Numeric Property Test Specifications

Purpose

Test parsing of property names that start with numbers.

Input

{
    1property = 'value1',
    2ndProperty = 'value2',
    3rdItem = 'value3'
}

Expected Result

• Property "1property" = "value1"
• Property "2ndProperty" = "value2"
• Property "3rdItem" = "value3"

Notes

Property names starting with digits are valid in TON.

Purpose

Test parsing of property names that are purely numeric.

Input

{
    123 = 'value123',
    456 = 'value456',
    789 = 'value789'
}

Expected Result

• Property "123" = "value123"
• Property "456" = "value456"
• Property "789" = "value789"

Notes

Pure numeric property names are treated as strings internally.

Purpose

Test using year values as property names.

Input

{
    2022 = 450000000,
    2023 = 520000000,
    2024 = 380000000
}

Expected Result

• Property "2022" = 450000000
• Property "2023" = 520000000
• Property "2024" = 380000000

Use Case

Common pattern for year-based data (revenues, statistics, etc.).

Purpose

Test document with mix of numeric and regular property names.

Input

{
    name = 'John',
    123 = 'numeric',
    age = 30,
    2024 = 'year',
    active = true
}

Expected Result

• 5 properties total
• Property "name" = "John"
• Property "123" = "numeric"
• Property "age" = 30
• Property "2024" = "year"
• Property "active" = true

Notes

Numeric and regular names can coexist in same object.

Purpose

Test numeric properties in nested object structures.

Input

{
    financials = {
        revenue = {
            2022 = 450000000,
            2023 = 520000000,
            2024 = 380000000
        },
        expenses = {
            2022 = 400000000,
            2023 = 450000000,
            2024 = 320000000
        }
    }
}

Expected Result

• Root has "financials" property
• financials.revenue.2022 = 450000000
• financials.revenue.2023 = 520000000
• financials.revenue.2024 = 380000000
• financials.expenses.2022 = 400000000
• financials.expenses.2023 = 450000000
• financials.expenses.2024 = 320000000

Purpose

Test that numeric property names serialize correctly.

Input Object

Properties:
  "2022" = 100
  "2023" = 200
  "2024" = 300
  "123" = "test"
  "regular" = "value"

Expected Serialization

{
    2022 = 100,
    2023 = 200,
    2024 = 300,
    123 = "test",
    regular = "value"
}

Notes

Numeric names don't need quotes in serialization.

Purpose

Test edge cases with zero and negative numeric property names.

Input

{
    0 = 'zero',
    00 = 'double zero',
    01 = 'zero one',
    10 = 'ten'
}

Expected Result

• Property "0" = "zero"
• Property "00" = "double zero"
• Property "01" = "zero one"
• Property "10" = "ten"

Notes

Leading zeros preserved as distinct property names.

Purpose

Test property names that look like floating point numbers.

Input

{
    3.14 = 'pi',
    2.718 = 'e',
    1.0 = 'one point zero'
}

Expected Result

• Property "3.14" = "pi"
• Property "2.718" = "e"
• Property "1.0" = "one point zero"

Notes

Decimal points allowed in property names.

Purpose

Test property names that resemble scientific notation.

Input

{
    1e3 = 'thousand',
    1e6 = 'million',
    1.5e2 = 'one fifty'
}

Expected Result

• Property "1e3" = "thousand"
• Property "1e6" = "million"
• Property "1.5e2" = "one fifty"

Purpose

Test programmatic access to numeric properties.

Operations

object.GetProperty("123")  // String key
object.GetProperty("2024") // Year as string
object["456"]              // Bracket notation

Expected Result

• All access methods work with string keys
• No automatic numeric conversion
• Consistent behavior across access patterns