OpenBB-finance/OpenBB · error · OpenBBError

Expected ISO-8859-1 encoding but got: {r.encoding}

Error message

Expected ISO-8859-1 encoding but got: {r.encoding}

What it means

Raised after a successful 200 response when the response object's detected encoding is not exactly 'ISO-8859-1'. The provider hard-codes this expectation because the Treasury CSV endpoint historically returned that charset; if the server stops advertising it (or an intermediary rewrites the Content-Type header), the strict equality check fails even though the payload may be fine. Note the code then decodes with utf-8 regardless, so the check is about the header, not the bytes.

Source

Thrown at openbb_platform/providers/government_us/openbb_government_us/models/treasury_prices.py:92

            "Origin": "https://treasurydirect.gov",
            "User-Agent": get_random_agent(),
        }

        payload = (
            f"priceDateDay={query.date.day}"  # type: ignore
            f"&priceDateMonth={query.date.month}"  # type: ignore
            f"&priceDateYear={query.date.year}"  # type: ignore
            "&fileType=csv"
            "&csv=CSV+FORMAT"
        )

        r = make_request(url=url, method="POST", headers=HEADERS, data=payload)

        if r.status_code != 200:
            raise OpenBBError("Error with the request: " + str(r.status_code))

        if r.encoding != "ISO-8859-1":
            raise OpenBBError(f"Expected ISO-8859-1 encoding but got: {r.encoding}")

        return r.content.decode("utf-8")

    @staticmethod
    def transform_data(
        query: GovernmentUSTreasuryPricesQueryParams,
        data: str,
        **kwargs: Any,
    ) -> list[GovernmentUSTreasuryPricesData]:
        """Transform the data."""
        # pylint: disable=import-outside-toplevel
        from math import isnan  # noqa
        from io import StringIO
        from pandas import Index, read_csv, to_datetime

        try:
            if not data:
                raise EmptyDataError("Data not found")

View on GitHub (pinned to 3e071fcc2c)

Solutions

1. Confirm with curl -I that the endpoint's Content-Type header still declares charset=ISO-8859-1.
2. If upstream changed the charset permanently, update to a newer openbb-platform release where the check was adjusted, or patch the check locally to accept the new encoding.
3. Remove any intermediary (proxy/cache) that rewrites response headers and retest.
4. If you control a fork, decode based on r.content directly instead of gating on r.encoding.

Example fix

# before (strict header equality)
if r.encoding != "ISO-8859-1":
    raise OpenBBError(f"Expected ISO-8859-1 encoding but got: {r.encoding}")

# after (tolerate header drift, decode bytes directly)
return r.content.decode("utf-8", errors="replace")

try:
    res = obb.equity.gov.treasury_prices(date=d)
except OpenBBError as e:
    if "Expected ISO-8859-1" in str(e):
        # header drift from upstream/proxy; verify with curl -I and file an upstream issue
        log.warning("encoding gate hit for %s", d)
    else:
        raise

When it happens

Trigger: The Treasury endpoint changes its Content-Type charset (e.g. to utf-8 or omits charset so the HTTP library guesses differently); a corporate proxy or cache rewrites response headers; using a different HTTP backend whose encoding detection differs.

Common situations: Upstream server configuration changes after Treasury site updates; environments with MITM proxies (Zscaler, corporate squid) that normalize headers; library upgrades that change default charset sniffing.

Related errors

• Error with the request: {r.status_code}
• Server error fetching report {report_id} -> {csv_text}
• Method must be GET or POST
• Invalid base64-encoded token.
• methods must be a list of strings

AI-assisted analysis of OpenBB-finance/OpenBB@3e071fcc2c (2026-08-14). Data as JSON: /api/errors/a63ad48d120bddc1. Report an issue: GitHub.