Spaces:
Sleeping
Sleeping
File size: 19,596 Bytes
e8d8f32 | 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 | """
data_normalizer.py - Smart Data Extraction + Normalization
===========================================================
Yeh module kya karta hai:
β
USD β INR Cr convert karta hai
β
Million/Billion/Lakh β Cr convert karta hai
β
Timeline ko standard format mein laata hai
β
Customer, Order Type, Date accurately nikalta hai
β
SEBI Annexure-A table structure samajhta hai (jo aapke PDFs mein hai)
Real-world examples jo handle karta hai:
USD 1,346,210 β 1.14 Cr (@ Rs.85/USD)
βΉ1,587.10 million β 158.71 Cr
Rs. 890.69 Million β 89.07 Cr
Rs. 17,750 Million β 1775.00 Cr
INR 20.58 Crores β 20.58 Cr
Rs. 1,827.33 β 1827.33 Cr (already in Cr)
Delivery by Oct 2027 β Oct 2027
Within 8 months β 8 Months
910 Days β 910 Days (~30 Months)
4 (04) Years β 4 Years (~48 Months)
18 Months β 18 Months
"""
import re
# βββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
# π± CURRENCY CONVERSION RATES
# Inhe update karte rahna agar rate badal jaye
# βββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
USD_TO_INR = 85.0 # 1 USD = Rs. 85 approx
MILLION_TO_CR = 0.1 # 1 Million = 0.1 Cr (10 Lakh)
BILLION_TO_CR = 100.0 # 1 Billion = 100 Cr
LAKH_TO_CR = 0.01 # 1 Lakh = 0.01 Cr
MN_TO_CR = 0.1 # MN = Million shorthand
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
# π° MAIN FUNCTION: Order Value Normalize karo
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
def normalize_order_value(text: str) -> dict:
"""
Kisi bhi format ki value ko INR Crore mein convert karta hai.
Input examples:
"USD 1,346,210"
"βΉ1,587.10 million (Including GST)"
"Rs. 890.69 Million"
"Rs. 1,827.33 Cr"
"INR 20.58 Crores"
"Rs. 17,750 Million"
Return:
{
"original": "USD 1,346,210",
"inr_crore": 1.14,
"inr_crore_display": "1.14 Cr",
"currency": "USD",
"note": "Converted @ Rs.85/USD"
}
"""
if not text or text.strip() in ["N/A", "", "-", "β"]:
return _empty_value()
clean = text.strip()
# ββ Pattern 1: USD amount ββ
# Matches: "USD 1,346,210" or "USD 1346210" or "US$ 1,346,210"
usd_match = re.search(
r"USD\s*[\$]?\s*([\d,]+(?:\.\d+)?)",
clean, re.IGNORECASE
)
if usd_match:
usd_val = float(usd_match.group(1).replace(",", ""))
inr_val = usd_val * USD_TO_INR
inr_cr = round(inr_val / 1e7, 2) # Convert to Crore
return {
"original": clean,
"inr_crore": inr_cr,
"inr_crore_display": f"{inr_cr:.2f} Cr",
"currency": "USD",
"note": f"USD {usd_val:,.0f} Γ Rs.{USD_TO_INR}/USD Γ· 1Cr"
}
# ββ Pattern 2: Rupees in Million ββ
# Matches: "Rs. 890.69 Million", "βΉ1,587.10 million", "INR 17,750 Million", "Rs. 17,750 Mn"
million_match = re.search(
r"(?:Rs\.?|βΉ|INR|INR\s*)?\s*([\d,]+(?:\.\d+)?)\s*(?:million|mn\.?)\b",
clean, re.IGNORECASE
)
if million_match:
mn_val = float(million_match.group(1).replace(",", ""))
cr_val = round(mn_val * MILLION_TO_CR, 2)
return {
"original": clean,
"inr_crore": cr_val,
"inr_crore_display": f"{cr_val:.2f} Cr",
"currency": "INR",
"note": f"Rs.{mn_val} Mn Γ· 10 = {cr_val} Cr"
}
# ββ Pattern 3: Rupees in Billion ββ
# Matches: "Rs. 2.5 Billion", "INR 1 Billion"
billion_match = re.search(
r"(?:Rs\.?|βΉ|INR)?\s*([\d,]+(?:\.\d+)?)\s*(?:billion|bn\.?)\b",
clean, re.IGNORECASE
)
if billion_match:
bn_val = float(billion_match.group(1).replace(",", ""))
cr_val = round(bn_val * BILLION_TO_CR, 2)
return {
"original": clean,
"inr_crore": cr_val,
"inr_crore_display": f"{cr_val:.2f} Cr",
"currency": "INR",
"note": f"Rs.{bn_val} Bn Γ 100 = {cr_val} Cr"
}
# ββ Pattern 4: Already in Crore ββ
# Matches: "Rs. 1,827.33 Cr", "INR 20.58 Crores", "Rs. 775.27 Cr. approx"
# Also: "Rs 148 crores", "approx. Rs. 33 Crore"
crore_match = re.search(
r"(?:Rs\.?|βΉ|INR)?\s*([\d,]+(?:\.\d+)?)\s*(?:crore|cr\.?|crores)\b",
clean, re.IGNORECASE
)
if crore_match:
cr_val = round(float(crore_match.group(1).replace(",", "")), 2)
return {
"original": clean,
"inr_crore": cr_val,
"inr_crore_display": f"{cr_val:.2f} Cr",
"currency": "INR",
"note": "Already in Crore"
}
# ββ Pattern 5: Rupees in Lakh ββ
# Matches: "Rs. 50 Lakh", "Rs. 2.5 L"
lakh_match = re.search(
r"(?:Rs\.?|βΉ|INR)?\s*([\d,]+(?:\.\d+)?)\s*(?:lakh|lac|l)\b",
clean, re.IGNORECASE
)
if lakh_match:
lakh_val = float(lakh_match.group(1).replace(",", ""))
cr_val = round(lakh_val * LAKH_TO_CR, 4)
return {
"original": clean,
"inr_crore": cr_val,
"inr_crore_display": f"{cr_val:.4f} Cr",
"currency": "INR",
"note": f"Rs.{lakh_val} L Γ· 100 = {cr_val} Cr"
}
# ββ Fallback: Kuch match nahi hua ββ
return {
"original": clean,
"inr_crore": None,
"inr_crore_display": clean, # As-is rakh do
"currency": "UNKNOWN",
"note": "Auto-convert nahi hua, manual check karo"
}
def _empty_value():
return {
"original": "", "inr_crore": None,
"inr_crore_display": "N/A", "currency": "", "note": ""
}
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
# β° TIMELINE NORMALIZE karo
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
def normalize_timeline(text: str) -> dict:
"""
Kisi bhi format ki timeline ko standard format mein convert karta hai.
Input examples:
"910 Days" β {"display": "910 Days (~30 Months)", "months": 30}
"4 (04) Years" β {"display": "4 Years (~48 Months)", "months": 48}
"Within 18 Months" β {"display": "18 Months", "months": 18}
"Delivery to be completed by Aug 31, 2026" β {"display": "Aug 2026", "months": None}
"Delivery by October 2027" β {"display": "Oct 2027", "months": None}
"Within 8 months from LOA" β {"display": "8 Months from LOA", "months": 8}
"11 months from Letter of Acceptance" β same pattern
"""
if not text or text.strip() in ["N/A", "", "-"]:
return {"display": "N/A", "months": None, "original": text}
clean = text.strip()
months = None
display = clean # Default: as-is
# ββ Pattern 1: Days ββ
# "910 Days", "30 days"
days_match = re.search(r"(\d+)\s*days?\b", clean, re.IGNORECASE)
if days_match:
days = int(days_match.group(1))
months = round(days / 30.4)
display = f"{days} Days (~{months} Months)"
return {"display": display, "months": months, "original": clean}
# ββ Pattern 2: Years ββ
# "4 Years", "4 (04) Years", "Four (04) Years", "2.5 years"
years_match = re.search(
r"(\d+(?:\.\d+)?)\s*(?:\(\d+\)\s*)?years?\b",
clean, re.IGNORECASE
)
if years_match:
years = float(years_match.group(1))
months = round(years * 12)
display = f"{int(years) if years == int(years) else years} Years (~{months} Months)"
return {"display": display, "months": months, "original": clean}
# ββ Pattern 3: Months ββ
# "18 Months", "Within 18 Months", "8 months from LOA"
months_match = re.search(r"(\d+)\s*months?\b", clean, re.IGNORECASE)
if months_match:
months = int(months_match.group(1))
# Context nikalo - "from LOA" jaisi cheez
context_match = re.search(
r"months?\s+(from\s+[^\.,]{3,40})", clean, re.IGNORECASE
)
suffix = f" {context_match.group(1)}" if context_match else ""
display = f"{months} Months{suffix}"
return {"display": display, "months": months, "original": clean}
# ββ Pattern 4: Specific Date ββ
# "Delivery by October 2027", "by August 31, 2026", "by April 2026"
date_match = re.search(
r"(?:by|until|before|on)\s+"
r"((?:Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)[a-z]*"
r"(?:\s+\d{1,2},?)?\s+\d{4})",
clean, re.IGNORECASE
)
if date_match:
display = date_match.group(1).strip()
# Short month format: "October 2027" β "Oct 2027"
display = _shorten_month(display)
return {"display": display, "months": None, "original": clean}
# ββ Fallback ββ
return {"display": clean[:60], "months": None, "original": clean}
def _shorten_month(date_str: str) -> str:
"""'October 2027' β 'Oct 2027'"""
month_map = {
"january": "Jan", "february": "Feb", "march": "Mar",
"april": "Apr", "may": "May", "june": "Jun",
"july": "Jul", "august": "Aug", "september": "Sep",
"october": "Oct", "november": "Nov", "december": "Dec"
}
for full, short in month_map.items():
date_str = re.sub(full, short, date_str, flags=re.IGNORECASE)
return date_str
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
# π SEBI ANNEXURE-A TABLE PARSER
# Yeh special function hai jo BSE filings ki standard table padh sakta hai
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
def parse_sebi_annexure_table(text: str) -> dict:
"""
SEBI Regulation 30 ke standard Annexure-A table se data nikalta hai.
Yeh table structure handle karta hai:
Row 1: Name of entity awarding β Customer
Row 2: Significant terms β Order Description
Row 4: Nature of order β Order Type
Row 6: Time period β Timeline
Row 7: Broad consideration / size β Order Value
Yeh almost SABB BSE filings mein hota hai!
"""
result = {
"customer": "",
"order_description": "",
"order_type": "",
"timeline_raw": "",
"value_raw": "",
"domestic_or_intl": "",
}
if not text:
return result
# Normalize whitespace
text_clean = re.sub(r"\s+", " ", text)
# ββ Row 1: Customer/Entity Name ββ
cust_patterns = [
r"(?:name of the entity awarding[^:]*?[:]\s*)([^\n\r]{5,100}?)(?:\n|\r|$|\d+\.)",
r"(?:entity awarding[^:]*?[:]\s*)([^\n\r]{5,100}?)(?:\n|\r|$)",
r"(?:awarded by|awarded from)[:\s]+([A-Z][^\n\r]{5,100}?)(?:\n|\r|$)",
]
for p in cust_patterns:
m = re.search(p, text_clean, re.IGNORECASE)
if m:
val = m.group(1).strip()
# Tabular "Details" header skip karo
if len(val) > 3 and "particular" not in val.lower():
result["customer"] = val[:120]
break
# ββ Row 2: Order Description / Terms ββ
desc_patterns = [
r"(?:significant terms[^:]*?[:]\s*)([^\n\r]{10,300}?)(?:\n|\r|$|\d+\.)",
r"(?:terms and conditions[^:]*?[:]\s*)([^\n\r]{10,300}?)(?:\n|\r|$)",
]
for p in desc_patterns:
m = re.search(p, text_clean, re.IGNORECASE)
if m:
result["order_description"] = m.group(1).strip()[:300]
break
# ββ Row 4: Nature of Order ββ
nature_patterns = [
r"(?:nature of order[^:]*?[:]\s*)([^\n\r]{5,200}?)(?:\n|\r|$|\d+\.)",
]
for p in nature_patterns:
m = re.search(p, text_clean, re.IGNORECASE)
if m:
result["order_type"] = m.group(1).strip()[:150]
break
# ββ Domestic/International ββ
dom_match = re.search(
r"(?:domestic or international[^:]*?[:]\s*)([^\n\r]{3,30}?)(?:\n|\r|$|\d+\.)",
text_clean, re.IGNORECASE
)
if dom_match:
result["domestic_or_intl"] = dom_match.group(1).strip()
# ββ Row 6: Timeline ββ
time_patterns = [
r"(?:time period[^:]*?[:]\s*)([^\n\r]{5,200}?)(?:\n|\r|$|\d+\.)",
r"(?:period by which[^:]*?[:]\s*)([^\n\r]{5,200}?)(?:\n|\r|$)",
r"(?:to be executed[^:]*?[:]\s*)([^\n\r]{5,200}?)(?:\n|\r|$)",
]
for p in time_patterns:
m = re.search(p, text_clean, re.IGNORECASE)
if m:
result["timeline_raw"] = m.group(1).strip()[:200]
break
# ββ Row 7: Order Value ββ
value_patterns = [
r"(?:broad consideration[^:]*?[:]\s*)([^\n\r]{5,200}?)(?:\n|\r|$|\d+\.)",
r"(?:size of the order[^:]*?[:]\s*)([^\n\r]{5,200}?)(?:\n|\r|$)",
r"(?:consideration or size[^:]*?[:]\s*)([^\n\r]{5,200}?)(?:\n|\r|$)",
r"(?:value of order[^:]*?[:]\s*)([^\n\r]{5,200}?)(?:\n|\r|$)",
]
for p in value_patterns:
m = re.search(p, text_clean, re.IGNORECASE)
if m:
result["value_raw"] = m.group(1).strip()[:200]
break
return result
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
# π― MASTER FUNCTION - Sab ek jagah
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
def extract_and_normalize(pdf_text: str, snippet: str, filing_meta: dict) -> dict:
"""
PDF text se FULL normalized record banata hai.
Pehle SEBI Annexure table parse karta hai (most reliable),
phir value + timeline normalize karta hai.
Return: Complete record dict for CSV
"""
# Step 1: SEBI standard table parse karo
annexure = parse_sebi_annexure_table(pdf_text)
# Step 2: Value normalize karo
raw_value = annexure["value_raw"] or snippet
value_info = normalize_order_value(raw_value)
# Step 3: Timeline normalize karo
raw_timeline = annexure["timeline_raw"] or ""
timeline_info = normalize_timeline(raw_timeline)
# Step 4: Customer finalize
customer = annexure["customer"]
if not customer:
# Fallback: snippet se try karo
m = re.search(
r"(?:from|by)\s+([A-Z][A-Za-z\s&\-]{5,60}"
r"(?:Ltd\.?|Limited|Authority|Board|NHAI|DDA|NTPC|ONGC|RAIL|NAVY|ARMY))",
snippet
)
if m:
customer = m.group(1).strip()
# Step 5: Order type finalize
order_type = annexure["order_type"] or filing_meta.get("title", "")
# Title se Order type nikalo agar table mein nahi mila
if "LOA" in order_type or "Letter of Award" in (pdf_text + snippet):
order_type = "Letter of Award"
elif "LOI" in order_type or "Letter of Intent" in (pdf_text + snippet):
order_type = "Letter of Intent"
elif "Work Order" in (pdf_text + snippet):
order_type = "Work Order"
return {
"Stock Name": filing_meta.get("stock", ""),
"Exchange": filing_meta.get("exchange", ""),
"Filing Title": filing_meta.get("title", ""),
"Customer": customer or "N/A",
"Order Description": annexure["order_description"] or "N/A",
"Order Type": order_type or "N/A",
"Domestic / International": annexure["domestic_or_intl"] or "N/A",
"Date Order Received": filing_meta.get("date", ""),
# β
NORMALIZED VALUES
"Order Value (INR Cr)": value_info["inr_crore_display"],
"Value (Raw)": value_info["original"],
"Conversion Note": value_info["note"],
# β
NORMALIZED TIMELINE
"Timeline": timeline_info["display"],
"Timeline (Months)": str(timeline_info["months"]) if timeline_info["months"] else "N/A",
"Timeline (Raw)": timeline_info["original"],
"Source PDF": filing_meta.get("pdf_url", ""),
"Screener Date": filing_meta.get("date", ""),
"Scraped At": __import__("datetime").datetime.now().strftime("%Y-%m-%d %H:%M"),
}
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
# π§ͺ TEST - Seedha run karo to verify
# ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
if __name__ == "__main__":
print("=" * 55)
print("π° ORDER VALUE CONVERSION TESTS")
print("=" * 55)
test_values = [
"USD 1,346,210",
"USD 1346210.00",
"βΉ1,587.10 million (Including GST)",
"Rs. 890.69 Million (including GST)",
"Rs. 17,750 Million",
"INR 20.58 Crores (Including Tax)",
"Rs. 1,827.33 Cr",
"Rs. 148 crores",
"Rs. 775.27 Cr. approx.",
"Rs. 33 Crore (Excl. GST)",
"Rs. 2.13 crores (excluding taxes)",
"Rs. 437.79 Cr. (Approx.)",
"Rs. 1,582.11", # Pure number - Cr assume karein
]
for val in test_values:
result = normalize_order_value(val)
print(f" Input : {val}")
print(f" Output: {result['inr_crore_display']} [{result['note']}]")
print()
print("=" * 55)
print("β° TIMELINE NORMALIZATION TESTS")
print("=" * 55)
test_timelines = [
"910 Days",
"Four (04) Years",
"4 (04) Years",
"Within 18 Months from signing of PPA",
"Delivery to be completed by August 31, 2026",
"Delivery by October 2027",
"Within 8 months from the Letter of Award",
"11 months from the date of issue of Letter of Acceptance",
"Delivery to be completed by April 2026",
"25 years from commissioning",
]
for tl in test_timelines:
result = normalize_timeline(tl)
months_str = f" | {result['months']} months" if result["months"] else ""
print(f" Input : {tl}")
print(f" Output: {result['display']}{months_str}")
print()
|