{

"intent": "issuelist",
"question": "issue list",
"score": 0.9,
"nodes": [
{
"id": "1",
"label": "Node 1",
"x": -4.625,
"y": 62.296875,
"config": {
"responseType": "api",
"text": "",
"value": {
"apiUrl": "http://127.0.0.1:8000/jira/projects",
"apiMethod": "GET",
"apiHeaders": "",
"apiBody": ""
}
}
},
{
"id": "2",
"label": "Node 2",
"x": 91.375,
"y": 201.296875,
"config": {
"responseType": "report",
"text": "",
"value": {
"title": "Projects",
"datafield": "{{issuelist#1}}",
"summary": "List of Projects",
"columns": [
{
"key": "value[0].name",
"label": "name"
}
]
}
}
}
],
"links": [
{
"from": "1",
"to": "2"
}
]
},
{
"intent": "projects",
"question": "show projects",
"score": 1.5,
"nodes": [
{
"id": "1",
"label": "Node 1",
"x": 89.75,
 "y": 98.203125,
"config": {
"responseType": "api",
"text": "",
"value": {
"apiUrl": "http://127.0.0.1:8000/jira/projects",
"apiMethod": "GET",
"apiHeaders": "",
"apiBody": ""
}
}
},
{
"id": "2",
"label": "Node 2",
"x": 94.75,
"y": 205.203125,
"config": {
"responseType": "form",
"text": "",
"value": [
{
"label": "Project",
"type": "dropdown",
"required": false,
"options": "",
"optionsType": "dynamic",
"datafield": "{{projects#1}}.value",
"datakey": "key",
"datalabel": "name"
}
]
}
},
{
"id": "3",
"label": "Node 3",
"x": 95.75,
"y": 299.203125,
"config": {
"responseType": "api",
"text": "",
"value": {
"apiUrl": "http://127.0.0.1:8000/jira/search_issues?
jql=project='{{projects#2}}.Project'",
"apiMethod": "GET",
"apiHeaders": "",
"apiBody": ""
}
}
},
{
"id": "4",
"label": "Node 4",
"x": 93.75,
"y": 408.203125,
"config": {
"responseType": "report",
"text": "",
 "value": {
"title": "Issues",
"datafield": "{{projects#3}}.value.issues",
"summary": "Issuelist",
"columns": [
{
"key": "key",
"label": "Task ID"
},
{
"key": "fields.summary",
"label": "Task Summary"
},
{
"key": "fields.description",
"label": "Description"
},
{
"key": "fields.assignee.displayName",
"label": "Assignee"
},
{
"key": "fields.status.name",
"label": "Status"
},
{
"key": "fields.priority.name",
"label": "Priority"
},
{
"key": "fields.duedate",
"label": "Due Date"
},
{
"key": "fields.created",
"label": "Created Date"
},
{
"key": "fields.resolutiondate",
"label": "Resolved Date"
},
{
"key": "fields.timespent",
"label": "Time Spent"
},
{
"key": "fields.issuetype.name",
"label": "Issue Type"
}
]
}
}
}
],
"links": [
{
"from": "1",
"to": "2"
},
 {
"from": "2",
"to": "3"
},
{
"from": "3",
"to": "4"
}
]
},

import nltk
from sentence_transformers import SentenceTransformer, util
import numpy as np

nltk.download('stopwords')
from nltk.corpus import stopwords

# Assuming you already have a Sentence Transformer model loaded

model = SentenceTransformer('all-mpnet-base-v2')

# Preprocessing function (as discussed in previous responses)

def preprocess_text(text):
return text
# Lowercase the text
text = text.lower()

# Remove special characters or digits (optional, depending on your needs)

text = re.sub(r'\W+', ' ', text)

# Remove stopwords (common words like "is", "the", "and", etc.)

stop_words = set(stopwords.words('english'))
text = ' '.join([word for word in text.split() if word not in stop_words])
print(text)
return text

# Intent matching function with vectorized similarity calculation

def getIntent( index,question):
db_data = db.load_db()

# Flatten all nodes from all workflows

#
intents =db_data['workflows']
processed_question = preprocess_text(question)
input_embedding = model.encode(processed_question, convert_to_tensor=True)

processed_intents = [preprocess_text(intent) for intent in intents]

intent_embeddings = model.encode(processed_intents, convert_to_tensor=True)

similarities = util.pytorch_cos_sim(input_embedding,
intent_embeddings).squeeze(0)
best_idx = torch.argmax(similarities).item()
best_intent = intents[best_idx]

print(f"Matched Intent: {best_intent}")

return best_intent
def getIntent1(index,question):
global model
# Input question

# Encode input and known questions

input_embedding = model.encode(question, convert_to_tensor=True)
best_intent = None
best_score = 0
db_data = db.load_db()

# Flatten all nodes from all workflows

#
intents =db_data['workflows'] #[node for workflow in db_data['workflows'] for
node in workflow["nodes"] if node["id"] == "1"]
for intent in intents:
#print(intent)
#print(intent["question"])

#print(intent["score"])
intent_embedding = model.encode(intent["question"], convert_to_tensor=True)
similarity = util.pytorch_cos_sim(input_embedding, intent_embedding).item()
#print(similarity)
if similarity >= intent["score"] and similarity > best_score:
best_score = similarity
best_intent = intent["intent"]
return intent

print("Matched Intent:", best_intent if best_intent else "No matching intent")