From 1b77de9e0051c199153a49eb210c9c65c7eb523a Mon Sep 17 00:00:00 2001
From: Ash Vardanian <1983160+ashvardanian@users.noreply.github.com>
Date: Sun, 1 Dec 2024 09:55:04 +0000
Subject: [PATCH] Add: Checksums in Python

---
 python/lib.c    | 35 ++++++++++++++++++++++++++++++++++-
 scripts/test.py | 12 +++++++++++-
 2 files changed, 45 insertions(+), 2 deletions(-)

diff --git a/python/lib.c b/python/lib.c
index b6f57646..ba435fdb 100644
--- a/python/lib.c
+++ b/python/lib.c
@@ -714,7 +714,40 @@ static PyObject *Str_like_hash(PyObject *self, PyObject *args, PyObject *kwargs)
     }
 
     sz_u64_t result = sz_hash(text.start, text.length);
-    return PyLong_FromSize_t((size_t)result);
+    return PyLong_FromUnsignedLongLong((unsigned long long)result);
+}
+
+static char const doc_like_checksum[] = //
+    "Compute the checksum of individual byte values in a string.\n"
+    "\n"
+    "This function can be called as a method on a Str object or as a standalone function.\n"
+    "Args:\n"
+    "  text (Str or str or bytes): The string to hash.\n"
+    "Returns:\n"
+    "  int: The checksum of individual byte values in a string.\n"
+    "Raises:\n"
+    "  TypeError: If the argument is not string-like or incorrect number of arguments is provided.";
+
+static PyObject *Str_like_checksum(PyObject *self, PyObject *args, PyObject *kwargs) {
+    // Check minimum arguments
+    int is_member = self != NULL && PyObject_TypeCheck(self, &StrType);
+    Py_ssize_t nargs = PyTuple_Size(args);
+    if (nargs < !is_member || nargs > !is_member + 1 || kwargs) {
+        PyErr_SetString(PyExc_TypeError, "checksum() expects exactly one positional argument");
+        return NULL;
+    }
+
+    PyObject *text_obj = is_member ? self : PyTuple_GET_ITEM(args, 0);
+    sz_string_view_t text;
+
+    // Validate and convert `text`
+    if (!export_string_like(text_obj, &text.start, &text.length)) {
+        wrap_current_exception("The text argument must be string-like");
+        return NULL;
+    }
+
+    sz_u64_t result = sz_checksum(text.start, text.length);
+    return PyLong_FromUnsignedLongLong((unsigned long long)result);
 }
 
 static char const doc_like_equal[] = //
diff --git a/scripts/test.py b/scripts/test.py
index 82f13376..93a01706 100644
--- a/scripts/test.py
+++ b/scripts/test.py
@@ -767,7 +767,7 @@ def test_translations(length: int):
 
 
 @pytest.mark.repeat(3)
-@pytest.mark.parametrize("length", range(1, 300))
+@pytest.mark.parametrize("length", list(range(0, 300)) + [1024, 4096, 100000])
 @pytest.mark.skipif(not numpy_available, reason="NumPy is not installed")
 def test_translations_random(length: int):
     body = get_random_string(length=length)
@@ -775,6 +775,16 @@ def test_translations_random(length: int):
     assert sz.translate(body, memoryview(lut)) == baseline_translate(body, lut)
 
 
+@pytest.mark.repeat(3)
+@pytest.mark.parametrize("length", list(range(0, 300)) + [1024, 4096, 100000])
+def test_checksums_random(length: int):
+    def sum_bytes(body: str) -> int:
+        return sum([ord(c) for c in body])
+
+    body = get_random_string(length=length)
+    assert sum_bytes(body) == sz.checksum(body)
+
+
 @pytest.mark.parametrize("list_length", [10, 20, 30, 40, 50])
 @pytest.mark.parametrize("part_length", [5, 10])
 @pytest.mark.parametrize("variability", [2, 3])