fix: CI lint complaints

I didn't touch any of these Python files in my PR which makes me think
either the linting changed recently or PRs that fail CI have been
getting merged. Anyway, happy to fix them up for the greater good :)

Makefile

@@ -324,7 +324,7 @@ check: check-src check-tests check-version
 ## check-src:               runs linters (source only, no tests)
 .PHONY: check-src
 check-src:
-	ruff . --select I,UP015,UP032,UP034,UP018,COM,C4,PT,SIM,PLR0402 --ignore PT011,PT012,SIM117
+	ruff . --select I,UP015,UP032,UP034,UP018,COM,C4,PT,SIM,PLR0402 --ignore COM812,PT011,PT012,SIM117
 	black --line-length 100 ${PACKAGE_NAME} --check
 	flake8 ${PACKAGE_NAME}
 	mypy ${PACKAGE_NAME} --ignore-missing-imports --check-untyped-defs

scripts/collect_env.py

@@ -40,7 +40,7 @@ def get_os_version():
     return platform.platform()
 
 
-def is_python_package_installed(package_name):
+def is_python_package_installed(package_name: str):
     """
     Check if a Python package is installed
 
@@ -57,14 +57,10 @@ def is_python_package_installed(package_name):
         check=True,
     )
 
-    for line in result.stdout.splitlines():
-        if line.lower().startswith(package_name.lower()):
-            return True
-
-    return False
+    return any(line.lower().startswith(package_name.lower()) for line in result.stdout.splitlines())
 
 
-def is_brew_package_installed(package_name):
+def is_brew_package_installed(package_name: str):
     """
     Check if a Homebrew package is installed
 
@@ -95,11 +91,7 @@ def is_brew_package_installed(package_name):
         check=True,
     )
 
-    for line in result.stdout.splitlines():
-        if line.lower().startswith(package_name.lower()):
-            return True
-
-    return False
+    return any(line.lower().startswith(package_name.lower()) for line in result.stdout.splitlines())
 
 
 def get_python_package_version(package_name):
@@ -221,8 +213,7 @@ def main():
     ):
         print(
             "PaddleOCR version: ",
-            get_python_package_version("paddlepaddle")
-            or get_python_package_version("paddleocr"),
+            get_python_package_version("paddlepaddle") or get_python_package_version("paddleocr"),
         )
     else:
         print("PaddleOCR is not installed")

scripts/performance/run_partition.py

@@ -13,11 +13,7 @@
 
     file_path = sys.argv[1]
     strategy = sys.argv[2]
-    model_name = None
-    if len(sys.argv) > 3:
-        model_name = sys.argv[3]
-    else:
-        model_name = os.environ.get("PARTITION_MODEL_NAME")
+    model_name = sys.argv[3] if len(sys.argv) > 3 else os.environ.get("PARTITION_MODEL_NAME")
     result = partition(file_path, strategy=strategy, model_name=model_name)
     # access element in the return value to make sure we got something back, otherwise error
     result[1]

unstructured/documents/html.py

@@ -444,11 +444,8 @@ def _construct_text(tag_elem: etree.Element, include_tail_text: bool = True) ->
     return text.strip()
 
 
-def _has_break_tags(tag_elem: etree.Element) -> bool:
-    for descendant in tag_elem.iterdescendants():
-        if descendant.tag in TEXTBREAK_TAGS:
-            return True
-    return False
+def _has_break_tags(tag_elem: etree._Element) -> bool:  # pyright: ignore[reportPrivateUsage]
+    return any(descendant.tag in TEXTBREAK_TAGS for descendant in tag_elem.iterdescendants())
 
 
 def _unfurl_break_tags(tag_elem: etree.Element) -> List[etree.Element]:

unstructured/partition/json.py

@@ -54,10 +54,7 @@ def partition_json(
         last_modification_date = get_last_modified_date_from_file(file)
 
         file_content = file.read()
-        if isinstance(file_content, str):
-            file_text = file_content
-        else:
-            file_text = file_content.decode()
+        file_text = file_content if isinstance(file_content, str) else file_content.decode()
         file.seek(0)
 
     elif text is not None:

unstructured/staging/prodigy.py

@@ -28,9 +28,7 @@ def _validate_prodigy_metadata(
         )
         if isinstance(id_error_index, int):
             raise ValueError(
-                'The key "id" is not allowed with metadata parameter at index: {index}'.format(
-                    index=id_error_index,
-                ),
+                f'The key "id" is not allowed with metadata parameter at index: {id_error_index}'
             )
         validated_metadata = metadata
     else: