Sedona

Primitives

Overview

The Sedona programming language contains the following built-in primitive types:

  • bool: boolean value
  • byte: unsigned 8-bit integer (as field or array only)
  • short: unsigned 16-bit integer (as field or array only)
  • int: signed 32-bit integer
  • long: signed 64-bit integer
  • float: 32-bit floating point
  • double: 64-bit floating point

In addition to the primitive types, the following class types have special language support for literal representation:

  • Str: string of ASCII characters terminated by 0 (like C string)
  • Buf: chunk of bytes in memory

These types are described in more detail below.

Bool

The bool type stores a boolean variable. Boolean literals are expressed using the true, false, and null keywords. Use null to indicate an invalid boolean value. If used in a boolean expression, null will evaluate to true (it is represented as 2 in memory).

A bool is stored in fields and arrays as an unsigned 8-bit integer. During stack manipulation bools are stored on the stack as signed 32-bit integers.

The representation for booleans:

Value Binary String
false 0 "false"
true 1 "true"
null 2 "null"

Integers

There are four integer types of varying widths:

  • byte: unsigned 8-bit integer (as field or array only)
  • short: unsigned 16-bit integer (as field or array only)
  • int: signed 32-bit integer
  • long: signed 64-bit integer

Both byte and short are special types that may only be used as fields or in arrays. Attempting to use byte or short as a return type, parameter type, or local variable type is a compiler error. Note that unlike Java both byte and short are unsigned. Currently there is no signed 8-bit or 16-bit integer type.

All integer operations on the SVM stack are performed using signed 32-bit integers. When a byte or short is loaded from a field or array it is automatically expanded into a 32-bit signed value. Likewise when it is stored back into a field or array it is narrowed from a 32-bit signed value.

Integer literals are decimal by default. If prefixed with "0x" they are hexadecimal. You may use the underbar "_" as separator in both decimal and hexadecimal formats. To specify a 64-bit long value, you must append an "L" to the number.

Char escapes

You may also use single quotes to specify a character as an integer literal. The following character escape sequences are supported:

      \0   zero/null terminator
      \n   newline
      \r   carriage return
      \t   horizontal tab
      \"   double quote
      \'   single quote
      \\   backslash
      \$   dollar sign ($ is used for str interpolation)
Examples of integer literals:

8
-78
0xABCD
10_000
0xffff_ffff
'x'
'\n'
0L
0x1234_5678_aabb_ccddL

Floating Point

The float type maps to a 32-bit floating point value and double to 64-bit floating point.

Floating point literals are expressed in decimal format using a '.' dot as the decimal point. The 'F' or 'f' character may be used as a suffix (required if not using a decimal point). The 'D' or 'd' character is required as a suffix for a 64-bit double. You may use the '_' underbar as a separator.

You can also specify floating point literals in scientific notation. All numbers given in scientific notation are of type float unless explicitly marked as a double using 'D' or 'd'. A floating-point literal has the following format:

The keyword null is used to represent not-a-number for situations requiring indication of an invalid float or double. The string representation for null floats and doubles is always 'null'. The '==' operator will return true when comparing two null floating point values (this is different from Java and IEEE). The evaluation of arithmetic and comparison operations with null operands, however, is unspecified for the Sedona VM.

3f
3.0
40_000F
-2.00D
0d
1e+5
5.86e12d
314159E-5
null

Time

The Sedona Framework represents time in nanosecond ticks, stored as a 64-bit long. When working with time, you can use a special literal representation for longs using the following suffixes on a decimal number:

Suffix Unit Nanosecond Multiplier
ns nanoseconds 1
ms milliseconds 1,000,000
sec seconds 1,000,000,000
min minutes 60,000,000,000
hr hours 3,600,000,000,000
days days 86,400,000,000,000

Examples of long time literals and what they represent:

5ns          // 5L
1ms          // 1_000_000L
10sec        // 10_000_000_000L
3min         // 180_000_000_000L
12hr         // 43_200_000_000_000L
0.5ms        // 500_000L
0.001sec     // 1_000_000L
0.25min      // 15_000_000_000L
0.5days      // 43_200_000_000_000L
1days        // 86_400_000_000_000L
36500days    // 31_53_600_000_000_000_000L

Str

The sys::Str class models a string of ASCII characters. Strings are stored in memory like C strings using a null terminator (a byte with a value of zero). You should use only 7-bit ASCII characters (clear high bit) to allow future UTF-8 and Unicode support.

In the SVM, the Str class makes use of the unsized class feature to create a byte[] of the correct length when the Str object is instantiated. No other fields are declared, so an instance of Str is stored in memory just like a byte[]. You can also treat a Str reference as a normal C string (char*) when writing native methods.

Because Str is an unsized class, you must specify the length of the string when declaring a Str field. For example to declare a Str that can hold a max of 8 characters (including the null terminator):

    Str someStr           // reference to Str stored elsewhere
    inline Str(8) myStr   // storage allocated here for 8 byte Str

Sedona also supports string interpolation when writing to an output stream.

Str literals are written using double quotes. You may use supported escape sequences for special characters inside the quotes. All Str literals are interned when compiling a scode image - this means that all Str literals with the same sequence of characters will share the same reference. Str literals should be considered read-only memory - never try to change the contents of a Str literal.

Examples of Str literals:

"hello"
"Hi there.\nHow are you?"

The compiler automatically adds the null terminator byte when interning the literal. For example a pointer to the literal "abc" is really a pointer to four bytes of memory containing "abc\0".

Buf

The sys::Buf class models a contiguous chunk of bytes in memory. Like sys::Str, it is an unsized class containing a byte[] that is allocated to the specified size when the Buf object is created. Unlike Str, however, Buf does not treat its contents as a string, so no null terminator is added. It also has fields that store the size of the buffer and the number of bytes used.

The syntax for a Buf literal is 0x[hexDigits]. You can use whitespace (including newlines) between bytes. For example:

static Buf literalA = 0x[cafe babe 03 dead beef]

Just like Str literals, Buf literals are interned and stored in scode memory space. So you should never attempt to write into a Buf literal\'s memory space - for example never try to set the bytesLen field or change the contents of the bytes field.

Array Literals

Although they are not free-form expressions, you can also declare array literals in code:

define Str[] colors = {"red", "green", "blue"}

See Array Literals for more details.


Last update: April 28, 2020