Method getTableStructureAndFormatFromFileImpl

src/Storages/StorageFile.cpp:1072–1118

}

std::pair<ColumnsDescription, String> StorageFile::getTableStructureAndFormatFromFileImpl(
    std::optional<String> format,
    const std::vector<String> & paths,
    const String & compression_method,
    const std::optional<FormatSettings> & format_settings,
    const ContextPtr & context,
    const std::optional<ArchiveInfo> & archive_info)
{
    if (format == "Distributed")
    {
        if (paths.empty())
            throw Exception(ErrorCodes::INCORRECT_FILE_NAME, "Cannot get table structure from file, because no files match specified name");

        return {ColumnsDescription(DistributedAsyncInsertSource(paths[0]).getOutputs().front().getHeader().getNamesAndTypesList()), *format};
    }

    if (((archive_info && archive_info->paths_to_archives.empty()) || (!archive_info && paths.empty()))
        && (!format || !FormatFactory::instance().checkIfFormatHasExternalSchemaReader(*format)))
    {
        if (format)
            throw Exception(
                ErrorCodes::CANNOT_EXTRACT_TABLE_STRUCTURE,
                "The table structure cannot be extracted from a {} format file, because there are no files with provided path. "
                "You can specify table structure manually", *format);

        throw Exception(
            ErrorCodes::CANNOT_EXTRACT_TABLE_STRUCTURE,
            "The data format cannot be detected by the contents of the files, because there are no files with provided path. "
            "You can specify the format manually");

    }

    if (archive_info)
    {
        ReadBufferFromArchiveIterator read_buffer_iterator(*archive_info, format, format_settings, context);
        if (format)
            return {readSchemaFromFormat(*format, format_settings, read_buffer_iterator, context), *format};

        return detectFormatAndReadSchema(format_settings, read_buffer_iterator, context);
    }

    ReadBufferFromFileIterator read_buffer_iterator(paths, format, compression_method, format_settings, context);
    if (format)
        return {readSchemaFromFormat(*format, format_settings, read_buffer_iterator, context), *format};

    return detectFormatAndReadSchema(format_settings, read_buffer_iterator, context);
}

ColumnsDescription StorageFile::getTableStructureFromFile(
    const DB::String & format,