update test to include levenshtein difference

tests/extraction/test_invoice_splitter_auto_extraction.py

@@ -9,6 +9,7 @@
 from mindee.product import InvoiceSplitterV1, InvoiceV4
 from tests.product import get_id, get_version
 from tests.test_inputs import PRODUCT_DATA_DIR
+from tests.utils import levenshtein_ratio
 
 
 @pytest.fixture
@@ -52,4 +53,4 @@ def test_pdf_should_extract_invoices_strict():
         PRODUCT_DATA_DIR / "invoices" / "response_v4" / "summary_full_invoice_p1.rst",
         invoice_0.document,
     )
-    assert test_string_rst_invoice_0 == str(invoice_0.document)
+    assert levenshtein_ratio(test_string_rst_invoice_0, str(invoice_0.document)) >= 0.97

tests/utils.py

@@ -25,3 +25,54 @@ def dummy_envvars(monkeypatch) -> None:
 
 
 EXTRAS_DIR = Path("./tests/data/extras/")
+
+
+def levenshtein_distance(reference_str: str, target_str: str) -> int:
+    """
+    Calculate the Levenshtein distance between two strings.
+
+    The Levenshtein distance is a measure of the difference between two sequences.
+    Informally, the Levenshtein distance between two words is the minimum number
+    of single-character edits (insertions, deletions or substitutions) required
+    to change one word into the other.
+
+
+    :param reference_str: The reference string.
+    :param target_str: The target string.
+
+    :return: The distance between the two strings.
+    """
+    reference_len, target_len = len(reference_str), len(target_str)
+    previous_row = list(range(target_len + 1))
+    current_row = [0] * (target_len + 1)
+
+    for i in range(reference_len):
+        current_row[0] = i + 1
+
+        for j in range(target_len):
+            deletion_cost = previous_row[j + 1] + 1
+            insertion_cost = current_row[j] + 1
+            substitution_cost = previous_row[j] if reference_str[i] == target_str[j] else previous_row[j] + 1
+
+            current_row[j + 1] = min(deletion_cost, insertion_cost, substitution_cost)
+
+        previous_row, current_row = current_row, previous_row
+
+    return previous_row[target_len]
+
+
+def levenshtein_ratio(ref_str: str, target_str: str) -> float:
+    """
+    Calculates the Levenshtein ratio between two strings.
+
+    :param ref_str: Reference string.
+    :param target_str: Target String.
+    :return: Ratio between the two strings
+    """
+    lev_distance = levenshtein_distance(ref_str, target_str)
+    max_len = max(len(ref_str), len(target_str))
+
+    if max_len == 0:
+        return 1.0
+
+    return 1.0 - (lev_distance / max_len)