• A variable that has been defined as being of Type Float may not be passed as an argument into a method that requires an Int, nor as an index into a List or an Array even if the variable contains no fractional part. It may, however, be converted into an Int before passing, using the function floor or ceiling.
• If you wish to define a variable to be of Type Float, but initialise it with an integer value, then add .0 on the end of the whole number, as in:
  variable hour set to 3.0

Boolean

A Boolean value is either true or false.

The keywords true and false must be written in lower case.

Dot method on a Boolean

String

A String represents text, i.e. a sequence of zero or more characters.

Notes

• A String is always delineated by double quotation marks.
• When typing from the keyboard into a literal string all the basic ASCII characters (0x20 to 0x7e) can be input. Similarly you can copy from elsewhere into a literal string provided only basic ASCII characters are include. You cannot use Alt-numeric-keypad
• As in most languages, strings are immutable. When you apply any operation or function with the intent of modifying an existing string, the existing string is never modified. Instead, the operation or function will return a new string that is based on the original, but with the specified differences.
• Strings can be built from other strings by concatenation using the plus operator, for example
• print "Hello" + " " + "World"
• A newline may be inserted within a string with \n, for example:
• print "Hello\nWorld"
• There are no 'substring' methods in Elan because you can always use an index range to get a substring, e.g. s[3..7] gives a string containing the fourth character to the seventh character inclusive of s. Note that the upper bound of the range is exclusive. See Indexed Value.
• There are, however, many dot methods that can be applied to strings:

Dot methods on a String

Interpolated strings

Strings are automatically interpolated, that is you can insert the values of variables or simple expressions within a string by enclosing them in braces. For example, assuming that the variables a and b are already defined as integers:

print "{a} times {b} equals {a*b}"

prints the sentence with the values and calculated expression inserted.

If velocity and fuel are already defined as Float then:

set message to "LANDED SAFELY AT SPEED {(velocity*100).floor()} FUEL {fuel.floor()}"

updates variable message with calculated (and printable) integer values.

Special characters

You may include single quotation marks 'within a string, but not double quotation marks ". Nor can you include the braces { or } within a literal string because of their special meanings. Instead, you use the constants quotes, openBrace and closeBrace respectively, e.g.

print "This is a double quotation mark: " + quotes

print "Here are the curly braces: {openBrace} and {closeBrace}"

Alternatively, you can insert their Unicode (codepoint) values (in decimal or hexadecimal) by means of function unicode:

print "Here are the curly braces: {unicode(123)} and {unicode(125)}"

print "This is an ASCII double quotation mark: " + unicode(0x22)

Character Sets

Any character except quote and control characters can be used in a String. Identifiers can only contain letters (without accents), numbers and underscore.

Tuple

• Holding a pair of x and y coordinates (each of Type Float) as a single unit.
• Allowing a function to pass back a result comprised of both a message in a String and a Boolean indicating whether the operation was successful.
• Returning both an updated copy of an immutable data structure and the value of an element from it, e.g. from method pop on a Stack.

Using a tuple

You may pass a tuple into a function, or return one from a function, for example:

An existing tuple (for example point1 below) may be deconstructed into new variables or named values (where the number of variables or names must match the number of elements in the tuple):

or into existing variables of the correct Type:

The 'discard' symbol _ (underscore) may also be used when deconstructing a tuple when there is no need to capture specific elements:

Notes

• As in most languages, an Elan tuple is immutable. Once defined they are effectively 'read only'. You cannot alter any of the elements in a tuple nor (unlike a List) can you create a new tuple from an existing one with specified differences.
• You cannot deconstruct a tuple into a mixture of new and existing variables.
• They may be nested: you can define a tuple within a tuple.

Standard (mutable) data structures

All four are 'mutable' – meaning that their contents may be changed directly by calling the various procedure dot methods that each Type defines. However, since procedure methods may be called only from within the main routine, or from within another procedure, it is also possible to make changes via function dot methods – which return a copy of the data structure, with the specified changes – which is why all such methods have names starting with....

All four data structures contain values of a single Type, that Type being specified either explicitly – as in <of Int> – or implicitly if the structure is created from its literal definition form.

A Dictionary is defined with two Types: one for its keys and one for their values.

Array

An array is a simple data structure containing n elements of a single Type, indexed from 0 to n-1.

The array elements' Type must be one of Int, Float, String, or Boolean.

As in most languages, individual elements may be referenced by their index.

To modify an element you may either call the put procedure method, or use the withPut function dot method.

An array's size is specified when it is created, along with the single value to which each element is initialised.

It is not possible to append further elements to an array, but you can convert an array to a list using method asList and then extend or reduce the list.

Procedure methods on an Array

Function dot methods on an Array

Array2D

The Array2D Type defines a 2-dimensional array of fixed size. The two dimensions, which are specified when the array is created, may be of the same size (a square array) or different (a rectangular array). If you want to create a 'jagged' array, you should use a List of Lists.

The convention is to refer to the position of an element in such an array by two index values (x,y), where x is the column number and y is the row number of a rectangular grid.

The Type for the elements is also specified when the Array is defined and must be one of: Int, Float, String, or Boolean.

Procedure methods on an Array2D

Function dot methods on an Array2D

• Since an Array2D has a known fixed size, methods rows and columns are of use only when passing such an array into a procedure or function that needs to discover its size.

You can read individual elements with a double index, for example:

List

A List is similar to an Array in that its n elements are all of one Type and are indexed from 0 to n-1, but is more flexible because:

• While the elements must all be of one Type, that Type is not limited to the simple value Types: the element Type may also be a user-defined Type such as a class, record, another List or an Array.
• A List is created either empty or initialised from a literal definition.
• A List's size can be dynamically changed.

List Deconstruction

A List or ListImmutable may be deconstructed into new variables or named values in a similar way to deconstructing a Tuple, but in this case:

• Always use two variable names.
• Use a colon : to separate the variable names.
• The first item in the list (the head) is assigned to the first variable.
• The rest of the list (the tail) is assigned to the second variable. If the original list only contains one element, this will be an empty list.
• Like deconstructing a Tuple, you can use an underscore _ as the variable name to discard either the head or the tail.

variable x:xs set to myList
set h:t to myList
set h:myList to myList

Discarding either the head or tail:

variable _:tail set to myList
variable head:_ set to myList

Procedure methods on a List

Function dot methods on a List

Dictionary

A Dictionary works like an Array, but instead of having a numeric (integer) index, each entry has a key. Each key is associated with a value, so the Dictionary is a set of key:value pairs. The key's Type must be one of Int, Float, String or Boolean, or it may be a user-defined record. The values may be of any Type, and the Types of both key and value are fixed when the Dictionary is created.

Procedure methods on a Dictionary

Function dot methods on a Dictionary

Immutable data structures

In contrast to the standard (mutable) data structures, immutable data structures cannot be modified directly – and hence define no procedure dot methods. Instead, changes are made by using function dot methods that copy the existing data structure, but with specified differences. Immutable data structures are intended specifically to facilitate the Functional Programming paradigm, but some are also useful within other programming paradigms.

The five immutable data structures defined in the Elan library are summarised in this table. More details on each Type are given below the table.

ListImmutable

A ListImmutable is like a List but is immutable (like a String). You can still insert, delete or change elements in a ListImmutable, but the methods for these operations do not modify the input ListImmutable: they return a new ListImmutable based on the input ListImmutable with the specified differences.

Type name

The Type is specified in the following ways:

• ListImmutable<of String> for a ListImmutable of Type String
• ListImmutable<of Int> for a ListImmutable of Type Int
• ListImmutable<of ListImmutable<of Int>> for a ListImmutable of such lists each of Type Int

Creating a ListImmutable

A ListImmutable may be defined in 'literal' form, delimited by curly brace with all the required elements separated by commas. The elements may be literal values but must all be of the same Type, for example:

ListImmutable Deconstruction

A ListImmutable can be deconstructed just like a List, using a colon : to separate two variable names.

set h:t to myListImmutable

Function dot methods on a ListImmutable

Try these examples:

DictionaryImmutable

A DictionaryImmutable is like a Dictionary but immutable (like a String). You can still insert, delete or change elements in a DictionaryImmutable, but the methods of these operations do not modify the input DictionaryImmutable: they return a new DictionaryImmutable with the specified differences.

A DictionaryImmutable may be defined in a constant.

Type name

In the following example, the keys are of Type String, and the values associated with the keys are of Type Int:

Defining a literal DictionaryImmutable

A literal DictionaryImmutable is defined as a comma-separated list of key:value pairs surrounded by curly braces:

Using an Immutable Dictionary

Try this example:

Function dot methods on a DictionaryImmutable

Set

A set is a standard data structure that works somewhat like a ListImmutable with the important difference that in a set a given element may appear only once. If an item being added to a Set is identical to an existing item in the Set then the Set remains the same length as before.

This enables a set to work like a mathematical set so that it is possible to perform standard set operations such as union or intersection. For the same reason, a Set is an immutable data structure: there are no methods that modify the set on which they are called, but several of them (including add, remove) return a new set that is based on the original set or sets, with specified differences.

Example of use:

Notes

• When creating a set, the Type of the elements must be specified in the form Set<of String>. This applies both when creating a new, empty set and when defining the Type of a parameter to be a Set.
• You can add elements: individually with add, or multiple elements with addFromList.
• You can create a new set from an existing list or ListImmutable by calling asSet on it.

Function dot methods on a Set

Stack and Queue

• Stack and Queue are similar data structures except that Stack is ‘LIFO’ (last in, first out), while Queue is FIFO (first in, first out). The names of the methods for adding/removing are different, but there are also common methods.
• Both a Stack and a Queue are defined with the Type of the items that they can contain, similarly to how List and ListImmutable have a specified item Type, though with different syntax. The Type is specified in the form shown below e.g. Stack<of String>, Queue<of Int>, Stack<of (Float, Float)>, Queue<of Square>.
• Both Stack and Queue are dynamically extensible, like List and ListImmutable. There is no need (or means) to specify a size limit as they will continue to expand until, eventually, the computer’s memory limit is reached.
• This same syntax is used to specify the Type if you want to pass a Stack or Queue into a function, or specify it as the return Type.
• Stack and Queue have two methods in common: length and peek. peek returns the next item to be removed, without actually removing it.
• The methods for adding and removing an item are different for Stack and Queue, as shown here:

Function dot methods on a Stack

Function dot methods on a Queue

Example use of a Stack:

Example use of a Queue:

Common dot methods

Dot methods that work on more than one fundamental Type

Although all applicable methods are described in earlier sections under each Type, this table lists those which are applicable to several Types.

Click on a Type to go to the Type's dot methods, or on the ✔ tick to go to the specific method on that Type.

Notes

• The method asString is so widely applicable because it enables you to print most variables and data structures to the Display for debugging purposes.
• The length of a Dictionary or DictionaryImmutable can be found by applying method length to the list returned by method keys on the dictionary:

print dict.keys().length() → 15

• Checking whether an item is contained in a Dictionary or DictionaryImmutable can be done by applying method contains to the list returned by methods keys and values on the dictionary:

print dict.keys().contains(42) → true or false

print dict.values().contains("widget") → true or false

Input/output

Reading keys ‘on the fly’

In some applications – especially in games, for example – you want the program to react to a key pressed by the user, but without holding up the program to wait for value to be input.

Whether your application makes use graphics, or just uses the Console for text, reading keystrokes ‘on the fly’ is done via one of two methods:

Notes

• When the getKey function is called, the system does not wait for a response. If a key has been pressed then that will be returned as a String e.g. "a".
• Non-printable keys will be returned in the form: "Backspace","Enter","ArrowDown",..
• If no key has been pressed (since the last time the method was called), it will return the empty string "".
• Pressing just the Shift, Ctrl (Cmd under macOS), or Alt keys will not be detected by getKey. To read those keys use…
• getKeyWithModifier which returns a 2-tuple containing the key pressed plus any ‘modifier’ key such a Shift, Ctrl, or Alt (or the empty string if no modifier key is pressed).
• Both of these getKey methods are System methods because they have a dependency on the system and so may only be used within a procedure or in main.

Use the procedure method clearKeyBuffer() if you want to enforce that the user cannot get too far ahead of the program by hitting keys in very rapid succession.

waitForKey waits for a key to be pressed, and returns it. pressAnyKeyToContinue gives an optional prompt and is used when you don't need to know which key was pressed.

Reading Text Files

The TextFileReader class is used to read textual data from a file.

An instance is created by the standalone system method openFileForReading.

The available procedure methods are:

These methods may be used to read a whole file in one go:

or to read a file line by line:

Notes

• openFileForReading will present the user with a dialog to select the file.
• readWholeFile returns a String containing every character in the file, without any trimming. It automatically closes the file after the read.
• readLine reads as far as the next newline character (\n) and then automatically trims the line to remove any spaces and/or carriage-returns (which some file systems insert after the newline automatically) from the resulting line returned as a String. If this behaviour is not desired, you can use readWholeFile, which does no trimming, and then parse the resulting String into separate lines.
• Calling file.close after reading line by line is strongly recommended to avoid any risk of leaving the file locked. It is not necessary to call it after using readWholeFile because that method automatically closes the file.
• Calling any method on a file that is already closed will result in a runtime error.

Writing text files

The TextFileWriter class is used to write textual data to a file.

An instance is created by the standalone system method createFileForWriting.

The available procedure methods are:

These methods may be used to write a whole file in one go:

or to write a file line by line:

Notes

• writeLine adds the string it is passed onto the end of any data previously written, with a newline character (\n) automatically appended.
• When execution reaches saveAndClose you will be presented with a dialog to confirm (or edit) the given filename and location where it is to be saved. It is not therefore strictly necessary to specify a filename when creating the file, since it can be specified by the user in the dialog so, in that case, you might put the empty string "" into the parameter of createFileForWriting.
• writeWholeFile puts the string it is given into the file and then automatically saves the file, so the user will be presented with the same dialog as if saveAndClose had been called.
• Calling any method on a file that has already been closed (by calling either saveAndClose or by writeWholeFile) will result in a runtime error.
• If the user were to hit Cancel on the save dialog, then the program will exit with an error. If you want to guard against this possibility (if, for example, it might mean the loss of important data) then you should perform the save and close within a try..catch like this:
+try call file.saveAndCloseprocedureName(arguments) +catch exception in evariableName print "File save cancelled"expression end try

or you could make the code offer the user options: to save again, or to continue without saving.

Displaying Html

If you attempt to embed Html in a string and then attempt to print it, you will see the string displayed literally - the Html tags will not be recognised - this is for security reasons. However, it is possible to display formatted Html on the Display. The following code:

will produce:

This Html forms another 'layer' of the Display. Any plain text that you print (using print or any of the print... procedures) will be overlaid on top of this.

You can programmatically clear just the Html display using the procedure clearHtml.

For specifying style or other attributes within Html tags, the attribute values should be enclosed in single quotation marks ' as shown above. Html will recognise single or double quotation marks, but entering double quotation marks would terminate the Elan string. Alternatively, you could use the constant quotes within braces as an interpolated field.

Using an embedded CSS stylesheet

You can also specify a style tag at the start of your Html string, to apply to the whole Html being displayed:

producing this result:

Displaying images

Printing an image on the Display

An image that can be accessed via a URL may be printed on the display using the image keyword followed by the URL. Here, the URL is not bounded by quotes:

As soon as you leave the field the text will change to show a thumbnail copy of the image:

print expression

If you edit the field again, the keyword and URL will be shown.

Acknowledgement: The idea of displaying an image value as a thumbnail within the code was inspired by a similar feature in the language Strype. We are grateful to Prof. Michael Kölling and the team at King's College, London responsible for Strype, who showed us this feature in confidence before its release, but nonetheless generously allowed us to mimic it in Elan.

When the code is run the image will be printed on the Display, but still thumbnail sized. You can specify dimensions by using a with clause like this:

The with clause also permits you to specify a title for the image, which shows up as a tooltip on the image. It is good practice to do this for meeting accessibility guidelines.The with clause also offers position properties x and y. These are ignored when printing an image, but are required in the context of drawing Vector Graphics.

An image specified in this way may also be assigned to a named value, or defined inline as an argument to a method, for example:

let sharkname be expression

The Type of a named value that holds an image is ImageVG – the 'VG' indicating that this Type is compatible with vector graphics, so an image may be added to a List<ofVectorGraphic> or displayed directly by:

In this usage the position may be controlled by specifying the x and y coordinates for the top left corner.

It is also possible to instantiate an ImageVG explicitly, but in this case you must provide the URL as a String, either as a literal (by surrounding it with quotes) or as a named value, for example:

Various properties can be set on an ImageVG using a with clause, for example:

The properties are listed in Vector Graphics Properties.

Graphics

Block graphics

Block graphics provides a simple way to create low resolution graphics, ideal for simple but engaging games for example.

The graphics are displayed on a grid that is 40 blocks wide by 30 blocks high.

Each block is be rendered as a solid colour.

An example of block graphics to produce a rapidly changing pattern of coloured blocks:

Notes

• The Array2D must be of TypeInt and of size 40 x 30.
• You may create multiple Array2Ds holding different patterns of blocks, and switch between them just by passing the required one as the argument to the displayBlocks method.
• A colour is specified as an Int, as described under Colours.

Turtle graphics

Turtle graphics are implemented in Elan with output to the Display pane on in the user interface, i.e. the 'paper' on which the Turtle draws.

The area is 200 turtle units wide by 150 turtle units high, and both integer and floating point values of turtle units can be used.

The origin for turtle units (0,0) is at the centre of the area: positive x rightwards, positive y upwards. So for example the top left corner of the display is at (-100,75).

Procedure methods on a Turtle

Properties of a Turtle

Example:

Output:

Notes

• If the turtle is placed or moved outside the 200 × 150 area boundary, it will not cause an error, but the location of the turtle and any lines outside the boundary will not be visible.
• You can move and turn the turtle, causing lines to be drawn, whether or not the turtle is shown.
• The current location and heading of the turtle may be read using the properties x, y, and heading.
• The turtle starts facing upwards (heading 0), so that move moves in the y-direction.

Here is a more sophisticated example, using a procedure and recursion, that produces a fractal snowflake:

Vector graphics

Vector graphics are implemented in Elan using the SVG (Scalable Vector Graphics) element that is part of the Html specification.

The names of the shapes broadly correspond to the names of SVG tags:

• CircleVG for <circle../>
• LineVG for <line../>
• RectangleVG for <rect../>
• ImageVG for <image../>

The properties of the Elan shapes reflect the names of the attributes used in the SVG tags but some differ. For example, SVG's stroke-width is set with property strokeWidth, to make it a valid Elan Identifier. Also stroke and fill are set by the properties strokeColour and fillColour.

Vector graphics are output to the Display pane in the user interface.

The area is 100 units wide by 75 units high, and both integer and floating point values of the units can be used.

The origin for SVG units (0,0) is a the top left corner of the display: positive x rightwards, positive y downwards. So for exxample the centre of the display is at (50, 37.5).

As with using SVG from Html, the shapes are drawn in the order in which they are added to the VectorGraphic instance: later shapes are positioned over earlier ones.

A set of SVG shapes is defined in either a List<of VectorGraphic> or a List of any specific Type of VectorGraphic (such as CircleVG) by using the List append method, and the assembled list is then displayed by a call to procedure displayVectorGraphics with the List as its argument. Examples are shown below the table of properties.

Properties of Vector graphic shapes

Note

The height and width of an image are dependent upon the dimensions of the original graphic file.

Example of a circle with a coloured circumference:

Example of a circle that changes between red and green every second:

Notes

• The constructor for each VG Type requires arguments corresponding to the Html attributes for the corresponding SVG Type.
• As with Block graphics the display is not updated until the displayVectorGraphics method is called, allowing you to make multiple changes before updating the display. Similarly, the method to add a shape returns a new instance of the VectorGraphics which must be assigned either to an existing variable, or to a new let.
• The fillColour and strokeColour properties may be specified as described under Colours. The fillColour only may also be specified as transparent (which has the value -1).
• VectorGraphic is the abstract superclass of all ...VG shapes. You use it if you want to define a List holding different types of shape.
• To make it easier for simple programs, the VectorGraphic shapes all have default values for their properties, and you only have to specify the ones that you want to differ. So the constructors do not take any properties, and instead you add a with clause listing those properties which you want to have different values. See New instance for more about with clauses.
• Individual properties of any of the VG Types may be modified by calling the corresponding set... procedure method or, if working within a function, by using the corresponding with... function dot method, for example setCentreX or withCentreX.

Combining graphic outputs

Program outputs, whether text or graphical, can be combined in the Display. In particular, Block graphics and text or Html printing can share the Display along with either Vector graphics or Turtle graphics (but not both).

If you want to share the Display in this way, remember that both text and Html print outputs appear sequentially down the Display (which can be scrolled), whereas the graphic outputs are positioned in the Display using their own absolute coordinate systems.

The order in which the outputs are displayed (and therefore overwrite) is:

1. Block graphics
2. Vector or Turtle graphics
3. Printed text or Html

So some care is needed to manage the layout in the Display.

Note that, while vector graphics are being drawn, printed text that exceeds the height of the display does not scroll until the graphic completes.

Other Types

Random

Generating random numbers within a function

It is not possible to use the system methods random() or randomInt() within a function because they create unseen side-effects. You may use those system methods outside the function and pass the resulting random number (as an Int or a Float) as an argument into a function.

It is possible to create and use random numbers within a function, but it requires a different approach and is a little more complex, using a special Type named Random (note that the R is in upper-case), as in this example:

The Random Type defines two ‘function methods’: next and nextInt.

Both of them return a 2-Tuple consisting of the random value (as either a Float or an Int respectively) plus a new Random. The new (returned) Random must be used for generating the subsequent random number (if more are required). If you call next repeatedly on the same instance of Random, you will always get the same value.

As shown in the example, when first created you should call initialiseWithClock() on it. If you remove that call statement from the code above, the program will still generate a sequence of randomised values, but the sequence will be exactly the same each time you run the program. Initialising from the clock ensures that you get a different sequence each run. Using Random without so initialising, however, can be extremely useful for testing purposes since the results are repeatable.

Func

A function may be passed as an argument into another function (or a procedure), or returned as the result of calling another function. This pattern is known as Higher-order Function (HoF), and is a key idea in the Functional Programming.

To define a function that takes in another function as a parameter, or returns a function, you need to specify the Type of the function, just as you would specify the Type of every parameter and the return Type for the function.

Type name

The Type of any function starts with the word Func followed by angle brackets defining the Type of each parameter, and the return Type for that function, following this syntax:

This example defines the Type for a function that defines three parameters of Type String, String, and Int respectively, and returns a Boolean value. This Type would match that of a function definition that started:

Constants

String constants

Boolean constants

Maths constant

Colours

A colour is specified as an Int value using one of these methods:

• the limited colours defined as library constants as in the above table.
• an integer in the decimal range 0 (black) to 2^24-1 (white).
• a six digit hexadecimal value in the range 0x000000 – 0xffffff using the same 'RGB' format as used in Html style, for example 0xff0000 for red.

Standalone functions

Standalone library functions always return a value and are therefore used in contexts that expect a value, such as in the right-hand side of a variable declaration or a set assignment, either on their own or within a more complex expression. All standalone library functions require at least one argument to be passed in brackets, corresponding to the parameters defined for that function.

unicode

Function unicode converts a Unicode (codepoint) value expressed as an integer value in decimal or hexadecimal notation, or as any expression evaluating to an Int, into a string of a single character. For example:

parseAsInt and parseAsFloat

Function parseAsInt attempts to parse the input String as an Int, and returns a 2-tuple, the first value of which is Boolean, with true indicating whether or not the parse has succeeded, and the second value being the resulting Int. parseAsFloat does the equivalent for floating point. Operation is illustrated by these tests:

Notes

• Any string that parses as an Int will also parse as a Float.
• If the parse fails, the second value will become zero, so you should always check the first value to see if the second value is a correct parse or just the default.
• You can 'deconstruct' the tuple into two variables: variable success, parsedValuename set to parseAsInt(myString)
• One use of these parsing methods is for validating user input, but note that an easier way to do this is to use the various input methods.

floor, ceiling, round, isNaN, and IsInfinite

All of these functions are called as 'dot methods' on a numeric value of Type Float or Int. NaN is short for 'Not A (Real) Number' Their use is illustrated with the following tests:

Maths functions

Examples of some maths functions and constants being tested:

Regular expressions

Elan's regular expressions are modelled on those of JavaScript, including the syntax for literal regular expressions. See, for example this Guide to Regular Expressions.

More functions for using regular expressions will be added in a future release of Elan. For now…

The method matchesRegExp is applied to a String using dot syntax and requires a RegExp parameter specified as a literal or as variable. It returns a Boolean. For example:

You can convert a valid string without /../ delimiters to a RegExp using function asRegExp:

Although it is recommended that literal regular expressions are written with /../ delimiters, the ability to convert a string allows you to input a regular expression into a running program.

Bitwise functions

These functions take in an integer value, and manipulate the bit representation of that value.

Examples of the bitwise functions being tested:

The result of bitNot(a) being -14 , when a is 13, might be a surprise. But this is because the bitwise functions assume that the arguments are represented as 32-bit signed binary integers. So 13 is represented as 00000000000000000000000000001101, and applying bitNot gives 11111111111111111111111111110010 which is the value -14 in signed two’s complement format, the left-most bit being the sign (0 positive, 1 negative).

Sequence

The sequence function is used to create a List containing a sequence of integer values as defined by the two (integer) parameters: start and end (both being inclusive values).

Here is an example which uses Higher-order Functions to print the prime numbers between 2 and 30:

Standalone procedures

Standalone procedures

All procedures are accessed via a call statement.

Method printTab helps in the layout of information printed to the Display, in particular, when printing columns of data. For example:

Right-align numeric output using a lambda function:

System methods

System methods

System methods appear to work like functions, because:

• they may require one or more arguments to be provided.
• they always return a value.
• they are used in expressions.

They are not, however, pure functions because:

• They may have a dependency on data that is not provided as an argument.
• They may generate side effects, such as changing the screen display or writing to a file.

Because of these properties, system methods may be used only within the main routine or within a procedure. They may not be used inside a function that you have defined, because to do so would mean that your function would not be pure.

System methods are all defined within the Elan standard library: you cannot write a system method yourself.

System methods are commonly associated with Input/output, but note that:

• Input/output may be performed in procedures as well as in main.
• Some system methods are not concerned with input/output, e.g. some generate random numbers.

The reason those are system methods is that they have a dependency on variable data that is not passed into them as arguments.

Note

• Methods random and randomInt cannot be used in a function because of their dependence on external factors, i.e. they are subject to side effects.
  To obtain random numbers within a function, use Type Random.

Higher-order Functions (HoFs)

A higher-order function (HoF) is one that takes in a reference to another function as a parameter, or (less commonly) that returns a reference to another function as its result.

Standard HoFs

The Elan standard library contains several HoFs that are widely recognised and used within functional programming.

Passing a function as a reference

To use a higher-order function (HoF), you have to pass in a reference to a function. This can be either a lambda or the keyword ref followed by the name of a function.

It is also possible to use references to functions not in the context of HoFs.

On most occasions when you write the name of an existing function elsewhere in code your intent is to evaluate the function and, to do so, you write the name of the function followed by brackets containing such arguments as are required by the function. For this reason if you forget to add the brackets, you will get an error, for example:

To evaluate function 'myfunction' add brackets. Or to create a reference to 'myfunction', precede it by 'ref'.

The second sentence in this error message is for when your intention is not to evaluate the function, but to create a reference to the function. This is a valid thing to do in functional programming but is not generally done in procedural programming. As the error message says, to create a reference to a function you need to precede it by ref and the name of the function should then not be followed by brackets (or any arguments). For example:

Defining your own HoFs

Library functions that process Lists

These dot methods are called on any List, ListImmutable or String. As Higher-order Functions they take either a lambda or a function reference as one of their arguments: see Passing a function as a reference.

Although several of these methods return a ListImmutable, they may be converted to a List using method asList.

These are not yet fully documented but, for readers familiar with HoFs from another programming language, some examples are shown below.

filter

Example function from demo program pathfinder.elan in which filter applies the lambda to the nodes (of class Node) to find one that contains p (of record Point):

map

Usage:

reduce

Usage:

maxBy and minBy

maxBy and minBy are dot functions of List and ListImmutable. They take a function as an argument, and return the list element corresponding to maximum or minimum of the values returned by the function. This prints "89" and "orange":

variable a set to {33, 4, 0, 92, 89, 55}
print a.maxBy(lambda x as Int => x mod 10)
variable fruit set to {"apple", "orange", "pear"}
print fruit.maxBy(lambda x as String => x.length())

For simply finding the maximum or minimum value in a List, you can use maxFloat, minFloat, maxInt or minInt.

sortBy

sortBy takes a lambda that takes two arguments (of the same Type as that of the ListImmutable being sorted) and compares them, returning an integer with one of the values -1, 0, or +1, to indicate whether the first argument should be placed respectively before, adjacent to or after the second argument in the sorted result, where ‘adjacent to’ means it does not matter whether before or after):

And similarly for string comparisons using methods isBefore and isAfter:

Elan Library Reference go to the top