File size: 9,368 Bytes
9406eac
6f14fd9
 
 
 
9406eac
5c06b65
9406eac
 
 
 
 
 
5c06b65
9406eac
 
5c06b65
9406eac
 
5c06b65
9406eac
 
 
 
c72b167
e657d8c
 
 
 
 
c72b167
e657d8c
c72b167
5c06b65
 
9406eac
5c06b65
 
9406eac
 
 
 
 
5c06b65
9406eac
 
c72b167
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9406eac
5c06b65
9406eac
 
 
5c06b65
 
 
 
 
 
 
 
 
 
 
 
 
9406eac
 
 
 
5c06b65
 
9406eac
 
5c06b65
 
 
 
 
 
9406eac
 
5c06b65
 
9406eac
 
5c06b65
c72b167
9406eac
5c06b65
9406eac
5c06b65
 
 
 
 
 
c72b167
9406eac
 
 
 
 
 
c72b167
9406eac
 
5c06b65
9406eac
5c06b65
9406eac
5c06b65
 
9406eac
 
5c06b65
9406eac
 
 
 
 
 
c72b167
9406eac
 
5c06b65
c72b167
9406eac
 
5c06b65
9406eac
5c06b65
9406eac
 
5c06b65
c72b167
5c06b65
c72b167
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e657d8c
5c06b65
 
 
 
 
 
 
 
c72b167
5c06b65
 
 
 
 
 
 
 
c72b167
 
 
5c06b65
c72b167
5c06b65
c72b167
 
 
 
5c06b65
c72b167
5c06b65
 
 
 
 
 
 
c72b167
5c06b65
 
 
 
 
c72b167
5c06b65
 
 
c72b167
 
5c06b65
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c72b167
5c06b65
 
 
 
 
c72b167
5c06b65
 
c72b167
5c06b65
 
 
 
 
9406eac
5c06b65
9406eac
5c06b65
 
 
 
 
 
9406eac
5c06b65
 
 
 
 
 
 
9406eac
 
5c06b65
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
import os
import tempfile
import time
import re
import logging
from datetime import datetime

import gradio as gr
import google.generativeai as genai
from PyPDF2 import PdfReader
from tika import parser

# Configure logging
LOG_FILE = "pdf_processor_log.txt"
logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s - %(levelname)s - %(message)s",
    handlers=[
        logging.StreamHandler(),
        logging.FileHandler(LOG_FILE)
    ]
)
logger = logging.getLogger("pdf_processor")

# Try Unstructured.io
try:
    from unstructured.partition.pdf import partition_pdf
    UNSTRUCTURED_AVAILABLE = True
except ImportError:
    UNSTRUCTURED_AVAILABLE = False
    logger.warning("unstructured.partition.pdf not available; skipping that method")

# Load Gemini API key from env (set in your Space Secrets)
API_KEY = os.getenv("GOOGLE_API_KEY")
if API_KEY:
    genai.configure(api_key=API_KEY)
else:
    logger.warning("GOOGLE_API_KEY not set in environment.")

EXTRACTED_TEXT = ""
PDF_SECTIONS = []
EXTRACTION_METHOD = ""


# --- Extraction Functions ---
def extract_text_with_unstructured(pdf_path):
    try:
        logger.info("Extracting via Unstructured.io...")
        elements = partition_pdf(filename=pdf_path, extract_images_in_pdf=False)
        sections, current = [], {"title": "Introduction", "content": ""}
        for e in elements:
            if hasattr(e, "text") and (t := e.text.strip()):
                if len(t) < 80 and (t.isupper() or t.endswith(":") or re.match(r"^[0-9]+\.?\s+", t)):
                    if current["content"]:
                        sections.append(current)
                    current = {"title": t, "content": ""}
                else:
                    current["content"] += t + "\n\n"
        if current["content"]:
            sections.append(current)
        return sections
    except Exception as e:
        # Bubble up so process_pdf can catch & log
        logger.error(f"Unstructured extraction error: {e}", exc_info=True)
        raise


def extract_text_with_pypdf(pdf_path):
    logger.info("Extracting via PyPDF2...")
    reader = PdfReader(pdf_path)
    full_text = ""
    for i, page in enumerate(reader.pages, start=1):
        txt = page.extract_text()
        if txt:
            full_text += f"\n\n--- Page {i} ---\n\n{txt}"
    parts = re.split(r"\n\s*([A-Z][A-Z\s]+:?|[0-9]+\.\s+[A-Z].*?)\s*\n", full_text)
    if len(parts) > 1:
        return [
            {"title": parts[i].strip(), "content": parts[i + 1].strip()}
            for i in range(1, len(parts), 2)
        ]
    return [{"title": "Document", "content": full_text}]


def extract_text_with_tika(pdf_path):
    logger.info("Extracting via Tika...")
    parsed = parser.from_file(pdf_path)
    lines = (parsed.get("content") or "").split("\n")
    sections, current = [], {"title": "Introduction", "content": ""}
    for ln in lines:
        ln = ln.strip()
        if not ln:
            continue
        if len(ln) < 80 and (ln.isupper() or ln.endswith(":") or re.match(r"^[0-9]+\.?\s+[A-Z]", ln)):
            if current["content"]:
                sections.append(current)
            current = {"title": ln, "content": ""}
        else:
            current["content"] += ln + "\n\n"
    if current["content"]:
        sections.append(current)
    return sections


# --- Gemini calls ---
def generate_greg_brockman_summary(content):
    model = genai.GenerativeModel("gemini-1.5-pro")
    prompt = f"""
You are an expert document analyst specializing in proposal evaluation.

# GREG BROCKMAN TEMPLATE STRUCTURE
1. GOAL: ...
... (rest of template) ...

CONTENT:
{content}
"""
    try:
        resp = model.generate_content(prompt)
        return resp.text, None
    except Exception as e:
        logger.error(f"Summary error: {e}")
        return None, str(e)


def answer_question_about_pdf(content, question):
    model = genai.GenerativeModel("gemini-1.5-pro")
    prompt = f"""
You are a precise document analysis assistant.

DOCUMENT CONTENT:
{content}

QUESTION: {question}
"""
    try:
        resp = model.generate_content(prompt)
        return resp.text, None
    except Exception as e:
        logger.error(f"Q&A error: {e}")
        return None, str(e)


# --- Handlers ---
def process_pdf(pdf_file, progress=gr.Progress()):
    global EXTRACTED_TEXT, PDF_SECTIONS, EXTRACTION_METHOD

    if not API_KEY:
        return None, None, "❌ Set GOOGLE_API_KEY in Secrets.", ""
    if pdf_file is None:
        return None, None, "❌ No file uploaded.", ""

    # Determine path & write bytes if needed
    tmp_dir = tempfile.gettempdir()
    # Case 1: NamedString (in‐memory) with .name & .data
    if hasattr(pdf_file, "name") and hasattr(pdf_file, "data"):
        path = os.path.join(tmp_dir, pdf_file.name)
        with open(path, "wb") as f:
            f.write(pdf_file.data)
    # Case 2: direct filepath (str)
    elif isinstance(pdf_file, str):
        path = pdf_file
    # Case 3: file‐like with .read()
    elif hasattr(pdf_file, "read"):
        path = os.path.join(tmp_dir, getattr(pdf_file, "name", "uploaded.pdf"))
        with open(path, "wb") as f:
            f.write(pdf_file.read())
    else:
        return None, None, "❌ Unrecognized upload type", ""

    # Try methods in order
    methods = []
    if UNSTRUCTURED_AVAILABLE:
        methods.append(("unstructured", extract_text_with_unstructured))
    methods += [
        ("pypdf", extract_text_with_pypdf),
        ("tika", extract_text_with_tika),
    ]

    sections = None
    last_err = ""
    for name, fn in methods:
        try:
            secs = fn(path)
            if secs:
                sections = secs
                EXTRACTION_METHOD = name
                break
        except Exception as e:
            last_err = f"{name} failed: {e}"
            logger.warning(last_err)

    if not sections:
        return None, None, "❌ Extraction failed", last_err

    # Combine & summarize
    combined, structure = "", ""
    for i, sec in enumerate(sections, 1):
        structure += f"{i}. {sec['title']}\n"
        chunk = f"## {sec['title']}\n{sec['content']}\n\n"
        combined += chunk if len(combined + chunk) < 30000 else f"## {sec['title']}\n[Truncated]\n\n"
    EXTRACTED_TEXT = combined
    PDF_SECTIONS = sections

    summary, err = generate_greg_brockman_summary(combined)
    if err:
        return None, structure, f"❌ {err}", combined

    return summary, structure, "✅ PDF processed", f"Used {EXTRACTION_METHOD}"

def ask_question(question):
    if not API_KEY:
        return "❌ Set GOOGLE_API_KEY in Secrets."
    if not EXTRACTED_TEXT:
        return "❌ Process a PDF first."
    if not question.strip():
        return "❌ Enter a question."

    ans, err = answer_question_about_pdf(EXTRACTED_TEXT, question)
    return ans if not err else f"❌ {err}"

def view_log():
    try:
        return open(LOG_FILE).read()
    except Exception as e:
        return f"Error reading log: {e}"

def save_summary(summary):
    if not summary:
        return "❌ No summary to save."
    fn = f"summary_{datetime.now():%Y%m%d_%H%M%S}.txt"
    with open(fn, "w", encoding="utf-8") as f:
        f.write(summary)
    return f"✅ Saved to {fn}"

def save_qa(question, answer):
    if not question or not answer:
        return "❌ Nothing to save."
    fn = f"qa_{datetime.now():%Y%m%d_%H%M%S}.txt"
    with open(fn, "w", encoding="utf-8") as f:
        f.write(f"Q: {question}\n\nA: {answer}")
    return f"✅ Saved to {fn}"


# --- Gradio UI ---
with gr.Blocks(title="PDF Analyzer with Gemini API") as app:
    gr.Markdown("# 📄 PDF Analyzer with Gemini API")
    gr.Markdown("Upload a PDF, get a Greg Brockman style summary, and ask questions.")

    with gr.Tab("Setup"):
        gr.Markdown("⚠️ Make sure `GOOGLE_API_KEY` is set in your Space's Secrets.")

    with gr.Tab("PDF Processing"):
        with gr.Row():
            pdf_file = gr.File(label="Upload PDF", file_types=[".pdf"])
            proc_btn = gr.Button("Process PDF", variant="primary")
            status = gr.Markdown("Awaiting upload…")
        summary_out = gr.Textbox(label="Summary", lines=15)
        structure_out = gr.Textbox(label="Structure", lines=8)
        log_info    = gr.Textbox(label="Internal Log", lines=5)
        proc_btn.click(
            fn=process_pdf,
            inputs=[pdf_file],
            outputs=[summary_out, structure_out, status, log_info]
        )
        save_sum_btn = gr.Button("Save Summary")
        save_sum_status = gr.Markdown()
        save_sum_btn.click(save_summary, inputs=[summary_out], outputs=[save_sum_status])

    with gr.Tab("Ask Questions"):
        question_in = gr.Textbox(label="Your Question", lines=2)
        ask_btn = gr.Button("Ask", variant="primary")
        answer_out = gr.Textbox(label="Answer", lines=10)
        ask_btn.click(ask_question, inputs=[question_in], outputs=[answer_out])
        save_qa_btn = gr.Button("Save Q&A")
        save_qa_status = gr.Markdown()
        save_qa_btn.click(save_qa, inputs=[question_in, answer_out], outputs=[save_qa_status])

    with gr.Tab("System Log"):
        refresh_btn = gr.Button("Refresh Log")
        sys_log = gr.Textbox(label="System Log", lines=20)
        refresh_btn.click(view_log, inputs=None, outputs=[sys_log])

if __name__ == "__main__":
    app.launch(server_name="0.0.0.0")