#include "HTTPDictionarySource.h" #include #include #include #include #include #include #include #include #include #include #include #include "DictionarySourceFactory.h" #include "DictionarySourceHelpers.h" #include "DictionaryStructure.h" #include "registerDictionaries.h" namespace DB { namespace ErrorCodes { extern const int LOGICAL_ERROR; } static const UInt64 max_block_size = 8192; HTTPDictionarySource::HTTPDictionarySource( const DictionaryStructure & dict_struct_, const Poco::Util::AbstractConfiguration & config, const std::string & config_prefix, Block & sample_block_, ContextPtr context_, bool check_config) : log(&Poco::Logger::get("HTTPDictionarySource")) , update_time{std::chrono::system_clock::from_time_t(0)} , dict_struct{dict_struct_} , url{config.getString(config_prefix + ".url", "")} , update_field{config.getString(config_prefix + ".update_field", "")} , format{config.getString(config_prefix + ".format")} , sample_block{sample_block_} , context(context_) , timeouts(ConnectionTimeouts::getHTTPTimeouts(context)) { if (check_config) context->getRemoteHostFilter().checkURL(Poco::URI(url)); const auto & credentials_prefix = config_prefix + ".credentials"; if (config.has(credentials_prefix)) { credentials.setUsername(config.getString(credentials_prefix + ".user", "")); credentials.setPassword(config.getString(credentials_prefix + ".password", "")); } const auto & headers_prefix = config_prefix + ".headers"; if (config.has(headers_prefix)) { Poco::Util::AbstractConfiguration::Keys config_keys; config.keys(headers_prefix, config_keys); header_entries.reserve(config_keys.size()); for (const auto & key : config_keys) { const auto header_key = config.getString(headers_prefix + "." + key + ".name", ""); const auto header_value = config.getString(headers_prefix + "." + key + ".value", ""); header_entries.emplace_back(std::make_tuple(header_key, header_value)); } } } HTTPDictionarySource::HTTPDictionarySource(const HTTPDictionarySource & other) : log(&Poco::Logger::get("HTTPDictionarySource")) , update_time{other.update_time} , dict_struct{other.dict_struct} , url{other.url} , header_entries{other.header_entries} , update_field{other.update_field} , format{other.format} , sample_block{other.sample_block} , context(Context::createCopy(other.context)) , timeouts(ConnectionTimeouts::getHTTPTimeouts(context)) { credentials.setUsername(other.credentials.getUsername()); credentials.setPassword(other.credentials.getPassword()); } BlockInputStreamPtr HTTPDictionarySource::createWrappedBuffer(std::unique_ptr http_buffer_ptr) { String http_request_compression_method_str = http_buffer_ptr->getCompressMethod(); auto in_ptr_wrapped = wrapReadBufferWithCompressionMethod(std::move(http_buffer_ptr), chooseCompressionMethod({}, http_request_compression_method_str)); auto input_stream = context->getInputFormat(format, *in_ptr_wrapped, sample_block, max_block_size); return std::make_shared>(input_stream, std::move(in_ptr_wrapped)); } void HTTPDictionarySource::getUpdateFieldAndDate(Poco::URI & uri) { if (update_time != std::chrono::system_clock::from_time_t(0)) { auto tmp_time = update_time; update_time = std::chrono::system_clock::now(); time_t hr_time = std::chrono::system_clock::to_time_t(tmp_time) - 1; WriteBufferFromOwnString out; writeDateTimeText(hr_time, out); uri.addQueryParameter(update_field, out.str()); } else { update_time = std::chrono::system_clock::now(); } } BlockInputStreamPtr HTTPDictionarySource::loadAll() { LOG_TRACE(log, "loadAll {}", toString()); Poco::URI uri(url); auto in_ptr = std::make_unique( uri, Poco::Net::HTTPRequest::HTTP_GET, ReadWriteBufferFromHTTP::OutStreamCallback(), timeouts, 0, credentials, DBMS_DEFAULT_BUFFER_SIZE, header_entries); return createWrappedBuffer(std::move(in_ptr)); } BlockInputStreamPtr HTTPDictionarySource::loadUpdatedAll() { Poco::URI uri(url); getUpdateFieldAndDate(uri); LOG_TRACE(log, "loadUpdatedAll {}", uri.toString()); auto in_ptr = std::make_unique( uri, Poco::Net::HTTPRequest::HTTP_GET, ReadWriteBufferFromHTTP::OutStreamCallback(), timeouts, 0, credentials, DBMS_DEFAULT_BUFFER_SIZE, header_entries); return createWrappedBuffer(std::move(in_ptr)); } BlockInputStreamPtr HTTPDictionarySource::loadIds(const std::vector & ids) { LOG_TRACE(log, "loadIds {} size = {}", toString(), ids.size()); auto block = blockForIds(dict_struct, ids); ReadWriteBufferFromHTTP::OutStreamCallback out_stream_callback = [block, this](std::ostream & ostr) { WriteBufferFromOStream out_buffer(ostr); auto output_stream = context->getOutputStreamParallelIfPossible(format, out_buffer, sample_block); formatBlock(output_stream, block); }; Poco::URI uri(url); auto in_ptr = std::make_unique( uri, Poco::Net::HTTPRequest::HTTP_POST, out_stream_callback, timeouts, 0, credentials, DBMS_DEFAULT_BUFFER_SIZE, header_entries); return createWrappedBuffer(std::move(in_ptr)); } BlockInputStreamPtr HTTPDictionarySource::loadKeys(const Columns & key_columns, const std::vector & requested_rows) { LOG_TRACE(log, "loadKeys {} size = {}", toString(), requested_rows.size()); auto block = blockForKeys(dict_struct, key_columns, requested_rows); ReadWriteBufferFromHTTP::OutStreamCallback out_stream_callback = [block, this](std::ostream & ostr) { WriteBufferFromOStream out_buffer(ostr); auto output_stream = context->getOutputStreamParallelIfPossible(format, out_buffer, sample_block); formatBlock(output_stream, block); }; Poco::URI uri(url); auto in_ptr = std::make_unique( uri, Poco::Net::HTTPRequest::HTTP_POST, out_stream_callback, timeouts, 0, credentials, DBMS_DEFAULT_BUFFER_SIZE, header_entries); return createWrappedBuffer(std::move(in_ptr)); } bool HTTPDictionarySource::isModified() const { return true; } bool HTTPDictionarySource::supportsSelectiveLoad() const { return true; } bool HTTPDictionarySource::hasUpdateField() const { return !update_field.empty(); } DictionarySourcePtr HTTPDictionarySource::clone() const { return std::make_unique(*this); } std::string HTTPDictionarySource::toString() const { Poco::URI uri(url); return uri.toString(); } void registerDictionarySourceHTTP(DictionarySourceFactory & factory) { auto create_table_source = [=](const DictionaryStructure & dict_struct, const Poco::Util::AbstractConfiguration & config, const std::string & config_prefix, Block & sample_block, ContextPtr context, const std::string & /* default_database */, bool check_config) -> DictionarySourcePtr { if (dict_struct.has_expressions) throw Exception(ErrorCodes::LOGICAL_ERROR, "Dictionary source of type `http` does not support attribute expressions"); auto context_local_copy = copyContextAndApplySettings(config_prefix, context, config); return std::make_unique( dict_struct, config, config_prefix + ".http", sample_block, context_local_copy, check_config); }; factory.registerSource("http", create_table_source); } }