From 63823a01de403cf1af6fece40ba270df2178f1fe Mon Sep 17 00:00:00 2001
From: Cristiano Ferreira <cristiano.ferreira@flowinn.biz>
Date: Wed, 3 Jan 2024 17:40:48 +0000
Subject: [PATCH] Add misspelled pairs to writing grading.

---
 app.py                          |   2 ++
 helper/speech_to_text_helper.py |  29 ++++++++++++++++++++++++++++-
 requirements.txt                | Bin 460 -> 574 bytes
 3 files changed, 30 insertions(+), 1 deletion(-)

diff --git a/app.py b/app.py
index fac3057..643e465 100644
--- a/app.py
+++ b/app.py
@@ -204,6 +204,7 @@ def grade_writing_task_1():
             response = make_openai_instruct_call(GPT_3_5_TURBO_INSTRUCT, message, token_count,
                                                  ["comment"],
                                                  GEN_QUESTION_TEMPERATURE)
+            response['misspelled_pairs'] = get_misspelled_pairs(answer)
             return response
         else:
             return {
@@ -256,6 +257,7 @@ def grade_writing_task_2():
             response = make_openai_instruct_call(GPT_3_5_TURBO_INSTRUCT, message, token_count,
                                                  ["comment"],
                                                  GEN_QUESTION_TEMPERATURE)
+            response['misspelled_pairs'] = get_misspelled_pairs(answer)
             return response
         else:
             return {
diff --git a/helper/speech_to_text_helper.py b/helper/speech_to_text_helper.py
index 6db79bb..b94c84b 100644
--- a/helper/speech_to_text_helper.py
+++ b/helper/speech_to_text_helper.py
@@ -1,8 +1,11 @@
+import string
 import whisper
 import os
 import nltk
 import boto3
 import random
+from spellchecker import SpellChecker
+
 nltk.download('words')
 from nltk.corpus import words
 from helper.constants import *
@@ -100,4 +103,28 @@ def divide_text(text, max_length=3000):
             divisions.append(text[current_position:next_position])
             current_position = next_position
 
-    return divisions
\ No newline at end of file
+    return divisions
+
+
+def get_misspelled_pairs(text):
+    spell = SpellChecker()
+
+    # Remove punctuation from the text
+    translator = str.maketrans("", "", string.punctuation)
+    text_without_punctuation = text.translate(translator)
+
+    # Split the text into words
+    words = text_without_punctuation.split()
+
+    # Find misspelled words
+    misspelled = spell.unknown(words)
+
+    # Create a list to store misspelled word pairs
+    misspelled_pairs = []
+
+    # Generate misspelled word pairs with their corrections
+    for word in misspelled:
+        correction = spell.correction(word)
+        misspelled_pairs.append({"misspelled": word, "correction": correction})
+
+    return misspelled_pairs
\ No newline at end of file
diff --git a/requirements.txt b/requirements.txt
index c00a2002f404897aa1824b03ed26ff03bed46150..d61d479d8e8cb926176d7313a0468430fdd0a1cb 100644
GIT binary patch
delta 122
zcmX@ZypLtW8O9=B1}=sIhDwHFAeqXL1BA&889-JtLpG2sVgSh(0a=9%r9iTnp#+FQ
cVkJO51q_)Cc?{`bemPKACe##&E?g!804HM?-~a#s

delta 7
OcmdnTa)x=s8AbpN0s{{K