File size: 7,873 Bytes
6d0080e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cf2feea
 
 
 
 
 
 
6d0080e
cf2feea
6d0080e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cf2feea
 
 
 
 
 
 
 
 
6d0080e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cf2feea
6d0080e
cf2feea
6d0080e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import gradio as gr
import requests
from bs4 import BeautifulSoup
import numpy as np
from sentence_transformers import SentenceTransformer
import faiss
from typing import List, Tuple
import re

model = SentenceTransformer('all-MiniLM-L6-v2')

doc_chunks = []
doc_embeddings = None
index = None
source_url = ""

def fetch_documentation(url: str) -> str:
    try:
        headers = {
            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
            'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
            'Accept-Language': 'en-US,en;q=0.5',
            'Accept-Encoding': 'gzip, deflate, br',
            'DNT': '1',
            'Connection': 'keep-alive',
            'Upgrade-Insecure-Requests': '1'
        }
        response = requests.get(url, headers=headers, timeout=15, allow_redirects=True)
        response.raise_for_status()
        
        soup = BeautifulSoup(response.content, 'html.parser')
        
        for script in soup(["script", "style", "nav", "footer", "header"]):
            script.decompose()
        
        text = soup.get_text()
        
        lines = (line.strip() for line in text.splitlines())
        chunks = (phrase.strip() for line in lines for phrase in line.split("  "))
        text = '\n'.join(chunk for chunk in chunks if chunk)
        
        return text
    except Exception as e:
        error_msg = str(e)
        if "403" in error_msg or "Forbidden" in error_msg:
            raise Exception(f"Access denied (403 Forbidden). This website blocks automated requests. Try: 1) Using the site's API if available, 2) A different documentation page, 3) GitHub raw content URLs work well (e.g., https://raw.githubusercontent.com/...)")
        elif "404" in error_msg:
            raise Exception(f"Page not found (404). Please check the URL is correct.")
        elif "timeout" in error_msg.lower():
            raise Exception(f"Request timeout. The website took too long to respond.")
        else:
            raise Exception(f"Error fetching URL: {error_msg}")

def chunk_text(text: str, chunk_size: int = 500, overlap: int = 50) -> List[str]:
    sentences = re.split(r'[.!?]+', text)
    chunks = []
    current_chunk = ""
    
    for sentence in sentences:
        sentence = sentence.strip()
        if not sentence:
            continue
            
        if len(current_chunk) + len(sentence) < chunk_size:
            current_chunk += sentence + ". "
        else:
            if current_chunk:
                chunks.append(current_chunk.strip())
            current_chunk = sentence + ". "
    
    if current_chunk:
        chunks.append(current_chunk.strip())
    
    return chunks

def process_documentation(url: str) -> str:
    global doc_chunks, doc_embeddings, index, source_url
    
    if not url:
        return "Please provide a URL"
    
    try:
        status = "Fetching documentation..."
        print(status)
        
        text = fetch_documentation(url)
        
        if len(text) < 100:
            return "Retrieved content is too short. Please check the URL."
        
        status = "Chunking text..."
        print(status)
        
        doc_chunks = chunk_text(text)
        
        if not doc_chunks:
            return "No content chunks created. The documentation might be empty."
        
        status = f"Creating embeddings for {len(doc_chunks)} chunks..."
        print(status)
        
        doc_embeddings = model.encode(doc_chunks, show_progress_bar=False)
        
        dimension = doc_embeddings.shape[1]
        index = faiss.IndexFlatL2(dimension)
        index.add(doc_embeddings.astype('float32'))
        
        source_url = url
        
        return f"Documentation processed successfully!\n\nStatistics:\n- Chunks created: {len(doc_chunks)}\n- Text length: {len(text)} characters\n- Ready to answer questions!"
        
    except Exception as e:
        return f"Error: {str(e)}"

def answer_question(question: str, top_k: int = 3) -> Tuple[str, str]:
    global doc_chunks, doc_embeddings, index, source_url
    
    if not question:
        return "Please enter a question", ""
    
    if index is None or not doc_chunks:
        return "Please process documentation first by entering a URL above", ""
    
    try:
        question_embedding = model.encode([question])
        
        distances, indices = index.search(question_embedding.astype('float32'), top_k)
        
        relevant_chunks = [doc_chunks[i] for i in indices[0]]
        
        context = "\n\n".join([f"[{i+1}] {chunk}" for i, chunk in enumerate(relevant_chunks)])
        
        answer = f"Based on the documentation at {source_url}:\n\n"
        answer += f"Relevant Information:\n\n{relevant_chunks[0]}"
        
        if len(relevant_chunks) > 1:
            answer += f"\n\nAdditional Context:\n\n{relevant_chunks[1]}"
        
        sources = "Retrieved Chunks:\n\n"
        for i, (chunk, dist) in enumerate(zip(relevant_chunks, distances[0])):
            sources += f"Chunk {i+1} (similarity: {1/(1+dist):.3f}):\n{chunk}\n\n---\n\n"
        
        return answer, sources
        
    except Exception as e:
        return f"Error: {str(e)}", ""

with gr.Blocks(theme=gr.themes.Soft(), title="Documentation RAG System") as demo:
    gr.Markdown("# Documentation RAG System\n\nEnter a documentation URL, process it, then ask questions about the content using AI-powered retrieval.")
    
    with gr.Row():
        with gr.Column():
            url_input = gr.Textbox(
                label="Documentation URL",
                placeholder="https://docs.python.org/3/tutorial/index.html",
                lines=1
            )
            process_btn = gr.Button("Process Documentation", variant="primary")
            status_output = gr.Textbox(
                label="Status",
                lines=6,
                interactive=False
            )
    
    gr.Markdown("---")
    
    with gr.Row():
        with gr.Column():
            question_input = gr.Textbox(
                label="Your Question",
                placeholder="What is this documentation about?",
                lines=3
            )
            
            top_k_slider = gr.Slider(
                minimum=1,
                maximum=5,
                value=3,
                step=1,
                label="Number of chunks to retrieve"
            )
            
            ask_btn = gr.Button("Ask Question", variant="primary")
    
    with gr.Row():
        with gr.Column():
            answer_output = gr.Textbox(
                label="Answer",
                lines=10,
                interactive=False
            )
        
        with gr.Column():
            sources_output = gr.Textbox(
                label="Source Chunks",
                lines=10,
                interactive=False
            )
    
    gr.Markdown("### Example URLs to try:")
    gr.Examples(
        examples=[
            ["https://raw.githubusercontent.com/python/cpython/main/README.rst"],
            ["https://docs.python.org/3/tutorial/introduction.html"],
            ["https://raw.githubusercontent.com/huggingface/transformers/main/README.md"],
            ["https://pytorch.org/docs/stable/torch.html"],
        ],
        inputs=url_input
    )
    
    process_btn.click(
        fn=process_documentation,
        inputs=[url_input],
        outputs=[status_output]
    )
    
    ask_btn.click(
        fn=answer_question,
        inputs=[question_input, top_k_slider],
        outputs=[answer_output, sources_output]
    )
    
    question_input.submit(
        fn=answer_question,
        inputs=[question_input, top_k_slider],
        outputs=[answer_output, sources_output]
    )

if __name__ == "__main__":
    demo.launch()