@Luis-manzur

this is great for the sample_caller - but to make this truly work we will need to force CL to abide by the rate limits set out in a particular court right?

I think we need to create a replacement for this code in CL


def get_binary_content(
    download_url: str,
    site: AbstractSite,
) -> bytes | str:
    """Downloads the file, covering a few special cases such as invalid SSL
    certificates and empty file errors.

    :param download_url: The URL for the item you wish to download.
    :param site: Site object used to download data

    :return: The downloaded and cleaned content
    :raises: NoDownloadUrlError, UnexpectedContentTypeError, EmptyFileError
    """
    if not download_url:
        raise NoDownloadUrlError(download_url)

    # noinspection PyBroadException
    if site.method == "LOCAL":
        # "LOCAL" is the method when testing
        url = os.path.join(settings.MEDIA_ROOT, download_url)
        mr = MockRequest(url=url)
        r = mr.get()
        s = requests.Session()
    else:
        # some sites require a custom ssl_context, contained in the Site's
        # session. However, we can't send a request with both a
        # custom ssl_context and `verify = False`
        has_cipher = hasattr(site, "cipher")
        s = site.request["session"] if has_cipher else requests.session()

        if site.needs_special_headers:
            headers = site.request["headers"]
        else:
            headers = {"User-Agent": "CourtListener"}

        # Note that we do a GET even if site.method is POST. This is
        # deliberate.
        r = s.get(
            download_url,
            verify=has_cipher,  # WA has a certificate we don't understand
            headers=headers,
            cookies=site.cookies,
            timeout=300,
        )

        # test for empty files (thank you CA1)
        if len(r.content) == 0:
            raise EmptyFileError(f"EmptyFileError: '{download_url}'")

        # test for expected content type (thanks mont for nil)
        if site.expected_content_types:
            # Clean up content types like "application/pdf;charset=utf-8"
            # and 'application/octet-stream; charset=UTF-8'
            content_type = (
                r.headers.get("Content-Type").lower().split(";")[0].strip()
            )
            m = any(
                content_type in mime.lower()
                for mime in site.expected_content_types
            )

            if not m:
                court_str = site.court_id.split(".")[-1].split("_")[0]
                fingerprint = [f"{court_str}-unexpected-content-type"]
                msg = f"'{download_url}' '{content_type}' not in {site.expected_content_types}"
                raise UnexpectedContentTypeError(msg, fingerprint=fingerprint)

        # test for and follow meta redirects
        r = follow_redirections(r, s)
        r.raise_for_status()

    content = site.cleanup_content(r.content)

    return content

Uh oh!

add rate_limit base method #1496

Description

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions