Feature request

Is your feature request related to a problem? Please describe.

Add mechanism to validate Abstract Syntax Tree got from Tokenizer::parse() to ensure that used variables has been declared, etc.

Describe the solution you'd like 🌟

New class or method in Tokenizer class.

Describe alternatives you've considered 🐾

None

Additional context 👍:

Feature request

Is your feature request related to a problem? Please describe.

Create syntax analyzer module for building an Abstract Syntax Tree.

Describe the solution you'd like 🌟

Node system:

• Create class for syntax node.
• Each node should contain 3 or more operators, value and its type.
• Type must be represented as enumeration.
• Syntax tree building can use recursive methods.

Describe alternatives you've considered 🐾

None

Additional context 👍:

Graphical representation of tree

Source code

while b ≠ 0
  if a > b
    a := a − b
  else
    b := b − a
return a

Describe the bug 👀

On error tokenizer does not release allocated memory for nodes.

How to reproduce 🏃:

1. Try to compile incorrect program.
2. See that nodes are not deleted after exception throwing.

Expected behavior 🙌

On error tokenizer must release allocated memory.

Screenshots ✨

System information

Desktop 💻:

• OS: Windows 10
• Browser None
• Version latest

Smartphone 📱:

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context 🌞

Feature request

Is your feature request related to a problem? Please describe.

Add multi-character operators support for Lexer.
Example:

• <<
• ==
• !=
• ...

Describe the solution you'd like 🌟

Token with type OPERATOR and wstring payload.

Describe alternatives you've considered 🐾

Analyzing several tokens in syntax analyzer.

Additional context 👍:

Describe the bug 👀

Incorrect number reading with Lexer in scientific notation.

How to reproduce 🏃:

1. Try to assign number 20e+20 to the variable.
2. Compile.
3. See error.

Expected behavior 🙌

I expected that 20e+23 is a correct number notation, but Lexer interprets that like several lexemes.

Screenshots ✨

System information

Desktop 💻:

• OS: Windows
• Browser None
• Version latest

Smartphone 📱:

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context 🌞

Feature request

Is your feature request related to a problem? Please describe.

Add validation errors array to allow compiler throw several errors per one compilation attempt.

Describe the solution you'd like 🌟

A vector of errors. After validation it's length must be checked. If this array has errors - then output them to stderr and throw exception.

Describe alternatives you've considered 🐾

Just one error per compilation.

Additional context 👍:

Describe the bug 👀

Tokenizer interprets math expressions wrongly. It must apply operations by operators priority. We can use Reverse polish nonation (RPN) to interpret math expressions correctly.

How to reproduce 🏃:

1. Try to build syntax tree for int test = 2 + 2 * 2 / 4;
2. Get result tree:

└─[op2] VAR: DeclarationNode | INT | test
   └─[op1] DIVIDE: Node
       └─[op1] MULTIPLY: Node
           └─[op1] ADD: Node
               └─[op1] CONST: IntegerNode | 2
               └─[op2] CONST: IntegerNode | 2
           └─[op2] CONST: IntegerNode | 2
       └─[op2] CONST: IntegerNode | 4

Expected behavior 🙌

Expected syntax tree:

└─[op2] VAR: DeclarationNode | INT | test
   └─[op1] ADD: Node
       └─[op1] DIVIDE: Node
           └─[op1] MULTIPLY: Node
               └─[op1] CONST: IntegerNode | 2
               └─[op2] CONST: IntegerNode | 2
           └─[op2] CONST: IntegerNode | 4
       └─[op2] CONST: IntegerNode | 2

Screenshots ✨

System information

Desktop 💻:

• OS: Windows 10
• Browser None
• Version latest

Smartphone 📱:

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context 🌞