Another outcome of Language Agnostic Code Audit SaaS I’m currently working on is a PropWise library.
It scans the project content, and does its best on detecting and reporting places suitable for property-based testing, with examples using StreamData or PropEr.
An excerpt from docs:
PropWise analyzes your Elixir code to find functions that would benefit from property-based testing. It examines the Abstract Syntax Tree (AST) of your code to:
▸ Detect pure functions (functions without side effects)
▸ Identify common patterns suitable for property testing
▸ Find inverse function pairs (encode/decode, serialize/deserialize, etc.)
▸ Score and rank candidates by testability
▸ Provide specific testing suggestions for each candidate
The example of ouroborosing itself:
❯ mix propwise
Compiling 8 files (.ex)
Generated propwise app
Analyzing ....
================================================================================
PropWise Analysis Report
================================================================================
Summary:
Total functions analyzed: 115
Property test candidates: 30
Candidates dropped (below threshold): 76
Coverage: 26.1%
--------------------------------------------------------------------------------
Top Candidates (sorted by score):
--------------------------------------------------------------------------------
PropWise.Reporter.format_json_report/1
Score: 10
Location: lib/prop_wise/reporter.ex:133
Type: private
Patterns:
- Numeric Algorithm: Arithmetic operations
- Data Transformation: Map transformation
- Collection Operation: Uses Enum collection operations
Testing suggestions:
- property "returns numeric result" do
check all n <- one_of([integer(), float()]) do
result = Reporter.format_json_report(n)
assert is_number(result)
end
end
- property "handles zero and negative inputs" do
check all n <- one_of([integer(), float()]) do
# Verify the function doesn't crash on edge-case numeric inputs.
_ = Reporter.format_json_report(n)
end
end
…
Enjoy!
