ReqCassette.
View Source
Normalizes request and response data for predictable template variable extraction.
To ensure consistent positional indexing with {{var.0}}, {{var.1}}, etc., this module
normalizes data before extraction by sorting JSON object keys and query parameters
alphabetically, while preserving the semantic order of arrays and URI paths.
Why Normalization Matters
Positional markers like {{sku.0}} rely on consistent extraction order. Without
sorting, minor changes in JSON key order or query param order would break template matching.
Problem without normalization
# Recording
Request: {"product": "SKU-1234", "category": "tools"}
Extracted order: ["SKU-1234"] # JSON parser happened to iterate product first
Template: {{sku.0}} = "SKU-1234"
# Replay - same data, different key order
Request: {"category": "tools", "product": "SKU-5678"}
Extracted order: Could differ!
Template: {{sku.0}} might not match ❌Solution with normalization
# Both recording and replay
1. Sort JSON keys: {"category": "tools", "product": "SKU-XXXX"}
2. Extract in sorted order: consistently finds product SKU
3. Positional markers work reliably ✅Normalization Rules
- ✅ JSON object keys - Sorted alphabetically (recursive)
- ✅ Query parameters - Sorted alphabetically by name
- ❌ JSON arrays - Order preserved (arrays are ordered data structures)
- ❌ URI path - Order preserved (paths have semantic ordering)
Examples
# JSON normalization
iex> normalize_json(%{"z" => 1, "a" => 2, "m" => 3})
%{"a" => 2, "m" => 3, "z" => 1}
iex> normalize_json(%{"b" => %{"y" => 1, "x" => 2}, "a" => 3})
%{"a" => 3, "b" => %{"x" => 2, "y" => 1}}
# Arrays preserve order
iex> normalize_json([3, 1, 2])
[3, 1, 2]
# Query string normalization
iex> normalize_query_string("z=1&a=2&m=3")
"a=2&m=3&z=1"
iex> normalize_query_string("")
""Usage
This module is used internally during template extraction to ensure consistent ordering:
# During recording
normalized_request = Normalizer.normalize_request(request)
variables = Extractor.extract(normalized_request, patterns)
# During replay
normalized_request = Normalizer.normalize_request(request)
variables = Extractor.extract(normalized_request, patterns)
# Same normalization = same extraction order = reliable matching ✅Summary
Functions
Normalizes JSON data by sorting object keys alphabetically (recursive).
Normalizes a query string by sorting parameters alphabetically.
Normalizes a request map for predictable template extraction.
Normalizes a response map for predictable template extraction.
Functions
Normalizes JSON data by sorting object keys alphabetically (recursive).
Arrays preserve their order. Only maps (JSON objects) are sorted.
Parameters
data- JSON data (map, list, or primitive)
Returns
Normalized data with sorted object keys
Examples
iex> normalize_json(%{"c" => 1, "a" => 2, "b" => 3})
%{"a" => 2, "b" => 3, "c" => 1}
iex> normalize_json(%{"outer" => %{"z" => 1, "a" => 2}})
%{"outer" => %{"a" => 2, "z" => 1}}
iex> normalize_json([%{"b" => 1, "a" => 2}, %{"d" => 3, "c" => 4}])
[%{"a" => 2, "b" => 1}, %{"c" => 4, "d" => 3}]
iex> normalize_json([3, 1, 2])
[3, 1, 2]
iex> normalize_json("text")
"text"
iex> normalize_json(123)
123Normalizes a query string by sorting parameters alphabetically.
Parameters
query_string- Query string like "z=1&a=2"
Returns
Sorted query string like "a=2&z=1"
Examples
iex> normalize_query_string("z=1&a=2&m=3")
"a=2&m=3&z=1"
iex> normalize_query_string("single=value")
"single=value"
iex> normalize_query_string("")
""
iex> normalize_query_string("b=2&a=1&a=3")
"a=1&a=3&b=2"Normalizes a request map for predictable template extraction.
Sorts JSON bodies and query parameters while preserving URI path and array order.
Parameters
request- Request map with keys: "method", "uri", "querystring", "headers", "body*"
Returns
Normalized request map with sorted JSON and query params
Examples
iex> request = %{
...> "method" => "GET",
...> "uri" => "https://api.example.com/path",
...> "query_string" => "z=1&a=2",
...> "body_json" => %{"z" => 1, "a" => 2}
...> }
iex> normalized = normalize_request(request)
iex> normalized["query_string"]
"a=2&z=1"
iex> normalized["body_json"]
%{"a" => 2, "z" => 1}Normalizes a response map for predictable template extraction.
Sorts JSON bodies while preserving array order.
Parameters
response- Response map with keys: "status", "headers", "body_*"
Returns
Normalized response map with sorted JSON
Examples
iex> response = %{
...> "status" => 200,
...> "body_json" => %{"z" => 1, "a" => 2}
...> }
iex> normalized = normalize_response(response)
iex> normalized["body_json"]
%{"a" => 2, "z" => 1}