Refactor and enhance type hints across multiple modules

- Updated the `from_tuple` method in `Prof` class to specify return type.
- Added type hints for various methods in `LehmannsClient`, `OpenAI`, `WebRequest`, and `ZoteroController` classes to improve code clarity and type safety.
- Modified `pdf_to_csv` function to return a string instead of a DataFrame.
- Enhanced error handling and type hints in `wordparser` and `xmlparser` modules.
- Removed unused UI file `Ui_medianadder.ts`.
- Improved the layout and structure of the `semesterapparat_ui` to enhance user experience.
- Updated file picker to support `.doc` files in addition to `.docx`.
- Added unique item handling in `Ui` class to prevent duplicates in apparat list.
- General code cleanup and consistency improvements across various files.

src/logic/openai.py

@@ -1,10 +1,12 @@
-from openai import OpenAI
-from src import settings
 import json
+from typing import Any
+
+from openai import OpenAI
+
+from src import settings
 
 
-
-def init_client():
+def init_client() -> OpenAI:
     """Initialize the OpenAI client with the API key and model from settings."""
     global client, model, api_key
     if not settings.openAI.api_key:
@@ -16,9 +18,11 @@ def init_client():
     api_key = settings.openAI.api_key
     client = OpenAI(api_key=api_key)
     return client
-def run_shortener(title:str, length:int):
+
+
+def run_shortener(title: str, length: int) -> list[dict[str, Any]]:
     client = init_client()
-    response = client.responses.create(
+    response = client.responses.create(  # type: ignore
         model=model,
         instructions="""you are a sentence shortener. The next message will contain the string to shorten and the length limit.
 You need to shorten the string to be under the length limit, while keeping as much detail as possible. The result may NOT be longer than the length limit.
@@ -27,27 +31,28 @@ based on that, please reply only the shortened string. Give me 5 choices. if the
     )
     answers = response.output_text
     return eval(answers)  # type: ignore
-    #answers are strings in json format, so we need to convert them to a list of dicts
+    # answers are strings in json format, so we need to convert them to a list of dicts
 
 
-def name_tester(name: str):
+def name_tester(name: str) -> dict:
     client = init_client()
-    response = client.responses.create(
-        model = model,
+    response = client.responses.create(  # type: ignore
+        model=model,
         instructions="""you are a name tester, You are given a name and will have to split the name into first name, last name, and if present the title. Return the name in a json format with the keys "title", "first_name", "last_name". If no title is present, set title to none. Do NOt return the answer in a codeblock, use a pure json string. Assume the names are in the usual german naming scheme""",
-        input = f'{{"name":"{name}"}}'
+        input=f'{{"name":"{name}"}}',
     )
     answers = response.output_text
 
     return json.loads(answers)
 
-def semester_converter(semester:str):
+
+def semester_converter(semester: str) -> str:
     client = init_client()
-    response = client.responses.create(
-        model = model,
+    response = client.responses.create(  # type: ignore
+        model=model,
         instructions="""you are a semester converter. You will be given a string. Convert this into a string like this: SoSe YY or WiSe YY/YY+1. Do not return the answer in a codeblock, use a pure string.""",
-        input = semester
+        input=semester,
     )
     answers = response.output_text
 
-    return answers
+    return answers