Running mypy on SDK resources (following up #3995)

opentelemetry-api/src/opentelemetry/attributes/__init__.py

@@ -11,7 +11,6 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-# type: ignore
 
 import logging
 import threading
@@ -31,7 +30,7 @@
 
 def _clean_attribute(
     key: str, value: types.AttributeValue, max_len: Optional[int]
-) -> Optional[types.AttributeValue]:
+) -> Optional[Union[types.AttributeValue, tuple[str | int | float, ...]]]:
     """Checks if attribute value is valid and cleans it if required.
 
     The function returns the cleaned value or None if the value is not valid.
@@ -59,7 +58,7 @@ def _clean_attribute(
         cleaned_seq = []
 
         for element in value:
-            element = _clean_attribute_value(element, max_len)
+            element = _clean_attribute_value(element, max_len)  # type: ignore
             if element is None:
                 cleaned_seq.append(element)
                 continue
@@ -110,7 +109,7 @@ def _clean_attribute(
 
 def _clean_attribute_value(
     value: types.AttributeValue, limit: Optional[int]
-) -> Union[types.AttributeValue, None]:
+) -> Optional[types.AttributeValue]:
     if value is None:
         return None
 
@@ -126,7 +125,7 @@ def _clean_attribute_value(
     return value
 
 
-class BoundedAttributes(MutableMapping):
+class BoundedAttributes(MutableMapping):  # type: ignore
     """An ordered dict with a fixed max capacity.
 
     Oldest elements are dropped when the dict is full and a new element is
@@ -149,28 +148,32 @@ def __init__(
         self.dropped = 0
         self.max_value_len = max_value_len
         # OrderedDict is not used until the maxlen is reached for efficiency.
-        self._dict = {}  # type: dict | OrderedDict
-        self._lock = threading.RLock()  # type: threading.RLock
+
+        self._dict: (
+            MutableMapping[str, types.AttributeValue]
+            | OrderedDict[str, types.AttributeValue]
+        ) = {}
+        self._lock = threading.RLock()
         if attributes:
             for key, value in attributes.items():
                 self[key] = value
         self._immutable = immutable
 
-    def __repr__(self):
+    def __repr__(self) -> str:
         return f"{dict(self._dict)}"
 
-    def __getitem__(self, key):
+    def __getitem__(self, key: str) -> types.AttributeValue:
         return self._dict[key]
 
-    def __setitem__(self, key, value):
-        if getattr(self, "_immutable", False):
+    def __setitem__(self, key: str, value: types.AttributeValue) -> None:
+        if getattr(self, "_immutable", False):  # type: ignore
             raise TypeError
         with self._lock:
             if self.maxlen is not None and self.maxlen == 0:
                 self.dropped += 1
                 return
 
-            value = _clean_attribute(key, value, self.max_value_len)
+            value = _clean_attribute(key, value, self.max_value_len)  # type: ignore
             if value is not None:
                 if key in self._dict:
                     del self._dict[key]
@@ -179,23 +182,23 @@ def __setitem__(self, key, value):
                 ):
                     if not isinstance(self._dict, OrderedDict):
                         self._dict = OrderedDict(self._dict)
-                    self._dict.popitem(last=False)
+                    self._dict.popitem(last=False)  # type: ignore
                     self.dropped += 1
 
-                self._dict[key] = value
+                self._dict[key] = value  # type: ignore
 
-    def __delitem__(self, key):
-        if getattr(self, "_immutable", False):
+    def __delitem__(self, key: str) -> None:
+        if getattr(self, "_immutable", False):  # type: ignore
             raise TypeError
         with self._lock:
             del self._dict[key]
 
-    def __iter__(self):
+    def __iter__(self):  # type: ignore
         with self._lock:
-            return iter(self._dict.copy())
+            return iter(self._dict.copy())  # type: ignore
 
-    def __len__(self):
+    def __len__(self) -> int:
         return len(self._dict)
 
-    def copy(self):
-        return self._dict.copy()
+    def copy(self):  # type: ignore
+        return self._dict.copy()  # type: ignore

opentelemetry-sdk/src/opentelemetry/sdk/resources/__init__.py

@@ -63,6 +63,8 @@
 import typing
 from json import dumps
 from os import environ
+from types import ModuleType
+from typing import List, Optional, Sequence, Union
 from urllib import parse
 
 from opentelemetry.attributes import BoundedAttributes
@@ -75,10 +77,14 @@
 from opentelemetry.util._importlib_metadata import entry_points, version
 from opentelemetry.util.types import AttributeValue
 
+psutil: Optional[ModuleType] = None
+
 try:
-    import psutil
+    import psutil as psutil_module
+
+    pustil = psutil_module
 except ImportError:
-    psutil = None
+    pass
 
 LabelValue = AttributeValue
 Attributes = typing.Mapping[str, LabelValue]
@@ -141,12 +147,15 @@
 TELEMETRY_AUTO_VERSION = ResourceAttributes.TELEMETRY_AUTO_VERSION
 TELEMETRY_SDK_LANGUAGE = ResourceAttributes.TELEMETRY_SDK_LANGUAGE
 
-_OPENTELEMETRY_SDK_VERSION = version("opentelemetry-sdk")
+_OPENTELEMETRY_SDK_VERSION: str = version("opentelemetry-sdk")
 
 
 class Resource:
     """A Resource is an immutable representation of the entity producing telemetry as Attributes."""
 
+    _attributes: BoundedAttributes
+    _schema_url: str
+
     def __init__(
         self, attributes: Attributes, schema_url: typing.Optional[str] = None
     ):
@@ -173,7 +182,7 @@ def create(
         if not attributes:
             attributes = {}
 
-        resource_detectors = []
+        resource_detectors: List[ResourceDetector] = []
 
         resource = _DEFAULT_RESOURCE
 
@@ -184,29 +193,29 @@ def create(
         if "otel" not in otel_experimental_resource_detectors:
             otel_experimental_resource_detectors.append("otel")
 
+        resource_detector: str
         for resource_detector in otel_experimental_resource_detectors:
             resource_detectors.append(
                 next(
                     iter(
                         entry_points(
                             group="opentelemetry_resource_detector",
                             name=resource_detector.strip(),
-                        )
+                        )  # type: ignore
                     )
                 ).load()()
             )
-
         resource = get_aggregated_resources(
             resource_detectors, _DEFAULT_RESOURCE
         ).merge(Resource(attributes, schema_url))
 
         if not resource.attributes.get(SERVICE_NAME, None):
             default_service_name = "unknown_service"
-            process_executable_name = resource.attributes.get(
+            process_executable_name: Optional[Union[int, float, Sequence[str], Sequence[int], Sequence[float] ]] = resource.attributes.get(
                 PROCESS_EXECUTABLE_NAME, None
             )
             if process_executable_name:
-                default_service_name += ":" + process_executable_name
+                default_service_name += ":" + str(process_executable_name)
             resource = resource.merge(
                 Resource({SERVICE_NAME: default_service_name}, schema_url)
             )
@@ -241,8 +250,8 @@ def merge(self, other: "Resource") -> "Resource":
         Returns:
             The newly-created Resource.
         """
-        merged_attributes = self.attributes.copy()
-        merged_attributes.update(other.attributes)
+        merged_attributes = self.attributes.copy()  # type: ignore
+        merged_attributes.update(other.attributes)  # type: ignore
 
         if self.schema_url == "":
             schema_url = other.schema_url
@@ -257,8 +266,7 @@ def merge(self, other: "Resource") -> "Resource":
                 other.schema_url,
             )
             return self
-
-        return Resource(merged_attributes, schema_url)
+        return Resource(merged_attributes, schema_url) # type: ignore
 
     def __eq__(self, other: object) -> bool:
         if not isinstance(other, Resource):
@@ -268,15 +276,15 @@ def __eq__(self, other: object) -> bool:
             and self._schema_url == other._schema_url
         )
 
-    def __hash__(self):
-        return hash(
-            f"{dumps(self._attributes.copy(), sort_keys=True)}|{self._schema_url}"
-        )
+    def __hash__(self) -> int:
+        attributes_json = dumps(self._attributes.copy(), sort_keys=True)  # type: ignore
+        return hash(f"{attributes_json}|{self._schema_url}")
 
-    def to_json(self, indent=4) -> str:
+    def to_json(self, indent: int = 4) -> str:
+        attributes = dict(self._attributes)  # type: ignore
         return dumps(
             {
-                "attributes": dict(self._attributes),
+                "attributes": attributes,  # type: ignore
                 "schema_url": self._schema_url,
             },
             indent=indent,
@@ -294,7 +302,7 @@ def to_json(self, indent=4) -> str:
 
 
 class ResourceDetector(abc.ABC):
-    def __init__(self, raise_on_error=False):
+    def __init__(self, raise_on_error: bool = False) -> None:
         self.raise_on_error = raise_on_error
 
     @abc.abstractmethod
@@ -365,16 +373,17 @@ def detect(self) -> "Resource":
             resource_info[PROCESS_PARENT_PID] = os.getppid()
 
         if psutil is not None:
-            process = psutil.Process()
-            resource_info[PROCESS_OWNER] = process.username()
+            process: psutil_module.Process = psutil.Process()
+            username = process.username()
+            resource_info[PROCESS_OWNER] = username
 
-        return Resource(resource_info)
+        return Resource(resource_info)  # type: ignore
 
 
 def get_aggregated_resources(
     detectors: typing.List["ResourceDetector"],
     initial_resource: typing.Optional[Resource] = None,
-    timeout=5,
+    timeout: int = 5,
 ) -> "Resource":
     """Retrieves resources from detectors in the order that they were passed
 

tox.ini

@@ -126,7 +126,7 @@ setenv =
   ; i.e: CONTRIB_REPO_SHA=dde62cebffe519c35875af6d06fae053b3be65ec tox -e <env to test>
   CONTRIB_REPO_SHA={env:CONTRIB_REPO_SHA:main}
   CONTRIB_REPO=git+https://github.com/open-telemetry/opentelemetry-python-contrib.git@{env:CONTRIB_REPO_SHA}
-  mypy: MYPYPATH={toxinidir}/opentelemetry-api/src/:{toxinidir}/tests/opentelemetry-test-utils/src/
+  mypy: MYPYPATH={toxinidir}/opentelemetry-api/src/:{toxinidir}/opentelemetry-semantic-conventions/src/:{toxinidir}/opentelemetry-sdk/src/:{toxinidir}/tests/opentelemetry-test-utils/src/
 
 commands_pre =
 
@@ -314,7 +314,9 @@ commands =
 
   coverage: {toxinidir}/scripts/coverage.sh
 
+  mypy: mypy --version
   mypy: mypy --install-types --non-interactive --namespace-packages --explicit-package-bases opentelemetry-api/src/opentelemetry/
+  mypy: mypy --install-types --non-interactive --namespace-packages --explicit-package-bases opentelemetry-sdk/src/opentelemetry/sdk/resources
   mypy: mypy --install-types --non-interactive --namespace-packages --explicit-package-bases opentelemetry-semantic-conventions/src/opentelemetry/semconv/
 
 ; For test code, we don't want to enforce the full mypy strictness