import requests, re import pymupdf fid = "1vZ4huyqmwJr1hTXb1qa7uTJIdKeBkkWu" sess = requests.Session() d = sess.get("https://drive.usercontent.google.com/download", params={"id": fid, "export": "download", "confirm": "t"}, headers={"User-Agent": "Mozilla/5.0"}, timeout=300) pdf = pymupdf.open(stream=d.content, filetype="pdf") print("pages:", pdf.page_count) VNRE = re.compile(r"[\u00e0\u00e1\u1ea3\u00e3\u1ea1\u0103\u1eb1\u1eaf\u1eb3\u1eb5\u1eb7\u00e2\u1ea7\u1ea5\u1ea9\u1eab\u1ead\u0111\u00e8\u00e9\u1ebb\u1ebd\u1eb9\u00ea\u1ec1\u1ebf\u1ec3\u1ec5\u1ec7\u00ec\u00ed\u1ec9\u0129\u1ecb\u00f2\u00f3\u1ecf\u00f5\u1ecd\u00f4\u1ed3\u1ed1\u1ed5\u1ed7\u1ed9\u01a1\u1edd\u1edb\u1edf\u1ee1\u1ee3\u00f9\u00fa\u1ee7\u0169\u1ee5\u01b0\u1eeb\u1ee9\u1eed\u1eef\u1ef1\u1ef3\u00fd\u1ef7\u1ef9\u1ef5]+", re.I) def get_code(t): t = (t or "").replace("*", "").strip() m = re.match(r"^(\d{1,3})\.(\d{2,3})\.(\d{2,3})$", t) if m: return f"{m.group(1)}.{m.group(2)}.{m.group(3)}" m = re.match(r"^(\d{1,2})[\. ](\d{3})[\. ](\d{3})$", t) if m: return f"{int(m.group(1))}.{m.group(2)}.{m.group(3)}" return None PRICE_RE = re.compile(r"^(\d{1,3}(?:,\d{3})+)$") MODEL_HDR = ("Model No.", "Model") BADG = {"Satin","Polished gold PVD","Satin black","Mocha","White","Black","SS304","Inox 304","Full overlay","Half overlay","Inset","Tr\u00f9m ngo\u00e0i","Tr\u00f9m n\u1eeda","L\u1ecdt l\u00f2ng","Steel","Th\u00e9p","Dust proof","WC indicator","T\u00ean h\u00e0ng","M\u00f4 t\u1ea3","M\u00e0u s\u1eafc","Color","M\u1ea7u s\u1eafc","M\u00e3 s\u1ed1","M\u00f4 s\u1ed1","Price/set","Price/pc","Gi\u00e1 b\u00e1n/b\u1ed9","Gi\u00e1 b\u00e1n/c\u00e1i","Gi\u00e1 b\u00e1n/thanh","S\u1ed1 v\u00f2ng bi","R 9.5","C\u00f3 \u0111i\u1ec1u ch\u1ec9nh","\u0110i\u1ec1u ch\u1ec9nh 3 chi\u1ec1u","\u0110i\u1ec1u ch\u1ec9nh","T\u1ea3i tr\u1ecdng","T\u1ea3i tr\u1ecdng t\u1ed1i \u0111a","Chi\u1ec1u cao","Chi\u1ec1u r\u1ed9ng","K\u00edch th\u01b0\u1edbc","B\u1ec1 m\u1eb7t","S\u01a1n t\u0129nh \u0111i\u1ec7n","Ngu\u1ed3n cung c\u1ea5p","T\u1ed1c \u0111\u1ed9","\u0110\u1ed9 \u1ed3n","G\u00f3c m\u1edf t\u1ed1i \u0111a","\u0110\u1ed9 d\u00e0y c\u00e1nh c\u1eeda","K\u00edch th\u01b0\u1edbc m\u1eb7t b\u00e0n","Ch\u1ea5t li\u1ec7u ch\u00ednh","Ch\u1ea5t li\u1ec7u"} ATTR = ("overall dim","k\u00edch th\u01b0\u1edbc t\u1ed5ng th\u1ec3","cabinet wid","chi\u1ec1u r\u1ed9ng t\u1ee7","loading","t\u1ea3i tr\u1ecdng","max opening","g\u00f3c m\u1edf","door thickness","\u0111\u1ed9 d\u00e0y c\u00e1nh","max cuttable","c\u1eaft ng\u1eafn","main material","ch\u1ea5t li\u1ec7u","side panel","suitable","ph\u00f9 h\u1ee3p","please specify","vui l\u00f2ng","the prices","gi\u00e1 ch\u01b0a","reference","certificate","powder coating","s\u01a1n t\u0129nh","input","ngu\u1ed3n cung","tabletop","m\u1eb7t b\u00e0n","height ","chi\u1ec1u cao","speed","t\u1ed1c \u0111\u1ed9","ultra-quiet","\u0111\u1ed9 \u1ed3n","soft-close","gi\u1ea3m ch\u1ea5n","(side","min ","max ","electr","\u0111i\u1ec7n","with","model no","tr\u1eafng ho\u1eb7c","vui l\u00f2ng \u0111\u1eb7t","ph\u1ee5 ki\u1ec7n tu\u1ef3 ch\u1ecdn","optional accessories","please order","m\u1ed7i b\u1ea3n l\u1ec1","white or black","tr\u1eafng ho\u1eb7c \u0111en") def is_bad_name(l): ll = (l or "").strip() if not ll or ll in BADG: return True llow = ll.lower() if any(llow.startswith(a) for a in ATTR): return True if re.match(r"^[\d.,/x*\u00b0\u00b7\-\s]*\d+[\d.,/x*\u00b0\u00b7\-\s]*(mm|kg|v|w|\u00b0)?$", llow): return True if len(ll) < 2: return True return False def parse_page(ipage): pg = pdf.load_page(ipage) lines = [l.strip() for l in pg.get_text().splitlines() if l.strip()] hdr = [j for j, l in enumerate(lines) if l in MODEL_HDR] out = [] for hi, h in enumerate(hdr): start = hdr[hi - 1] + 1 if hi > 0 else 0 pre = lines[start:h] name = "" for l in pre: if VNRE.search(l) and not is_bad_name(l): name = l; break if not name: for l in pre: if re.search(r"[A-Za-z]", l) and not is_bad_name(l): name = l; break if not name: name = "" end = hdr[hi + 1] if hi + 1 < len(hdr) else len(lines) j = h + 1 while j < end: l = lines[j]; c = get_code(l) if c: price = ""; color = ""; jj = j + 1 while jj < end: l2 = lines[jj]; c2 = get_code(l2) if c2: break m = PRICE_RE.match(l2.replace(" ", "")) if m: price = m.group(1) else: low = l2.lower() if low in ("white","mocha","satin","satin black","black","polished gold pvd","grey","gray") and not color: color = l2 jj += 1 out.append({"code": c, "name": name, "color": color, "price": price}) j = jj else: j += 1 seen = set(); res = [] for r in out: if r["code"] in seen: continue seen.add(r["code"]); res.append(r) return res RANGES = [(14,27,"B\u1ea3n l\u1ec1 t\u1ee7 b\u1ebfp"),(28,39,"Kh\u00f3a"),(40,47,"B\u1ed9 kh\u00f3a tay n\u1eafm"),(48,53,"Ph\u1ee5 ki\u1ec7n t\u1ee7 \u00e1o"),(54,57,"K\u1eb9p k\u00ednh"),(58,63,"B\u1ea3n l\u1ec1 c\u1eeda"),(64,69,"Kh\u00f3a \u0111i\u1ec7n t\u1eed"),(70,75,"Ray h\u1ed9p"),(76,87,"Ru\u1ed9t kh\u00f3a"),(88,95,"Ph\u1ee5 ki\u1ec7n t\u1ee7 b\u1ebfp"),(96,105,"Ph\u1ee5 ki\u1ec7n t\u1ee7 \u00e1o"),(106,109,"Chi\u1ebfu s\u00e1ng"),(110,113,"Tay n\u1eafm c\u1eeda k\u00ednh"),(118,118,"H\u1ec7 c\u1eeda tr\u01b0\u1ee3t"),(121,121,"Tay n\u1eafm"),(124,139,"Ray tr\u01b0\u1ee3t"),(140,145,"Tay n\u1eafm t\u1ee7"),(146,151,"Ph\u1ee5 ki\u1ec7n n\u1ed9i th\u1ea5t th\u00f4ng minh"),(152,185,"B\u1ea3n l\u1ec1 s\u00e0n"),(186,199,"Tay n\u1eafm"),(200,211,"Ph\u1ee5 ki\u1ec7n tho\u00e1t hi\u1ec3m"),(212,229,"Thanh nh\u00f4m"),(230,240,"Ph\u1ee5 ki\u1ec7n c\u1eeda k\u00ednh ph\u00f2ng t\u1eafm")] def cat_of(pdf_pg): cp = pdf_pg * 2 - 1 for a, b, v in RANGES: if a <= cp <= b: return v return "Ph\u1ee5 ki\u1ec7n" ALL = [] for i in range(pdf.page_count): rows = parse_page(i); cat = cat_of(i + 1) for r in rows: r["cat"] = cat; r["page"] = i + 1; ALL.append(r) print("total rows:", len(ALL)) seen = set(); UNIQ = [] for r in ALL: if r["code"] in seen: continue seen.add(r["code"]); UNIQ.append(r) print("unique:", len(UNIQ), "no-name:", sum(1 for r in UNIQ if not r["name"])) for i in [15,16,80,101,102,103]: print(f"-- page {i+1}:") for r in parse_page(i): print(f" {r['code']} | {r['name'][:62]} | c={r['color']} p={r['price']}")