From 7c68d0ce7b547bd45f52ee028dd80ea2901cba26 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Maciej=20=C5=9Acigacz?= <macsci1999@gmail.com>
Date: Wed, 14 Jun 2023 00:03:39 +0200
Subject: [PATCH] add tone and style services

---
 application/__init__.py                |  4 +++
 application/functions/style.py         | 33 +++++++++++++++++++++
 application/functions/tone.py          | 40 ++++++++++++++++++++++++++
 application/services/style_services.py | 16 +++++++++++
 application/services/tone_services.py  | 16 +++++++++++
 5 files changed, 109 insertions(+)
 create mode 100644 application/functions/style.py
 create mode 100644 application/functions/tone.py
 create mode 100644 application/services/style_services.py
 create mode 100644 application/services/tone_services.py

diff --git a/application/__init__.py b/application/__init__.py
index 70d39d6..8098b5b 100644
--- a/application/__init__.py
+++ b/application/__init__.py
@@ -8,9 +8,13 @@ def create_app():
     from application.services.sentiment_service import sentiment_service
     from application.services.errors_service import errors_service
     from application.services.irony_service import irony_service
+    from application.services.style_services import style_service
+    from application.services.tone_services import tone_service
     
     application.register_blueprint(sentiment_service)
     application.register_blueprint(errors_service)
     application.register_blueprint(irony_service)
+    application.register_blueprint(style_service)
+    application.register_blueprint(tone_service)
 
     return application
\ No newline at end of file
diff --git a/application/functions/style.py b/application/functions/style.py
new file mode 100644
index 0000000..73ac6a5
--- /dev/null
+++ b/application/functions/style.py
@@ -0,0 +1,33 @@
+from transformers import pipeline
+import re
+
+pipe = pipeline('text-classification', model="jagiyahh/simple-polish-stylistic-errors", tokenizer = 'dkleczek/bert-base-polish-uncased-v1')
+
+def style_prediction(data):
+    result = pipe(data)
+    
+    return result
+
+def clear_data(data):
+    data = [re.sub(r"[^A-Za-zżźćńółęąśŻŹĆĄŚĘŁÓŃ ']+", r"", i) for i in data['sentences']]
+    data = [x for x in data if x != '']
+    data = [i.strip() for i in data]
+    data = [i.lower() for i in data]
+
+    return data
+
+def count_predictions(predictions):
+    l0 = 0
+    l1 = 0
+    all = {}
+
+    for i in predictions:
+        if i['label'] == 'LABEL_0':
+            l0 += 1
+        if i['label'] == 'LABEL_1':
+            l1 += 1
+
+    all['stylistically_positive'] = l0
+    all['stylistically_negative'] = l1
+
+    return all
diff --git a/application/functions/tone.py b/application/functions/tone.py
new file mode 100644
index 0000000..3de5fe4
--- /dev/null
+++ b/application/functions/tone.py
@@ -0,0 +1,40 @@
+from transformers import BertTokenizer, BertForSequenceClassification
+import torch
+
+model = BertForSequenceClassification.from_pretrained('jagiyahh/simple-polish-tone-recognition')
+tokenizer = BertTokenizer.from_pretrained('jagiyahh/simple-polish-tone-recognition')
+
+labels = ['controversial', 'intriguing', 'formal']
+
+def clear_data(data):
+    data = [i.strip() for i in data]
+    data = [i.lower() for i in data]
+
+    return data
+
+
+def predict_labels(texts):
+    encodings = tokenizer(texts, truncation=True, padding=True, return_tensors='pt')
+    input_ids = encodings['input_ids']
+    attention_mask = encodings['attention_mask']
+    
+    with torch.no_grad():
+        outputs = model(input_ids, attention_mask=attention_mask)
+    logits = outputs.logits
+    
+    probabilities = torch.sigmoid(logits)
+    
+    threshold = 0.5
+    predictions = (probabilities > threshold).int()
+    
+    predicted_labels = []
+    for pred in predictions:
+        label_indices = torch.nonzero(pred).flatten().tolist()
+        predicted_labels.append([labels[i] for i in label_indices])
+    
+    return predicted_labels
+
+def tone_prediction(data):
+    prediction = predict_labels(data)
+    
+    return prediction
diff --git a/application/services/style_services.py b/application/services/style_services.py
new file mode 100644
index 0000000..458a02f
--- /dev/null
+++ b/application/services/style_services.py
@@ -0,0 +1,16 @@
+from flask import(
+    request,   
+    jsonify, 
+    Blueprint,
+    )
+from application.functions.style import style_prediction, clear_data
+
+style_service = Blueprint("style_service", __name__)    
+
+@style_service.route("/get_style_data", methods=['POST'])
+def get_data():
+    data = request.get_json()
+    data_clear = clear_data(data)
+    predicitons = style_prediction(data_clear)
+
+    return jsonify({"predictions": predicitons})
\ No newline at end of file
diff --git a/application/services/tone_services.py b/application/services/tone_services.py
new file mode 100644
index 0000000..6dcc3af
--- /dev/null
+++ b/application/services/tone_services.py
@@ -0,0 +1,16 @@
+from flask import(
+    request,   
+    jsonify, 
+    Blueprint,
+    )
+from application.functions.tone import tone_prediction, clear_data
+
+tone_service = Blueprint("tone_service", __name__)    
+
+@tone_service.route("/get_tone_data", methods=['POST'])
+def get_data():
+    data = request.get_json()
+    data_clear = clear_data(data['sentences'])
+    predicitons = tone_prediction(data_clear)
+
+    return jsonify({"predictions": predicitons})
\ No newline at end of file