PyDAXLexer 0.3.0

PyDAX is designed to analyze DAX, it can extract comments, remove comments, and identify columns and measures referenced in DAX expressions.

PyDAX

PyDAX is a package designed to analyze DAX expressions. It can extract comments, remove comments, and identify columns and measures referenced in DAX expressions.

Installation

To install the package, use pip:

pip install PyDAXLexer

Usage

Here's how to use PyDAXLexer to analyze a DAX expression and surface helpful insights.

The example below intentionally violates a few best-practice rules (notably Unused Variables and FILTER patterns), and demonstrates how to:

• Extract comments and remove comments
• Extract table/column/measure references
• Verify best practices and list violating tokens
• Generate HTML highlighting violations

from PyDAX import DAXExpression

if __name__ == '__main__':
    # Intentionally problematic DAX (for demo purposes):
    # Two unused variables
    # FILTER on table+column inside CALCULATE
    # FILTER on table with measure predicate inside CALCULATE
    dax_expression = """
    // Demo measure with intentional violations
    VAR UnusedVar1 = 123
    VAR SalesPerCustomer = SUM(Sales[Amount]) / COUNTROWS(VALUES(Customers[CustomerID]))
    VAR UnusedVar2 = IFERROR(SUM('Sales'[DiscountAmount]), 0)
    RETURN
    CALCULATE(
        [Total Sales],
        FILTER('Sales', 'Sales'[Quantity] > 10), // column filter (violation)
        FILTER('Sales', [Total Sales] > 1000) // table + measure filter (violation)
    )
    """

    # Initialize the analyzer
    expression = DAXExpression(dax_expression)

    #Comments and comment-free expression
    print("Comments:", expression.comments)
    print("Expression without comments:", expression.dax_expression_no_comments)

    #Table/Column/Measure references
    print("Table/Artifact references:")
    for ref in expression.table_column_references:
        # DAXReference(table_name: str, artifact_name: str)
        print(f" - Table='{ref.table_name}' Artifact='{ref.artifact_name}'")

    #Best practices
    expression.print_best_practices_violations()
    print("Total best-practice violations:", expression.number_of_violations)

    # HTML highlighting (optional)
    # html_code = expression.generate_html_with_violations(name="Demo")
    # expression.save_html_with_violations_to_file("demo_violations.html", name="Demo")

Additional Features

The DAXExpression class provides several utility methods:

• remove_comments(): Returns the expression with comments removed (accessible as dax_expression_no_comments).
• extract_comments(): Returns a list of comment strings (accessible as comments).
• extract_artifact_references(): Returns table/column/measure references as DAXReference objects (accessible as table_column_references).
• generate_html(light: bool): Generates HTML output with syntax coloring.
• generate_html_with_violations(name: str, light: bool): Generates HTML and highlights best-practice violations.
• save_html_to_file(file_name: str): Saves the syntax-colored HTML output to a file.
• save_html_with_violations_to_file(file_name: str): Saves the violations-highlighted HTML output to a file.

Best-practices overview

When DAXExpression is created, it initializes a set of best-practice rules and verifies them by default. You can access:

• best_practice_rules: List of rule instances
• number_of_violations: Total count of violations across all rules
• print_best_practices_violations(): Print rule names and violating tokens with locations

Included rules (subject to change):

• Use DIVIDE instead of division operator
• Avoid IFERROR
• Use TREATAS instead of INTERSECT
• Filter column values with proper syntax
• Filter measure values by columns, not tables
• Unused variables
• Avoid 1 - x/y syntax
• Avoid EvaluateAndLog in production

License

This project is licensed under the MIT License.

Acknowledgments

TabularEditor

The lexer grammar used in this project is adapted from the TabularEditor GitHub repository.

• Source: TabularEditor GitHub Repository
• License: MIT License

ANTLR

This project uses ANTLR (ANother Tool for Language Recognition) to generate the lexer and parser for DAX expressions.

• Project: The ANTLR Project
• License: BSD-3-Clause License