ArachNode — rbz

import re
import scrapy

def clean_text(s: str) -> str:
    return re.sub(r"\s+", " ", (s or "").replace("\xa0", " ")).strip()

def clean_number(s: str) -> str:
    # keeps digits, dot, comma, minus; then remove thousands separators
    s = clean_text(s)
    s = re.sub(r"[^0-9,\.\-]", "", s)
    return s.replace(",", "")

class MySpider(scrapy.Spider):
    name = "rbz_rates"
    allowed_domains = ["rbz.co.zw"]

# Use whichever page actually contains that "baTab0" HTML.
    # If you know the exact page, replace this with it.
    start_urls = ["https://www.rbz.co.zw/"]

custom_settings = {
        "ROBOTSTXT_OBEY": True,
        "USER_AGENT": "Mozilla/5.0 (compatible; rbz-rates-scraper/1.0)",
    }

def parse(self, response):
        # The "Exchange Rates" table is inside <div id="baTab0"> ... <table> ... </table>
        tab = response.css("#baTab0")
        if not tab:
            self.logger.warning("Could not find #baTab0 on %s", response.url)
            return

table = tab.css("table").get()
        rows = tab.css("table tbody tr")

# Attempt to grab the date from the header line like:
        # "EXCHANGE RATES 27-02-2026"
        header_text = clean_text(" ".join(tab.css("strong *::text, strong::text").getall()))
        m = re.search(r"EXCHANGE\s+RATES\s+(\d{2}-\d{2}-\d{4})", header_text, flags=re.I)
        rates_date = m.group(1) if m else None

# Data rows look like: CURRENCY | BID | ASK | AVG
        # In your snippet, those start after the header rows, so we filter by
        # "has at least 4 cells with text".
        for tr in rows:
            tds = tr.css("td")
            if len(tds) < 4:
                continue

currency = clean_text(" ".join(tds[0].css("::text, *::text").getall()))
            bid = clean_number(" ".join(tds[1].css("::text, *::text").getall()))
            ask = clean_number(" ".join(tds[2].css("::text, *::text").getall()))
            avg = clean_number(" ".join(tds[3].css("::text, *::text").getall()))

# Skip header-ish rows
            if not currency or currency.upper() in {"CURRENCY", "INTERBANK RATES"}:
                continue
            # Skip blank numeric rows
            if not any([bid, ask, avg]):
                continue

yield {
                "source_url": response.url,
                "date": rates_date,
                "currency_pair": currency,
                "bid": float(bid) if bid else None,
                "ask": float(ask) if ask else None,
                "avg": float(avg) if avg else None,
            }