Merge pull request #210 from rtibblesbot/issue-209-cf2cc6

Fix language regex to allow multi-subtag codes like zh-Hans-CN

Author: bjester

js/EmbedContentRequest.js

@@ -18,7 +18,7 @@ export const SCHEMA = {
     "language": {
       "type": "string",
       "description": "Language code from https://github.com/learningequality/le-utils/blob/main/le_utils/resources/languagelookup.json",
-      "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)?$"
+      "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)*$"
     },
     "url": {
       "type": "string",

js/EmbedTopicsRequest.js

@@ -26,7 +26,7 @@ export const SCHEMA = {
     "language": {
       "type": "string",
       "description": "Language code from https://github.com/learningequality/le-utils/blob/main/le_utils/resources/languagelookup.json",
-      "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)?$"
+      "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)*$"
     },
     "level": {
       "type": "integer",

le_utils/constants/embed_content_request.py

@@ -24,7 +24,7 @@
         "language": {
             "type": "string",
             "description": "Language code from https://github.com/learningequality/le-utils/blob/main/le_utils/resources/languagelookup.json",
-            "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)?$",
+            "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)*$",
         },
         "url": {
             "type": "string",

le_utils/constants/embed_topics_request.py

@@ -29,7 +29,7 @@
         "language": {
             "type": "string",
             "description": "Language code from https://github.com/learningequality/le-utils/blob/main/le_utils/resources/languagelookup.json",
-            "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)?$",
+            "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)*$",
         },
         "level": {
             "type": "integer",

spec/schema-embed_content_request.json

@@ -13,7 +13,7 @@
     "language": {
       "type": "string",
       "description": "Language code from https://github.com/learningequality/le-utils/blob/main/le_utils/resources/languagelookup.json",
-      "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)?$"
+      "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)*$"
     },
     "url": {
       "type": "string",

spec/schema-embed_topics_request.json

@@ -21,7 +21,7 @@
     "language": {
       "type": "string",
       "description": "Language code from https://github.com/learningequality/le-utils/blob/main/le_utils/resources/languagelookup.json",
-      "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)?$"
+      "pattern": "^[a-z]{2,3}(?:-[a-zA-Z]+)*$"
     },
     "level": {
       "type": "integer",

tests/test_schemas.py

@@ -2,6 +2,8 @@
 from __future__ import unicode_literals
 
 import contextlib
+import json
+import pkgutil
 import uuid
 
 import pytest
@@ -61,6 +63,14 @@ def _validate_completion_criteria(data):
     )
 
 
+def _load_language_codes():
+    """Return all language codes from languagelookup.json."""
+    data = pkgutil.get_data("le_utils", "resources/languagelookup.json")
+    codes = list(json.loads(data).keys())
+    assert len(codes) > 0, "languagelookup.json should not be empty"
+    return codes
+
+
 @contextlib.contextmanager
 def _assert_not_raises(not_expected):
     try:
@@ -479,6 +489,31 @@ def test_embed__topics__missing_channel_id():
         )
 
 
+@skip_if_jsonschema_unavailable
+def test_embed__topics__all_languagelookup_codes_valid():
+    """Every language code in languagelookup.json must be accepted by the topics schema.
+
+    Regression test for https://github.com/learningequality/le-utils/issues/209
+    where codes like zh-Hans-CN were rejected.
+    """
+    test_id = str(uuid.uuid4())
+    for code in _load_language_codes():
+        with _assert_not_raises(jsonschema.ValidationError):
+            _validate_embed_topics_request(
+                {
+                    "topics": [
+                        {
+                            "id": test_id,
+                            "channel_id": test_id,
+                            "title": "Test topic",
+                            "description": "Test description",
+                            "language": code,
+                        }
+                    ],
+                }
+            )
+
+
 @skip_if_jsonschema_unavailable
 def test_embed__content__valid():
     with _assert_not_raises(jsonschema.ValidationError):
@@ -703,6 +738,34 @@ def test_embed__content__invalid_preset_files():
         )
 
 
+@skip_if_jsonschema_unavailable
+def test_embed__content__all_languagelookup_codes_valid():
+    """Every language code in languagelookup.json must be accepted by the content schema.
+
+    Regression test for https://github.com/learningequality/le-utils/issues/209
+    where codes like zh-Hans-CN were rejected.
+    """
+    test_id = str(uuid.uuid4())
+    for code in _load_language_codes():
+        with _assert_not_raises(jsonschema.ValidationError):
+            _validate_embed_content_request(
+                {
+                    "resources": [
+                        {
+                            "id": test_id,
+                            "channel_id": test_id,
+                            "title": "Resource title",
+                            "description": "Resource description",
+                            "text": "Resource text",
+                            "language": code,
+                            "content_id": test_id,
+                            "channel_version": 1,
+                        },
+                    ],
+                }
+            )
+
+
 def _validate_learning_objectives(data):
     """
     :param data: Dictionary of data to validate