General Information

This chapter contains general information about the Ring programming language

1. Ring Architecture
2. Memory Management
3. Data Representation

Ring Architecture

We have the next architecture

1. Ring Applications (Your Code) - Written in Ring - See folder : ring/applications
2. Ring Libraries (StdLib, WebLib, GameEngine, etc) - Written in Ring - See folder : ring/ringlibs
3. Ring Extensions (RingAllegro, RingQt, etc) - Written in C/C++ (may include Ring code) - See folder : ring/extensions
4. Ring Virtual Machine (Ring VM) - Written in C language
5. Operating System (Current Platform) - (Windows, Linux, macOS, Android, etc)

The extensions are just dynamic libraries (DLL, So, Dylib) You can update the extensions without the need to update your code.

Folder (ringlibdepwin) ====> C libraries used for building Ring Extensions (written in C) on Windows platform.

Folder (ringringlibs) ====> Ring libraries written in Ring itself (StdLib, WebLib, GameEngine, etc)

Folder (ringrnoteexe) ====> Just C source file to create executable file that will run rnote.ring

Folder (ringvisualsrc) ====> The Visual Source Code of the Ring Compiler & Ring VM developed using Programming Without Coding Technology (PWCT)

We use the term Ring Library —> When the library code is written in Ring We use the term Ring Extension —> When the library code is Written in C or C++

Memory Management

1. When we call function, we get a new scope for this function, inside this scope we store variables.

Also we get Temp. memory for this function. Inside this temp. memory we store temp. lists

All of this will be deleted directly after the end of the function call (end of scope)

2. In Ring to delete memory, you have the next options

2.1 Wait until the end of the function scope

2.2 Use the Assignment operator

2.3 Use callgc() function (delete temp. lists only)

3. Ring Garbage Collector uses Escape Analysis and Reference Counting

In 90% of cases, the Escape Analysis is used so we don’t waste time in running the garbage collector We directly know what will be deleted and what will remain in the memory.

https://en.wikipedia.org/wiki/Escape_analysis

In 10% of cases (or less than that) Ring may use Reference Counting.

For example when we pass a list and sub list to a function Ring will pass the lists by reference, but what will happens if we deleted the Parent List? In this case, the Sub List will use reference counting, and when deleting the Parent List, it will stay in memory until the end of the function.

Remember that Ring encourage us to avoid using references, and the Assignment Operator will copy lists by value, so Ring usage of reference counting is very limited to special cases and in most cases the Escape Analysis is enough which is very fast.

Starting from Ring 1.9 we extended the Reference Counting support to Ring Extensions and low level C pointers. So we don’t have to care about using fclose() when we use fopen() for example. and the same for other extensions like RingODBC, RingSQLite, RingMySQL, RingQt, etc.

All of the allocated resources will be cleaned by the Garbage Collector when we finish using it (when we lost the last reference).

Data Representation

1. In Ring, The String is just (Array of bytes)

Ring is 8-bit clean, Each character in the string is 8 bits (1 byte)

So these functions (Int2Bytes(), Float2Bytes() and Double2Bytes()) just return a string.

Also we can store binary data in strings

mystring = read("myfile.exe")

2. Remember this, when you think about variables

• Value —> What we have (What we are storing in the computer memory as data) - Low Level Concept
• Type —> What we can do with what we have or how we do things with what we have (Just a Logical Concept)

Computer memory —-> Just store [Bytes] - Each byte is 8-bit - (Here we avoid the memory word concept)

These bytes could be grouped together when moved between the memory and the processor registers. Here we have (The register size) and things like 32-bit and 64-bit for example. Also we have the bytes order.

Programming Languages —-> Add Types (Just as a concept) to these bytes so we can determine what to do with them and how operations should be done.

And programming language could allow (Type Conversion) —> Because the Type is a logical concept in most cases, What we really have is just data (Bytes, Bytes Count, Bytes Order, etc)

Ring Stings —-> You have these bytes (each byte is 8-bit) and Ring know the string size (stored as number in the String structure)

So we don’t check the NULL character or add it to the end of the string (Not Required)

All operations inside Ring VM, will check the Ring size and deal with the string as binary data (each character is 8-bit)

In the C language —> The normal case is adding NULL character (0) to the end of each string

And the string functions check this character, This is not suitable for binary data.

Signed vs Unsigned —> Is a logical concept which is important when you do arithmetic operations on the data, but when storing the data, if you will include all of the (8-bits) and will not ignore any of them —> Then don’t care.

In Ring, don’t think about these details, we are hiding it from you, so you can focus on your application and what you will do.

Think in C when you write C code which could be (based on need) low level code to have control on everything.

—-> Good for performance and memory management

Think in Ring when you write Ring code which let you ignore a lot of details and concentrate only on the result —–> Good for productivity and delivering software quickly

The good news (We can mix between Ring and C in our projects)

3. The functions Int2Bytes(), Float2Bytes() and Double2Bytes()

These function take input as (Number) —> Convert it to group of bytes based on the number type (int|float|double) —> Then return a Ring string that contains these bytes

Int2Bytes() —> Ring string (Group of bytes) and the string size = sizeof(int)

Float2Bytes() —> Ring string (Group of bytes) and the string size = sizeof(float)

Double2Bytes() —> Ring string (Group of bytes) and the string size = sizeof(double)

Example:

? len( int2bytes(1) )
? len( float2bytes(1) )
? len( double2bytes(1) )

Output:

4
4
8

4. Storing Numbers

When we use a number, Ring always use the (Double) data type for representing these numbers in memory. This is important to know when we do arithmetic operations on numbers.

But when we convert the number to a String using “” + number or using string(number) we get a string where each digit is represented in 1 byte (Not good idea for storage, but useful for string processing)

If you need the number to be represented in specific size (int|float|double) for storage then use bytes2int() , bytes2float() and bytes2double() when writing the data to binary files.

Ring Number (double) —-> int2bytes() - will cast the number from double to int then return the bytes —-> 4 bytes (Ring String)

Ring Number (double) —-> float2bytes() - will cast the number from double to float then return the bytes —-> 4 bytes (Ring String)

Ring Number (double) —-> double2bytes() - will use the number (double) to return the bytes —-> 8 bytes (Ring String)

The (int) type is used only for internal Ring operations, but Ring applications|code will use only the (double) type for numbers.

5. The Unsigned() Function

The function unsigned() expect the first and the second parameters as numbers

unsigned(nNumber1,nNumber2,cOperator)

We can use the bytes2int() function to convert the bytes to a number

Example:

B = list(4)

for k=1 to 4
{
        B[k]= Space(4)
        for kk=1 to 4 { B[k][kk]= char(60+4*k +kk) }
        ? " B" +k +": " +B[k]
}

A12= Space(4)     A12= bytes2int(B[1]) ^ bytes2int(B[2])
? "A12: " +string(A12)
A34= Space(4)     A34= bytes2int(B[3]) ^ bytes2int(B[4])
? "A34: " +string(A34)
A12= space(4)     A12= Unsigned(bytes2int(B[1]),bytes2int(B[2]),"^")
? "unsigned A12: " +A12
A34= space(4)     A34= Unsigned(bytes2int(B[3]),bytes2int(B[4]),"^")
? "unsigned A34: " +A34

Output:

B1: ABCD
B2: EFGH
B3: IJKL
B4: MNOP
A12: 201589764
A34: 470025220
unsigned A12: 201589764
unsigned A34: 470025220