From fdacdf2c8aac1c174db04d9f132bf79cfc397148 Mon Sep 17 00:00:00 2001
From: Joshua Wendland <80349780+joshuawe@users.noreply.github.com>
Date: Fri, 9 Feb 2024 17:54:41 +0100
Subject: [PATCH] fix errors

---
 requirements.txt                   |  2 +-
 src/delphi/eval/token_labelling.py | 28 +++++++---------------------
 2 files changed, 8 insertions(+), 22 deletions(-)

diff --git a/requirements.txt b/requirements.txt
index 07a340c6..7bda2f6c 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,4 +1,4 @@
-# torch==2.1.2
+torch==2.1.2
 datasets==2.16.1
 transformers==4.36.2
 tqdm==4.66.1
diff --git a/src/delphi/eval/token_labelling.py b/src/delphi/eval/token_labelling.py
index 5ec134e1..163d9bb1 100644
--- a/src/delphi/eval/token_labelling.py
+++ b/src/delphi/eval/token_labelling.py
@@ -4,7 +4,7 @@
 Additionally, it can visualize the sentences and their poart-of-speech (POS) tags.
 """
 
-from typing import Callable, Optional, Union
+from typing import Callable, Optional
 
 import spacy  # pylint: disable=import-error
 from spacy.tokens import Doc  # pylint: disable=import-error
@@ -98,7 +98,7 @@ def label_single_Token(token: Token) -> dict[str, bool]:
     return labels
 
 
-def label_sentence(tokens: Union[Doc, list[Token]]) -> list[dict[str, bool]]:
+def label_sentence(tokens: Doc | list[Token]) -> list[dict[str, bool]]:
     """
     Labels spaCy Tokens in a sentence. Takes the context of the token into account
     for dependency labels (e.g. subject, object, ...), IF dependency labels are turned on.
@@ -121,7 +121,7 @@ def label_sentence(tokens: Union[Doc, list[Token]]) -> list[dict[str, bool]]:
 
 
 def label_batch_sentences(
-    sentences: Union[list[str], list[list[str]]],
+    sentences: list[str] | list[list[str]],
     tokenized: bool = True,
     verbose: bool = False,
 ) -> list[list]:
@@ -146,7 +146,6 @@ def label_batch_sentences(
         corresponding token length where each entry provides the labels/categories
         for the token. Sentence -> Token -> Labels
     """
-    assert isinstance(sentences, list)
     # Load english language model
     nlp = spacy.load("en_core_web_sm")
     # labelled tokens, list holding sentences holding tokens holding corresponding token labels
@@ -168,18 +167,10 @@ def label_batch_sentences(
         labelled_tokens = list()  # list holding labels for all tokens of sentence
         labelled_tokens = label_sentence(doc)
 
-        # go through each token in the sentence
-        for token, labelled_token in zip(doc, labelled_tokens):
-            # labelled_token = label_single_Token(token)
-            # labels = list()  #  The list holding labels of a single token
-            # for _, category_check in TOKEN_LABELS.items():
-            #     label = category_check(token)
-            #     labels.append(label)
-            # add current token's to the list
-            # labelled_tokens.append(labelled_token)
-
-            # print the token and its labels to console
-            if verbose is True:
+        # print the token and its labels to console
+        if verbose is True:
+            # go through each token in the sentence
+            for token, labelled_token in zip(doc, labelled_tokens):
                 print(f"Token: {token}")
                 print(" | ".join(list(TOKEN_LABELS.keys())))
                 printable = [
@@ -198,11 +189,6 @@ def label_batch_sentences(
 
 
 if __name__ == "__main__":
-    # result = label_tokens(
-    #     ["Hi, my name is Joshua.".split(" "), "The highway is full of car s, Peter.".split(" ")],
-    #     tokenized=True,
-    #     verbose=True,
-    # )
     result = label_batch_token(
         ["Hi, my name is Joshua.", "The highway is full of car s, Peter."],
         tokenized=False,