Method URI

src/IO/S3Common.cpp:324–442

    }

    URI::URI(const Poco::URI & uri_, bool parse_region)
    {
        /// Case when bucket name represented in domain name of S3 URL.
        /// E.g. (https://bucket-name.s3.Region.amazonaws.com/key)
        /// https://docs.aws.amazon.com/AmazonS3/latest/dev/VirtualHosting.html#virtual-hosted-style-access

        static const RE2 virtual_hosted_style_pattern(R"((.+)\.(s3|cos|tos)([.\-][a-z0-9\-.:]+))");


        /// Case when bucket name and key represented in path of S3 URL.
        /// E.g. (https://s3.Region.amazonaws.com/bucket-name/key)
        /// https://docs.aws.amazon.com/AmazonS3/latest/dev/VirtualHosting.html#path-style-access
        static const RE2 path_style_pattern("^/([^/]*)/(.*)");

        static constexpr auto S3 = "S3";
        static constexpr auto COSN = "COSN";
        static constexpr auto COS = "COS";
        static constexpr auto TOS = "TOS";


        uri = uri_;
        storage_name = S3;

        if (uri.getHost().empty())
            throw Exception("Host is empty in S3 URI: " + uri.toString(), ErrorCodes::BAD_ARGUMENTS);

        if (isS3Scheme(uri.getScheme()))
        {
            // URI has format s3://bucket/key
            endpoint = "";
            bucket = uri.getAuthority();
            validateBucket(bucket, uri);
            if (uri.getPath().length() <= 1)
                throw Exception("Invalid S3 URI: no key: " + uri.toString(), ErrorCodes::BAD_ARGUMENTS);
            if (!uri.getQuery().empty())
                key = uri.getPathAndQuery().substr(1);
            else
                key = uri.getPath().substr(1);
            is_virtual_hosted_style = false;
            return;
        }

        String name;
        String endpoint_authority_from_uri;
        String endpoint_without_scheme;

        if (re2::RE2::FullMatch(uri.getAuthority(), virtual_hosted_style_pattern, &bucket, &name, &endpoint_authority_from_uri))
        {
            is_virtual_hosted_style = true;
            endpoint_without_scheme = name + endpoint_authority_from_uri;
            endpoint = uri.getScheme() + "://" + endpoint_without_scheme;

            /// S3 specification requires at least 3 and at most 63 characters in bucket name.
            /// https://docs.aws.amazon.com/awscloudtrail/latest/userguide/cloudtrail-s3-bucket-naming-requirements.html
            if (bucket.length() < 3 || bucket.length() > 63)
                throw Exception("Bucket name length is out of bounds in virtual hosted style S3 URI: "
                    + quoteString(bucket) + "(" + uri.toString() + ")", ErrorCodes::BAD_ARGUMENTS);