fix: Comments and typos

Author: n0nSmoker

.cursorrules

@@ -61,7 +61,6 @@ Never use bare `except:`. Log at appropriate levels (DEBUG, INFO, WARNING, ERROR
 **Project-Specific:**
 - Logger: `logging.getLogger("serializer")` with WARN level by default
 - Custom exception: `IsNotSerializable` for unserializable types
-- Iterable serialization swallows `IsNotSerializable` exceptions (see FIXME in code)
 - Use debug logging for serialization flow: `logger.debug("Serialize key:%s type:%s", ...)`
 
 ## Code Practices
@@ -133,9 +132,7 @@ Review all AI-generated code—especially complex logic and security-sensitive c
 ## Known Issues & Workarounds
 
 **Known Issues:**
-1. Iterable serialization swallows `IsNotSerializable` exceptions (see FIXME in `serialize_iter`)
-2. Schema checks can be optimized: TODO comments about skipping checks when not greedy
-3. Custom serializers cannot access format or tzinfo (documented limitation in README)
+1. Custom serializers cannot access format or tzinfo (documented limitation in README)
 
 **Common Pitfalls:**
 - One-element tuples must have trailing comma: `serialize_only = ('field',)` not `('field')`

docs/ARCHITECTURE.md

@@ -385,21 +385,14 @@ graph TD
 - Strict: Explicit control, security (don't leak fields)
 - `serialize_only` switches to strict mode
 
-### 6. Swallowing Exceptions in Iterables
-**Decision**: Catch `IsNotSerializable` in `serialize_iter()` and continue
-**Rationale**:
-- Allows partial serialization of mixed-type collections
-- Prevents one bad item from breaking entire serialization
-- **Known Issue**: May hide legitimate errors (see FIXME in code)
-
-### 7. LRU Cache for Field Introspection
+### 6. LRU Cache for Field Introspection
 **Decision**: Cache `get_serializable_keys()` results
 **Rationale**:
 - SQLAlchemy inspection is relatively expensive
 - Field names don't change at runtime
 - Significant performance improvement for repeated serialization
 
-### 8. Options Pattern
+### 7. Options Pattern
 **Decision**: Use `namedtuple` for serialization options
 **Rationale**:
 - Immutable configuration

docs/DEVELOPMENT.md

@@ -391,7 +391,6 @@ print(serializer.schema._tree)  # Inspect rule tree
 - Cache key is model instance (works because field names don't change)
 
 **Schema optimization**:
-- TODO: Skip `is_included()` checks when not greedy (see code comments)
 - Consider caching schema lookups for repeated serialization
 
 **Type checking optimization**:

docs/KNOWLEDGE.md

@@ -298,16 +298,6 @@ The library caches field names using `@functools.lru_cache`:
 
 **Impact**: Significant performance improvement for repeated serialization of same model type.
 
-### Schema Checks
-
-**Current behavior**: `is_included()` is called for every field, even in strict mode.
-
-**Optimization opportunity** (TODO in code):
-- Skip `is_included()` checks when not greedy (strict mode)
-- In strict mode, only iterate over `schema.keys`
-
-**Impact**: Minor performance improvement for strict mode serialization.
-
 ### Type Checking
 
 **Current implementation**: `isinstance(value, types)` with tuple of types.
@@ -525,16 +515,6 @@ class MyModel(Base, SerializerMixin):
 
 **Note**: This is a documented limitation. May be implemented in future.
 
-### Iterable Serialization Swallows Errors
-
-**Symptom**: Some items missing from serialized list, no error
-
-**Cause**: `serialize_iter()` catches `IsNotSerializable` and continues
-
-**Behavior**: By design - allows partial serialization of mixed-type collections
-
-**Workaround**: Check logs for warnings, or serialize items individually to catch errors.
-
 ## API Decisions
 
 ### Why Mixin Instead of Base Class?
@@ -587,17 +567,6 @@ class MyModel(Base, SerializerMixin):
 - Easy to exclude sensitive fields with negative rules
 - Can switch to strict mode when needed
 
-### Why Swallow Exceptions in Iterables?
-
-**Decision**: Catch `IsNotSerializable` in `serialize_iter()` and continue
-
-**Rationale**:
-- Allows partial serialization of mixed-type collections
-- Prevents one bad item from breaking entire serialization
-- Common use case: list with mostly serializable items
-
-**Trade-off**: May hide legitimate errors (see FIXME in code)
-
 ### Why LRU Cache for Field Introspection?
 
 **Decision**: Cache `get_serializable_keys()` results

sqlalchemy_serializer/serializer.py

@@ -333,7 +333,7 @@ def serialize_iter(self, value: Iterable) -> list:
     def serialize_dict(self, value: dict) -> dict:
         res = {}
         for k, v in value.items():
-            if self.schema.is_included(k):  # TODO: Skip check if is NOT greedy
+            if self.schema.is_included(k):
                 logger.debug("Serialize key:%s type:%s of dict", k, get_type(v))
 
                 result = self.serialize(value=v, key=k)
@@ -352,7 +352,7 @@ def serialize_model(self, value) -> dict:
             keys.update(get_serializable_keys(value))
 
         for k in keys:
-            if self.schema.is_included(key=k):  # TODO: Skip check if is NOT greedy
+            if self.schema.is_included(key=k):
                 v = getattr(value, k)
                 logger.debug(
                     "Serialize key:%s type:%s of model:%s",

tests/test_serializer_serialize_model__function.py

@@ -14,7 +14,7 @@ def test_model(get_instance):
 @pytest.mark.parametrize(
     "only, expected",
     [
-        # FIXME: last `key` ignored and all the rist is returned is it expected?
+        # FIXME: last `key` ignored and all the rest is returned is it expected?
         # (("model.list.key",), {"model": {"list": [{"key": 123}]}}),
         (("dict.key",), {"dict": {"key": 123}}),
         (("bool",), {"bool": False}),