File size: 37,004 Bytes
c06fe3c
 
 
 
 
 
 
 
 
 
 
5a27270
0e60fcf
 
5a27270
254234b
 
1478faa
 
 
 
c06fe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a9db436
c06fe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a9db436
 
c06fe3c
 
a9db436
 
c06fe3c
 
 
 
 
 
 
a9db436
 
 
 
 
c06fe3c
 
 
a9db436
 
 
 
 
 
 
 
 
 
c06fe3c
a9db436
 
c06fe3c
 
a9db436
 
 
 
c06fe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c72d516
c06fe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a9db436
c06fe3c
a9db436
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c06fe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a9db436
 
c06fe3c
 
 
a9db436
c06fe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a9db436
 
c06fe3c
 
 
a9db436
c06fe3c
 
a9db436
 
 
 
 
 
 
c06fe3c
 
 
 
a9db436
c06fe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a9db436
c06fe3c
 
 
 
 
 
 
 
 
 
a9db436
c06fe3c
 
 
 
 
 
 
a9db436
 
 
 
 
 
c06fe3c
 
a9db436
c06fe3c
a9db436
 
 
 
c06fe3c
 
 
 
 
a9db436
c06fe3c
 
 
 
 
 
c72d516
c06fe3c
 
 
a9db436
 
 
 
 
839ddf6
a9db436
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c72d516
 
a9db436
c06fe3c
a9db436
 
 
 
 
839ddf6
a9db436
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c72d516
 
a9db436
 
 
 
c06fe3c
 
 
a9db436
 
 
 
 
 
 
 
 
 
c72d516
 
a9db436
 
 
c72d516
a9db436
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c72d516
 
a9db436
 
c72d516
a9db436
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c06fe3c
 
660ae67
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
import gradio as gr
import sqlite3
import bcrypt
from datetime import datetime
import re
import torch
import torch.nn as nn
import torch.nn.functional as F
from transformers import AutoTokenizer, AutoModel, AutoModelForSequenceClassification
import os
import logging
from openai import OpenAI
import json
from fpdf import FPDF

print("ENV:", os.environ)  # 👈 Add this for debugging

api_key = os.getenv("OPENAI_API_KEY")
if not api_key:
    raise RuntimeError("OPENAI_API_KEY environment variable not found.")
client = OpenAI(api_key=api_key)
import json
from fpdf import FPDF

# --------------------------
# Environment Setup
# --------------------------
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print("Using device:", device)

# --------------------------
# Global Tokenizer and Hybrid Model for Treatment Prediction
# --------------------------
tokenizer = AutoTokenizer.from_pretrained("emilyalsentzer/Bio_ClinicalBERT")


class HybridMentalHealthModel(nn.Module):
    def __init__(self, bert_model, num_genders, num_medications, num_therapies, hidden_size=128):
        super(HybridMentalHealthModel, self).__init__()
        self.bert = AutoModel.from_pretrained(bert_model)
        bert_output_size = self.bert.config.hidden_size
        self.age_fc = nn.Linear(1, 16)
        self.gender_fc = nn.Embedding(num_genders, 16)
        self.fc = nn.Linear(bert_output_size + 32, hidden_size)
        self.medication_head = nn.Linear(hidden_size, num_medications)
        self.therapy_head = nn.Linear(hidden_size, num_therapies)

    def forward(self, input_ids, attention_mask, age, gender):
        bert_output = self.bert(input_ids=input_ids, attention_mask=attention_mask).last_hidden_state[:, 0, :]
        age_out = self.age_fc(age)
        gender_out = self.gender_fc(gender)
        combined = torch.cat((bert_output, age_out, gender_out), dim=1)
        hidden = torch.relu(self.fc(combined))
        return self.medication_head(hidden), self.therapy_head(hidden)


# --------------------------
# Global Label Mappings and Age Scaler
# --------------------------
medication_classes = ["Anxiolytics", "Benzodiazepines", "Antidepressants", "Mood Stabilizers", "Antipsychotics", "Stimulants"]
therapy_classes = ["Cognitive Behavioral Therapy", "Dialectical Behavioral Therapy", "Interpersonal Therapy", "Mindfulness-Based Therapy"]  # Update with your types
gender_classes = ["Male", "Female", "Other"]

medication_encoder = {name: idx for idx, name in enumerate(medication_classes)}
inv_medication_encoder = {idx: name for name, idx in medication_encoder.items()}
therapy_encoder = {name: idx for idx, name in enumerate(therapy_classes)}
inv_therapy_encoder = {idx: name for name, idx in therapy_encoder.items()}
gender_encoder = {name: idx for idx, name in enumerate(gender_classes)}

mean_age = 50
std_age = 10

def scale_age(age):
    return (age - mean_age) / std_age

# --------------------------
# Load the Hybrid Model (Treatment Prediction)
# --------------------------
num_genders = len(gender_classes)
num_medications = len(medication_classes)
num_therapies = len(therapy_classes)
MODEL_SAVE_PATH = "22.03.2025-16.02-ML128E10"  # Update accordingly

model = HybridMentalHealthModel("emilyalsentzer/Bio_ClinicalBERT", num_genders, num_medications, num_therapies)
state_dict = torch.load(MODEL_SAVE_PATH, map_location=device)
if "gender_fc.weight" in state_dict:
    del state_dict["gender_fc.weight"]
model.load_state_dict(state_dict, strict=False)
model.to(device)
model.eval()

# --------------------------
# Global Diagnosis Model (Mental Health Diagnosis)
# --------------------------
diagnosis_tokenizer = AutoTokenizer.from_pretrained("ethandavey/mental-health-diagnosis-bert")  # Update with your model ID
diagnosis_model = AutoModelForSequenceClassification.from_pretrained("ethandavey/mental-health-diagnosis-bert")  # Update with your model ID
diagnosis_model.to(device)
diagnosis_model.eval()

def predict_disease(text):
    inputs = diagnosis_tokenizer(text, return_tensors="pt", padding=True, truncation=True, max_length=128)
    inputs = {k: v.to(device) for k, v in inputs.items()}
    with torch.no_grad():
        outputs = diagnosis_model(**inputs)
        probabilities = F.softmax(outputs.logits, dim=1).squeeze()
    label_mapping = {0: "Anxiety", 1: "Normal", 2: "Depression", 3: "Suicidal", 4: "Stress"}

    topk = torch.topk(probabilities, k=3)
    top_preds = [(label_mapping[i.item()], probabilities[i].item()) for i in topk.indices]
    return top_preds


def predict_med_therapy(symptoms, age, gender):
    encoding = tokenizer(symptoms, return_tensors="pt", truncation=True, padding='max_length', max_length=128)
    input_ids = encoding["input_ids"].to(device)
    attention_mask = encoding["attention_mask"].to(device)
    age_norm = torch.tensor([[scale_age(age)]], dtype=torch.float32).to(device)
    gender_idx = gender_encoder.get(gender, 0)
    gender_tensor = torch.tensor([gender_idx], dtype=torch.long).to(device)
    with torch.no_grad():
        med_logits, therapy_logits = model(input_ids, attention_mask, age_norm, gender_tensor)
    med_probabilities = torch.softmax(med_logits, dim=1)
    therapy_probabilities = torch.softmax(therapy_logits, dim=1)
    med_pred = torch.argmax(med_probabilities, dim=1).item()
    therapy_pred = torch.argmax(therapy_probabilities, dim=1).item()
    med_confidence = med_probabilities[0][med_pred].item()
    therapy_confidence = therapy_probabilities[0][therapy_pred].item()
    predicted_med = inv_medication_encoder.get(med_pred, "Unknown")
    predicted_therapy = inv_therapy_encoder.get(therapy_pred, "Unknown")
    return (predicted_med, med_confidence), (predicted_therapy, therapy_confidence)

# --------------------------
# OpenAI Functions (Summarization and Explanation)
# --------------------------
def get_concise_rewrite(text, max_tokens, temperature=0.7):
    messages = [
        {"role": "system", "content": "You are an expert rewriting assistant. Rewrite the given statement into a concise version while preserving its tone and vocabulary."},
        {"role": "user", "content": text}
    ]
    try:
        response = client.chat.completions.create(model="gpt-4o-mini", messages=messages, max_tokens=max_tokens, temperature=temperature)
        concise_text = response.choices[0].message.content.strip()
    except Exception as e:
        concise_text = f"API call failed: {e}"
    return concise_text

def get_explanation(patient_statement, predicted_diagnosis):
    messages = [
        {"role": "system", "content": "You are an expert mental health assistant. Provide a concise, evidence-based explanation of how the patient's statement supports the diagnosis."},
        {"role": "user", "content": f"Patient statement: {patient_statement}\nPredicted diagnosis: {predicted_diagnosis}\nExplain briefly."}
    ]
    try:
        response = client.chat.completions.create(model="gpt-4o-mini", messages=messages, max_tokens=256)
        explanation = response.choices[0].message.content.strip()
    except Exception as e:
        explanation = "API call failed."
    return explanation

# --------------------------
# Database Functions
# --------------------------
def init_db():
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    c.execute("""
        CREATE TABLE IF NOT EXISTS users (
            id INTEGER PRIMARY KEY AUTOINCREMENT,
            username TEXT UNIQUE NOT NULL,
            password TEXT NOT NULL,
            full_name TEXT,
            email TEXT
        )
    """)
    c.execute("""
        CREATE TABLE IF NOT EXISTS chat_history (
            id INTEGER PRIMARY KEY AUTOINCREMENT,
            username TEXT NOT NULL,
            message TEXT NOT NULL,
            response TEXT NOT NULL,
            timestamp DATETIME DEFAULT CURRENT_TIMESTAMP
        )
    """)
    c.execute("""
        CREATE TABLE IF NOT EXISTS patient_sessions (
            id INTEGER PRIMARY KEY AUTOINCREMENT,
            username TEXT,
            patient_name TEXT,
            age REAL,
            gender TEXT,
            symptoms TEXT,
            diagnosis TEXT,
            medication TEXT,
            therapy TEXT,
            summary TEXT,
            explanation TEXT,
            pdf_report TEXT,
            session_timestamp DATETIME DEFAULT CURRENT_TIMESTAMP,
            appointment_date DATE
        )
    """)
    conn.commit()
    conn.close()

def register_user(username, password, full_name, email):
    if not re.fullmatch(r"[^@]+@[^@]+\.[^@]+", email):
        return "Invalid email format."
    if len(password) <= 8:
        return "Password must be more than 8 characters."
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    hashed = bcrypt.hashpw(password.encode(), bcrypt.gensalt())
    try:
        c.execute("INSERT INTO users (username, password, full_name, email) VALUES (?, ?, ?, ?)", (username, hashed, full_name, email))
        conn.commit()
        return "User registered successfully."
    except sqlite3.IntegrityError:
        return "Username already exists."
    finally:
        conn.close()

def login_user(username, password):
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    c.execute("SELECT password FROM users WHERE username = ?", (username,))
    user = c.fetchone()
    conn.close()
    if user and bcrypt.checkpw(password.encode(), user[0]):
        return True
    return False

def get_user_info(username):
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    c.execute("SELECT username, email, full_name FROM users WHERE username = ?", (username,))
    user = c.fetchone()
    conn.close()
    if user:
        return f"Username: {user[0]}\nFull Name: {user[2]}\nEmail: {user[1]}"
    else:
        return "User info not found."


def get_chat_history(username):
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    c.execute("SELECT message, response, timestamp FROM chat_history WHERE username = ? ORDER BY timestamp DESC", (username,))
    history = c.fetchall()
    conn.close()
    return history

def get_patient_sessions(filter_name="", filter_date=""):
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    query = "SELECT patient_name, age, gender, symptoms, diagnosis, medication, therapy, summary, explanation, pdf_report, session_timestamp FROM patient_sessions WHERE 1=1"
    params = []
    if filter_name:
        query += " AND patient_name LIKE ?"
        params.append(f"%{filter_name}%")
    if filter_date:
        query += " AND DATE(session_timestamp)=?"
        params.append(filter_date)
    c.execute(query, params)
    sessions = c.fetchall()
    conn.close()
    return sessions

def insert_patient_session(session_data):
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    c.execute("""
        INSERT INTO patient_sessions (username, patient_name, age, gender, symptoms, diagnosis, medication, therapy, summary, explanation, pdf_report, appointment_date)
        VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
    """, (
        session_data.get("username"), session_data.get("patient_name"), session_data.get("age"), session_data.get("gender"),
        session_data.get("symptoms"), session_data.get("diagnosis"), session_data.get("medication"),
        session_data.get("therapy"), session_data.get("summary"), session_data.get("explanation"),
        session_data.get("pdf_report"), session_data.get("appointment_date")))
    conn.commit()
    conn.close()

# --------------------------
# PDF Report Generation Function
# --------------------------
def generate_pdf_report(session_data):
    pdf = FPDF()
    pdf.add_page()
    pdf.set_font("Arial", size=12)
    # Use safe_text to ensure the title is safe for latin-1 encoding
    pdf.cell(200, 10, txt=safe_text("Patient Session Report"), ln=True, align='C')
    pdf.ln(10)
    for key, value in session_data.items():
        # Convert each line to a safe text version before writing it
        pdf.multi_cell(0, 10, txt=safe_text(f"{key.capitalize()}: {value}"))
    reports_dir = "reports"
    os.makedirs(reports_dir, exist_ok=True)
    timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
    filename = f"{reports_dir}/{session_data.get('patient_name')}_{timestamp}.pdf"
    pdf.output(filename)
    return filename


def safe_text(txt):
    # Encode the text to latin-1, replacing characters that can't be encoded
    return txt.encode("latin-1", "replace").decode("latin-1")

# --------------------------
# Helper: Autofill Previous Patient Info
# --------------------------
def get_previous_patients():
    # Use the current logged-in user from user_session.value
    username = user_session.value
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    c.execute("SELECT DISTINCT patient_name FROM patient_sessions WHERE username=?", (username,))
    records = c.fetchall()
    conn.close()
    return [r[0] for r in records]

def get_previous_patient_info(selected_patient):
    # Use the current logged-in user from user_session.value
    username = user_session.value
    conn = sqlite3.connect("users.db")
    c = conn.cursor()
    c.execute(
        "SELECT patient_name, age, gender FROM patient_sessions WHERE username=? AND patient_name=? ORDER BY session_timestamp DESC LIMIT 1",
        (username, selected_patient)
    )
    record = c.fetchone()
    conn.close()
    if record:
        return record[0], record[1], record[2]
    else:
        return "", None, ""

# --------------------------
# Gradio UI Setup with External CSS
# --------------------------
with gr.Blocks(css=open("styles.css", "r").read(), theme="soft") as app:
    user_session = gr.State(value="")
    profile_visible = gr.State(value=False)
    session_data_state = gr.State(value="")

    with gr.Row(elem_id="header") as header_row:
        with gr.Column(scale=8):
            gr.Markdown("## Mental Health Chatbot")
        with gr.Column(visible=False, elem_id="profile_container") as profile_container:
            profile_button = gr.Button("👤", elem_id="profile_button", variant="secondary")
            with gr.Column(visible=False, elem_id="profile_info_box") as profile_info_box:
                profile_info = gr.HTML()
                logout_button = gr.Button("Logout", elem_id="logout_button")

    with gr.Column(visible=True, elem_id="login_page") as login_page:
        gr.Markdown("## Login")
        with gr.Row():
            username_login = gr.Textbox(label="Username")
            password_login = gr.Textbox(label="Password", type="password")
        login_btn = gr.Button("Login")
        login_output = gr.Textbox(label="Login Status", interactive=False)
        gr.Markdown("New user? Click below to register.")
        go_to_register = gr.Button("Go to Register")

    with gr.Column(visible=False, elem_id="register_page") as register_page:
        gr.Markdown("## Register")
        new_username = gr.Textbox(label="New Username")
        new_password = gr.Textbox(label="New Password", type="password")
        full_name = gr.Textbox(label="Full Name")
        email = gr.Textbox(label="Email")
        register_btn = gr.Button("Register")
        register_output = gr.Textbox(label="Registration Status", interactive=False)
        gr.Markdown("Already have an account?")
        back_to_login = gr.Button("Back to Login")

    with gr.Tabs(visible=False, elem_id="main_panel") as main_panel:
        with gr.Tab("Chatbot"):
            with gr.Row():
                with gr.Column(scale=1):
                    previous_patient = gr.Dropdown(label="Previous Patients", choices=[], interactive=True)
                    patient_name_input = gr.Textbox(placeholder="Enter patient name", label="Patient Name")
                    gender_input = gr.Dropdown(choices=list(gender_encoder.keys()), label="Gender")
                    age_input = gr.Number(label="Age")
                    symptoms_input = gr.Textbox(placeholder="Describe symptoms", label="Symptoms", lines=4)
                    submit = gr.Button("Submit")
                    generate_report_btn = gr.Button("Generate Report", visible=False)
                with gr.Column(scale=1):
                    with gr.Row():
                        with gr.Column(scale=4, min_width=240):  # Textbox column
                            diagnosis_textbox = gr.Textbox(label="Diagnosis", 
                                                        interactive=False)
                        with gr.Column(scale=1, min_width=120):  # Confidence column
                            diagnosis_conf_html = gr.HTML(elem_classes=["confidence-container"])

                    with gr.Row():
                        with gr.Column(scale=4, min_width=240):
                            medication_textbox = gr.Textbox(label="Medication", 
                                                        interactive=False)
                        with gr.Column(scale=1, min_width=120):
                            medication_conf_html = gr.HTML(elem_classes=["confidence-container"])

                    with gr.Row():
                        with gr.Column(scale=4, min_width=240):
                            therapy_textbox = gr.Textbox(label="Therapy", 
                                                    interactive=False)
                        with gr.Column(scale=1, min_width=120):
                            therapy_conf_html = gr.HTML(elem_classes=["confidence-container"])
                    summary_textbox = gr.Textbox(label="Concise Summary", interactive=False)
                    explanation_textbox = gr.Textbox(label="Explanation", interactive=False)
            with gr.Row():
                report_download = gr.File(label="Download Report", interactive=False)

            def handle_chat_extended(patient_name, gender, age, symptoms):
                if age is None or age <= 0:
                    error_msg = "Age must be greater than 0."
                    return (error_msg, "", error_msg, "", error_msg, "", error_msg, error_msg, gr.update(visible=False))
                
                if age > 150:
                    error_msg2 = "Age must be lower than 150"
                    return (error_msg2, "", error_msg2, "", error_msg2, "", error_msg2, error_msg2, gr.update(visible=False))
                
                if len(symptoms.split()) > 512:
                    msg = "Input exceeds maximum allowed length of 512 words."
                    return (msg, "", msg, "", msg, "", msg, msg, gr.update(visible=False))

                full_statement = f"Patient Name: {patient_name}, Gender: {gender}, Age: {age}, Symptoms: {symptoms}"
                summary = get_concise_rewrite(full_statement, max_tokens=150, temperature=0.7)

                # Predict top 3 diagnoses
                diagnosis_preds = predict_disease(full_statement)  # Now returns list of (label, confidence)
                diagnosis_display = "\n".join([f"{label}" for label, _ in diagnosis_preds])

                def get_confidence_class(percentage):
                    if percentage <= 50:
                        return "confidence-low"
                    elif percentage <= 74:
                        return "confidence-medium"
                    else:
                        return "confidence-high"

                diagnosis_conf_html_val = "<div class='confidence-multi'>" + "<br>".join([
                    f"<div class='confidence-display'><span class='confidence-value {get_confidence_class(conf * 100)}'>{conf * 100:.1f}% confidence</span></div>"
                    for _, conf in diagnosis_preds
                ]) + "</div>"

                # Predict medication and therapy
                (med_pred, med_conf), (therapy_pred, therapy_conf) = predict_med_therapy(symptoms, age, gender)
                med_percentage = med_conf * 100
                therapy_percentage = therapy_conf * 100

                def get_conf_html(percentage):
                    return f"""
                    <div class="confidence-display">
                        <span class="confidence-value {get_confidence_class(percentage)}">
                            {percentage:.1f}% confidence
                        </span>
                    </div>
                    """

                medication_conf_html_val = get_conf_html(med_percentage)
                therapy_conf_html_val = get_conf_html(therapy_percentage)

                # Explanation
                top_diag_labels = ", ".join([label for label, _ in diagnosis_preds])
                explanation = get_explanation(full_statement, f"{top_diag_labels}, {med_pred} and {therapy_pred}")

                # Prepare session data
                top_diag_with_conf = ", ".join([f"{label} ({conf * 100:.1f}%)" for label, conf in diagnosis_preds])
                session_data = {
                    "patient_name": patient_name,
                    "age": age,
                    "gender": gender,
                    "symptoms": symptoms,
                    "diagnosis": top_diag_with_conf,
                    "medication": f"{med_pred} ({med_percentage:.1f}% confidence)",
                    "therapy": f"{therapy_pred} ({therapy_percentage:.1f}% confidence)",
                    "summary": summary,
                    "explanation": explanation,
                    "session_timestamp": datetime.now().strftime("%Y-%m-%d %H:%M:%S")
                }
                session_data_state.value = json.dumps(session_data)

                # Save to chat history
                conn = sqlite3.connect("users.db")
                c = conn.cursor()
                if user_session.value:
                    c.execute("INSERT INTO chat_history (username, message, response) VALUES (?, ?, ?)",
                            (user_session.value, full_statement, top_diag_with_conf))
                    conn.commit()
                conn.close()

                return (
                    diagnosis_display, diagnosis_conf_html_val,
                    med_pred, medication_conf_html_val,
                    therapy_pred, therapy_conf_html_val,
                    summary, explanation,
                    gr.update(visible=True)
                )


            submit.click(handle_chat_extended,
                         inputs=[patient_name_input, gender_input, age_input, symptoms_input],
                         outputs=[diagnosis_textbox, diagnosis_conf_html, medication_textbox, medication_conf_html,
                                  therapy_textbox, therapy_conf_html, summary_textbox, explanation_textbox,
                                  generate_report_btn])

            def handle_generate_report():
                try:
                    # Try to load session data and generate the PDF report.
                    data = json.loads(session_data_state.value)
                    pdf_file = generate_pdf_report(data)
                    data["username"] = user_session.value
                    data["appointment_date"] = ""
                    data["pdf_report"] = pdf_file
                    insert_patient_session(data)
                    return pdf_file
                except Exception as e:
                    # Create an error file that contains the error message.
                    error_msg = f"Error generating PDF report: {str(e)}"
                    reports_dir = "reports"
                    os.makedirs(reports_dir, exist_ok=True)
                    error_filename = f"{reports_dir}/error_report_{datetime.now().strftime('%Y%m%d_%H%M%S')}.txt"
                    with open(error_filename, "w", encoding="utf-8") as f:
                        f.write(error_msg)
                    return error_filename


            generate_report_btn.click(handle_generate_report, outputs=report_download)

            def autofill_previous(selected_patient):
                name, age_val, gender_val = get_previous_patient_info(selected_patient)
                return name, age_val, gender_val

            previous_patient.change(autofill_previous,
                                    inputs=[previous_patient],
                                    outputs=[patient_name_input, age_input, gender_input])

        with gr.Tab("Chat History"):
            history_output = gr.Textbox(label="Chat History", interactive=False)
            load_history_btn = gr.Button("Load History")

            def load_history():
                if user_session.value:
                    history = get_chat_history(user_session.value)
                    chat_history_text = "\n".join([f"[{h[2]}] {h[0]}\nBot: {h[1]}" for h in history])
                    return f"Username: {user_session.value}\n\n{chat_history_text}"
                else:
                    return "Please log in to view history."


            load_history_btn.click(load_history, outputs=history_output)

        with gr.Tab("Book an Appointment"):
            with gr.Row():
                with gr.Column():
                    patient_name_appt = gr.Textbox(label="Patient Name", placeholder="Enter your name")
                    doctor_name = gr.Dropdown(choices=["Dr. Smith", "Dr. Johnson", "Dr. Lee"], label="Select Doctor")
                    appointment_date = gr.Textbox(label="Appointment Date", placeholder="YYYY-MM-DD")
                    appointment_time = gr.Textbox(label="Appointment Time", placeholder="HH:MM (24-hour format)")
                    reason = gr.TextArea(label="Reason for Visit", placeholder="Describe your symptoms or reason for the visit")
                    book_button = gr.Button("Book Appointment")
                with gr.Column():
                    booking_output = gr.Textbox(label="Booking Confirmation", interactive=False)

            def book_appointment(patient_name, doctor_name, appointment_date, appointment_time, reason):
                if not user_session.value:
                    return "Please log in to book an appointment."
                patient_name = (patient_name or "").strip()
                doctor_name = (doctor_name or "").strip()
                appointment_date = (appointment_date or "").strip()
                appointment_time = (appointment_time or "").strip()
                reason = (reason or "").strip()
                if not (patient_name and doctor_name and appointment_date and appointment_time and reason):
                    return "Please fill in all the fields."
                if not re.fullmatch(r"[A-Za-z ]+", patient_name):
                    return "Patient name should contain only letters and spaces."
                try:
                    # Parse the appointment date and time strings
                    appointment_date_obj = datetime.strptime(appointment_date, "%Y-%m-%d")
                except ValueError:
                    return "Appointment date must be in YYYY-MM-DD format."
                try:
                    appointment_time_obj = datetime.strptime(appointment_time, "%H:%M")
                except ValueError:
                    return "Appointment time must be in HH:MM (24-hour) format."
                
                # Combine date and time into a single datetime object
                appointment_datetime = datetime.combine(appointment_date_obj.date(), appointment_time_obj.time())
                now = datetime.now()
                if appointment_datetime <= now:
                    return "Appointment date/time has already passed. Please select a future date and time."
                
                confirmation = (f"Appointment booked for {patient_name} with {doctor_name} on {appointment_date} at {appointment_time}.\n\n"
                                f"Reason: {reason}")
                return confirmation


            book_button.click(book_appointment,
                              inputs=[patient_name_appt, doctor_name, appointment_date, appointment_time, reason],
                              outputs=booking_output)

        with gr.Tab("Patient Sessions"):
            gr.Markdown("### Search Patient Sessions")
            search_name = gr.Textbox(label="Patient Name (optional)")
            search_date = gr.Textbox(label="Date (YYYY-MM-DD, optional)")
            search_button = gr.Button("Search")
            sessions_output = gr.Textbox(label="Sessions", interactive=False)

            def search_sessions(name, date):
                sessions = get_patient_sessions(filter_name=name, filter_date=date)
                if sessions:
                    output = "\n\n".join([f"Patient: {s[0]}\nAge: {s[1]}\nGender: {s[2]}\nSymptoms: {s[3]}\nDiagnosis: {s[4]}\nMedication: {s[5]}\nTherapy: {s[6]}\nSummary: {s[7]}\nExplanation: {s[8]}\nReport: {s[9]}\nSession Time: {s[10]}" for s in sessions])
                    return output
                else:
                    return "No sessions found."

            search_button.click(search_sessions, inputs=[search_name, search_date], outputs=sessions_output)

    

    def handle_register(username, password, full_name, email):
        return register_user(username, password, full_name, email)

    def go_to_register_page():
        return gr.update(visible=False), gr.update(visible=True)

    def back_to_login_page():
        return gr.update(visible=True), gr.update(visible=False)


    go_to_register.click(go_to_register_page, outputs=[login_page, register_page])
    register_btn.click(handle_register,
                       inputs=[new_username, new_password, full_name, email],
                       outputs=register_output)
    back_to_login.click(back_to_login_page, outputs=[login_page, register_page])


    
    
    
    def toggle_profile(current_visible):
        #print("toggle_profile called with user:", user_session.value)  # Debug print
        if not user_session.value:
            return gr.update(visible=False), False, ""
        new_visible = not current_visible
        info = get_user_info(user_session.value) if new_visible else ""
        return gr.update(visible=new_visible), new_visible, info
    
   
    
            


    # Connect profile button click with correct input order:
    profile_button.click(
        toggle_profile,
        inputs=[profile_visible],
        outputs=[profile_info_box, profile_visible, profile_info]
    )


    # Handle login: update previous patients dropdown
    def handle_login(username, password):
        prev_choices = []
        if login_user(username, password):
            user_session.value = username
            prev_choices = get_previous_patients()
            return (
                f"Welcome, {username}!",           # login_output
                gr.update(visible=True),           # main_panel
                gr.update(visible=False),          # login_page
                gr.update(visible=True),           # header_row
                gr.update(choices=prev_choices, value=None),   # previous_patient
                "",                                # patient_name_input
                None,                              # age_input
                None,                              # gender_input
                "",                                # symptoms_input
                "",                                # diagnosis_textbox
                "",                                # diagnosis_conf_html
                "",                                # medication_textbox
                "",                                # medication_conf_html
                "",                                # therapy_textbox
                "",                                # therapy_conf_html
                "",                                # summary_textbox
                "",                                # explanation_textbox
                gr.update(visible=False),          # generate_report_btn
                None,                              # report_download
                "",                                # session_data_state
                "",                                # search_name (Patient Sessions tab)
                "",                                # search_date (Patient Sessions tab)
                "",                                # booking_output (Book an Appointment tab)
                "",                                # patient_name_appt (Booking tab field)
                "",                                # appointment_date (Booking tab field)
                "",                                # appointment_time (Booking tab field)
                "",                                # reason (Booking tab field)
                gr.update(visible=True)            # profile_container: show profile icon now
            )
        else:
            return (
                "Invalid credentials.",            # login_output
                gr.update(),                       # main_panel
                gr.update(),                       # login_page
                gr.update(),                       # header_row
                gr.update(choices=[], value=None),   # previous_patient (cleared)
                "",                                # patient_name_input
                None,                              # age_input
                None,                              # gender_input
                "",                                # symptoms_input
                "",                                # diagnosis_textbox
                "",                                # diagnosis_conf_html
                "",                                # medication_textbox
                "",                                # medication_conf_html
                "",                                # therapy_textbox
                "",                                # therapy_conf_html
                "",                                # summary_textbox
                "",                                # explanation_textbox
                gr.update(visible=False),          # generate_report_btn
                None,                              # report_download
                "",                                # session_data_state
                "",                                # search_name
                "",                                # search_date
                "",                                # booking_output
                "",                                # patient_name_appt
                "",                                # appointment_date
                "",                                # appointment_time
                "",                                # reason
                gr.update(visible=False)           # profile_container: hide profile icon on failure
            )





    login_btn.click(
    handle_login,
    inputs=[username_login, password_login],
    outputs=[
        login_output, main_panel, login_page, header_row, previous_patient,
        patient_name_input, age_input, gender_input, symptoms_input,
        diagnosis_textbox, diagnosis_conf_html,
        medication_textbox, medication_conf_html,
        therapy_textbox, therapy_conf_html,
        summary_textbox, explanation_textbox,
        generate_report_btn, report_download, session_data_state,
        search_name, search_date, booking_output, patient_name_appt, appointment_date, appointment_time, reason,
        profile_container  # new output for profile container
    ]
    )


    def handle_logout():
        user_session.value = ""
        return (
            gr.update(visible=False),  # Hide main_panel
            gr.update(visible=True),   # Show login_page
            gr.update(visible=False),  # Hide header_row
            gr.update(visible=False),  # Hide profile_info_box
            False,                     # Reset profile_visible
            "",                        # Clear profile_info
            "",                        # Clear login_output
            "",                        # Clear history_output
            "",                        # Clear username_login textbox
            "",                        # Clear password_login textbox
            "",                        # Clear new_username textbox (register page)
            "",                        # Clear new_password textbox (register page)
            "",                        # Clear full_name textbox (register page)
            "",                        # Clear email textbox (register page)
            gr.update(visible=False)   # profile_container: hide profile icon
        )


    logout_button.click(
    handle_logout,
    outputs=[
        main_panel,
        login_page,
        header_row,
        profile_info_box,
        profile_visible,
        profile_info,
        login_output,
        history_output,
        username_login,
        password_login,
        new_username,
        new_password,
        full_name,
        email
    ]
    )

def main():
    init_db()
    app.launch()

if __name__ == "__main__":
    main()