Merge branch 'release/3.6.5'

Author: mgoldman-globality

.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 3.6.4
+current_version = 3.6.5
 commit = False
 tag = False
 

microcosm_flask/fields/__init__.py

@@ -2,7 +2,7 @@
 Custom fields.
 
 """
-from microcosm_flask.fields.ascii_encoded_string_field import AsciiEncodedStringField  # noqa: F401
+from microcosm_flask.fields.cleaned_field import CleanedField  # noqa: F401
 from microcosm_flask.fields.enum_field import EnumField  # noqa: F401
 from microcosm_flask.fields.language_field import LanguageField  # noqa: F401
 from microcosm_flask.fields.query_string_list import QueryStringList  # noqa: F401

…ask/fields/ascii_encoded_string_field.py microcosm_flask/fields/cleaned_field.pymicrocosm_flask/fields/ascii_encoded_string_field.py renamed to microcosm_flask/fields/cleaned_field.py microcosm_flask/fields/ascii_encoded_string_field.py renamed to microcosm_flask/fields/cleaned_field.py

@@ -1,16 +1,17 @@
-from marshmallow import fields
-
-
 """
-AsciiEncodedStringField Marshmallow field
-An extension of the marshmallow fields.String that removes unicode
+CleanedField Marshmallow field
+
+An extension of the marshmallow fields.String that removes unicode control
 characters on deserialization
 """
 
+import regex
+from marshmallow import fields
+
 
-class AsciiEncodedStringField(fields.String):
+class CleanedField(fields.String):
     def _deserialize(self, value, *args, **kwargs):
         if value is not None:
-            value_encoded = value.encode("ascii", "ignore")
-            value = value_encoded.decode()
+            # remove control characters
+            value = regex.sub(r'\p{C}', '', value)
         return super()._deserialize(value, *args, **kwargs)

microcosm_flask/swagger/parameters/default.py

@@ -6,7 +6,7 @@
 from marshmallow.fields import Field
 from microcosm_logging.decorators import logger
 
-from microcosm_flask.fields import AsciiEncodedStringField, LanguageField, URIField
+from microcosm_flask.fields import CleanedField, LanguageField, URIField
 from microcosm_flask.swagger.parameters.base import ParameterBuilder
 
 
@@ -16,7 +16,7 @@
 FIELD_MAPPINGS = {
     LanguageField: FieldInfo("string", "language"),
     URIField: FieldInfo("string", "uri"),
-    AsciiEncodedStringField: FieldInfo("string", "ascii-encoded-string"),
+    CleanedField: FieldInfo("string", "cleaned"),
     fields.Boolean: FieldInfo("boolean", None),
     fields.Date: FieldInfo("string", "date"),
     fields.DateTime: FieldInfo("string", "date-time"),

microcosm_flask/tests/fields/test_ascii_string.py

@@ -0,0 +1,36 @@
+"""
+Test cleaned field.
+
+"""
+from hamcrest import assert_that, equal_to, is_
+from marshmallow import Schema
+
+from microcosm_flask.fields import CleanedField
+
+
+printable_ascii = "~}|{zyxwvutsrqponmlkjihgfedcba`_^]\\[ZYXWVUTSRQPONMLKJIHGFEDCBA@?>="\
+    + "<;:9876543210/.-,+*)(\'&%$#\"! "
+control_chars = "\u0000\u0001\u0002\u0003\u0004\u0005\u0006\u0007\u0008\u0009\u000a"\
+    + "\u000b\u000c\u000d\u000e\u000f\u0010\u0011\u0012\u0013\u0014\u0015\u0016\u0017"\
+    + "\u0018\u0019\u001a\u001b\u001c\u001d\u001e\u001f"
+spanish_chars = "áéíóúüñ¿¡ÁÁÉÍÓÚÜÑ"
+portuguese_chars = "ãáàâçéêíõóôúüÃÁÀÂÇÉÊÍÕÓÔÚÜ"
+
+comment_text = printable_ascii + control_chars + spanish_chars + portuguese_chars
+comment_text_cleaned = printable_ascii + spanish_chars + portuguese_chars
+
+
+class CleanedStringSchema(Schema):
+    str = CleanedField(required=True)
+
+
+def test_load():
+    schema = CleanedStringSchema()
+
+    result = schema.load(dict(
+        str=comment_text,
+    ))
+    assert_that(
+        result["str"],
+        is_(equal_to(comment_text_cleaned)),
+    )

microcosm_flask/tests/fields/test_cleaned_field.py

@@ -3,7 +3,7 @@
 
 
 project = "microcosm-flask"
-version = "3.6.4"
+version = "3.6.5"
 
 
 setup(
@@ -33,6 +33,7 @@
         "python-dateutil>=2.7.3",
         "PyYAML>=3.13",
         "rfc3986>=1.2.0",
+        "regex>=2021.8.21",
     ],
     extras_require={
         "metrics": "microcosm-metrics>=2.2.0",