Spaces:
Sleeping
Sleeping
MakPr016 commited on
Commit ·
265e719
1
Parent(s): eac74fb
Updated output
Browse files- .gitignore +1 -0
- rfq_parser.py +68 -3
.gitignore
CHANGED
|
@@ -15,3 +15,4 @@ venv/
|
|
| 15 |
build/
|
| 16 |
dist/
|
| 17 |
*.egg-info/
|
|
|
|
|
|
| 15 |
build/
|
| 16 |
dist/
|
| 17 |
*.egg-info/
|
| 18 |
+
response.json
|
rfq_parser.py
CHANGED
|
@@ -297,6 +297,64 @@ def _looks_like_item_continuation(table):
|
|
| 297 |
|
| 298 |
|
| 299 |
def _extract_rows(rows, idx_map, num_cols, seen_srs, items):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 300 |
for row in rows:
|
| 301 |
row_clean = [_clean(c) for c in row]
|
| 302 |
row_clean = (row_clean + [""] * num_cols)[:num_cols]
|
|
@@ -342,13 +400,20 @@ def _extract_rows(rows, idx_map, num_cols, seen_srs, items):
|
|
| 342 |
continue
|
| 343 |
seen_srs.add(key)
|
| 344 |
|
|
|
|
|
|
|
|
|
|
| 345 |
# --- NEW: classify the item ---
|
| 346 |
-
category = determine_item_category(desc, unit_val)
|
| 347 |
|
| 348 |
items.append({
|
| 349 |
"sr": sr_val if sr_val is not None else len(items) + 1,
|
| 350 |
-
"description": desc,
|
| 351 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 352 |
"qty": qty_val,
|
| 353 |
"unit_price": None,
|
| 354 |
"total_price": None,
|
|
|
|
| 297 |
|
| 298 |
|
| 299 |
def _extract_rows(rows, idx_map, num_cols, seen_srs, items):
|
| 300 |
+
def _parse_description_parts(raw_desc):
|
| 301 |
+
text = raw_desc.strip()
|
| 302 |
+
if not text:
|
| 303 |
+
return "", "", ""
|
| 304 |
+
|
| 305 |
+
# Pull dosage-like fragments such as "156 Mg/5ml" or "500 mg".
|
| 306 |
+
dosage_match = re.search(
|
| 307 |
+
r"\b\d+(?:\.\d+)?\s*(?:mg|mcg|g|iu|ml|mg/ml|mcg/ml|g/ml)\b(?:\s*/\s*\d+(?:\.\d+)?\s*ml)?",
|
| 308 |
+
text,
|
| 309 |
+
flags=re.IGNORECASE,
|
| 310 |
+
)
|
| 311 |
+
dosage = dosage_match.group(0) if dosage_match else ""
|
| 312 |
+
|
| 313 |
+
# Common dosage forms that appear in descriptions.
|
| 314 |
+
form_match = re.search(
|
| 315 |
+
r"\b(tablet|tab|capsule|cap|suspension|syrup|injection|inj|vial|ampoule|amp|drops|inhaler|ointment|cream|gel|lotion|suppository|supp|solution|powder|elixir|serum)\b",
|
| 316 |
+
text,
|
| 317 |
+
flags=re.IGNORECASE,
|
| 318 |
+
)
|
| 319 |
+
form = form_match.group(0) if form_match else ""
|
| 320 |
+
|
| 321 |
+
cleaned = text
|
| 322 |
+
for fragment in [dosage, form]:
|
| 323 |
+
if fragment:
|
| 324 |
+
cleaned = re.sub(re.escape(fragment), "", cleaned, flags=re.IGNORECASE)
|
| 325 |
+
cleaned = re.sub(r"\s{2,}", " ", cleaned).strip(" ,.-")
|
| 326 |
+
|
| 327 |
+
return cleaned, dosage, form
|
| 328 |
+
|
| 329 |
+
def _parse_pack_from_unit(raw_unit):
|
| 330 |
+
text = raw_unit.strip()
|
| 331 |
+
if not text:
|
| 332 |
+
return "", 0, ""
|
| 333 |
+
|
| 334 |
+
# Match patterns like "Pack of 20 Tablet" or "Box of 100".
|
| 335 |
+
pack_match = re.search(r"\b(pack|box|bottle|bag|tube|vial|ampoule|amp|ea|each|single unit)\b", text, flags=re.IGNORECASE)
|
| 336 |
+
unit_type = pack_match.group(0) if pack_match else ""
|
| 337 |
+
|
| 338 |
+
qty_match = re.search(r"\b(\d+(?:\.\d+)?)\b", text)
|
| 339 |
+
pack_size = 0
|
| 340 |
+
if qty_match:
|
| 341 |
+
try:
|
| 342 |
+
pack_size_val = float(qty_match.group(1))
|
| 343 |
+
pack_size = int(pack_size_val) if pack_size_val.is_integer() else pack_size_val
|
| 344 |
+
except Exception:
|
| 345 |
+
pack_size = 0
|
| 346 |
+
|
| 347 |
+
pack_unit = ""
|
| 348 |
+
trailing = text
|
| 349 |
+
if qty_match:
|
| 350 |
+
trailing = text[qty_match.end():]
|
| 351 |
+
if trailing:
|
| 352 |
+
m = re.search(r"\b([a-zA-Z]+(?:\s+[a-zA-Z]+)?)\b", trailing)
|
| 353 |
+
if m:
|
| 354 |
+
pack_unit = m.group(1).strip()
|
| 355 |
+
|
| 356 |
+
return unit_type.title() if unit_type else "", pack_size, pack_unit.title() if pack_unit else ""
|
| 357 |
+
|
| 358 |
for row in rows:
|
| 359 |
row_clean = [_clean(c) for c in row]
|
| 360 |
row_clean = (row_clean + [""] * num_cols)[:num_cols]
|
|
|
|
| 400 |
continue
|
| 401 |
seen_srs.add(key)
|
| 402 |
|
| 403 |
+
clean_desc, dosage, form = _parse_description_parts(desc)
|
| 404 |
+
unit_type, pack_size, pack_unit = _parse_pack_from_unit(unit_val)
|
| 405 |
+
|
| 406 |
# --- NEW: classify the item ---
|
| 407 |
+
category = determine_item_category(clean_desc or desc, unit_val)
|
| 408 |
|
| 409 |
items.append({
|
| 410 |
"sr": sr_val if sr_val is not None else len(items) + 1,
|
| 411 |
+
"description": clean_desc or desc,
|
| 412 |
+
"dosage": dosage,
|
| 413 |
+
"form": form.title() if form else "",
|
| 414 |
+
"pack_size": pack_size,
|
| 415 |
+
"pack_unit": pack_unit,
|
| 416 |
+
"unit": unit_type,
|
| 417 |
"qty": qty_val,
|
| 418 |
"unit_price": None,
|
| 419 |
"total_price": None,
|