From 7447f8664b7d985a24d12279c6852529664762de Mon Sep 17 00:00:00 2001
From: Titusz Pan <titusz.pan@gmail.com>
Date: Thu, 8 Feb 2024 13:36:42 +0100
Subject: [PATCH] Add support for MSCDI validation

---
 CHANGELOG.md           | 1 +
 docs/changelog.md      | 1 +
 iscc_core/codec.py     | 5 ++---
 iscc_core/constants.py | 4 ++++
 tests/test_codec.py    | 9 +++++++++
 5 files changed, 17 insertions(+), 3 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index e7bec2c..656f4ee 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -2,6 +2,7 @@
 
 ## [1.0.9] - Unreleased
 - Updated dependencies
+- Added support for MSCDI validation
 
 ## [1.0.8] - 2024-01-30
 - Added implementors guide to README.md
diff --git a/docs/changelog.md b/docs/changelog.md
index e7bec2c..656f4ee 100644
--- a/docs/changelog.md
+++ b/docs/changelog.md
@@ -2,6 +2,7 @@
 
 ## [1.0.9] - Unreleased
 - Updated dependencies
+- Added support for MSCDI validation
 
 ## [1.0.8] - 2024-01-30
 - Added implementors guide to README.md
diff --git a/iscc_core/codec.py b/iscc_core/codec.py
index 51d6ed3..37b269d 100644
--- a/iscc_core/codec.py
+++ b/iscc_core/codec.py
@@ -1,6 +1,5 @@
 # -*- coding: utf-8 -*-
 import math
-import re
 import uvarint
 from typing import List, Tuple
 import base58
@@ -511,10 +510,10 @@ def iscc_validate(iscc, strict=True):
     """
 
     # Basic regex validation
-    match = re.match("^ISCC:[A-Z2-7]{10,60}$", iscc)
+    match = CANONICAL_REGEX.match(iscc)
     if not match:
         if strict:
-            raise ValueError("ISCC string does not match ^ISCC:[A-Z2-7]{10,60}$")
+            raise ValueError("ISCC string does not match ^ISCC:[A-Z2-7]{10,68}$")
         else:
             return False
 
diff --git a/iscc_core/constants.py b/iscc_core/constants.py
index 2fc9e9d..7e9f16f 100644
--- a/iscc_core/constants.py
+++ b/iscc_core/constants.py
@@ -1,4 +1,5 @@
 # -*- coding: utf-8 -*-
+import re
 import enum
 from typing import Tuple, Union
 
@@ -171,6 +172,7 @@ class LN(enum.IntEnum):
     L192 = 192
     L224 = 224
     L256 = 256
+    L320 = 320
 
 
 class MULTIBASE(str, enum.Enum):
@@ -245,3 +247,5 @@ class MULTIBASE(str, enum.Enum):
     "MM",
     "OA",
 ]
+
+CANONICAL_REGEX = re.compile("^ISCC:[A-Z2-7]{10,68}$")
diff --git a/tests/test_codec.py b/tests/test_codec.py
index fd35b93..89e0c4f 100644
--- a/tests/test_codec.py
+++ b/tests/test_codec.py
@@ -637,3 +637,12 @@ def test_iscc_validate_bad_length():
     with pytest.raises(ValueError) as excinfo:
         ic.iscc_validate(sample, strict=True)
     assert str(excinfo.value) == "Header expects 32 but got 26 bytes"
+
+
+def test_iscc_validate_mscdi():
+    sample = "ISCC:KEDRRHYRYJ7XELW7HAO5FFGQRX75HJUKSUSZVWTTRNHTF2YL5SKP7XIUFXM4KMKXEZZA"
+    assert ic.iscc_validate(sample, strict=False) is True
+    assert ic.iscc_validate(sample, strict=True) is True
+    with pytest.raises(ValueError) as excinfo:
+        ic.iscc_validate(sample + "A", strict=True)
+    assert str(excinfo.value) == "ISCC string does not match ^ISCC:[A-Z2-7]{10,68}$"