| 1 | #include <Processors/Formats/Impl/TemplateRowInputFormat.h> |
| 2 | #include <Formats/FormatFactory.h> |
| 3 | #include <Formats/verbosePrintString.h> |
| 4 | #include <IO/Operators.h> |
| 5 | #include <DataTypes/DataTypeNothing.h> |
| 6 | #include <Interpreters/Context.h> |
| 7 | #include <DataTypes/DataTypeNullable.h> |
| 8 | |
| 9 | namespace DB |
| 10 | { |
| 11 | |
| 12 | namespace ErrorCodes |
| 13 | { |
| 14 | extern const int ATTEMPT_TO_READ_AFTER_EOF; |
| 15 | extern const int CANNOT_READ_ALL_DATA; |
| 16 | extern const int CANNOT_PARSE_ESCAPE_SEQUENCE; |
| 17 | extern const int CANNOT_PARSE_QUOTED_STRING; |
| 18 | extern const int SYNTAX_ERROR; |
| 19 | } |
| 20 | |
| 21 | |
| 22 | TemplateRowInputFormat::TemplateRowInputFormat(const Block & , ReadBuffer & in_, const Params & params_, |
| 23 | FormatSettings settings_, bool ignore_spaces_, |
| 24 | ParsedTemplateFormatString format_, ParsedTemplateFormatString row_format_, |
| 25 | std::string row_between_delimiter_) |
| 26 | : RowInputFormatWithDiagnosticInfo(header_, buf, params_), buf(in_), data_types(header_.getDataTypes()), |
| 27 | settings(std::move(settings_)), ignore_spaces(ignore_spaces_), |
| 28 | format(std::move(format_)), row_format(std::move(row_format_)), |
| 29 | default_csv_delimiter(settings.csv.delimiter), row_between_delimiter(std::move(row_between_delimiter_)) |
| 30 | { |
| 31 | /// Validate format string for result set |
| 32 | bool has_data = false; |
| 33 | for (size_t i = 0; i < format.columnsCount(); ++i) |
| 34 | { |
| 35 | if (format.format_idx_to_column_idx[i]) |
| 36 | { |
| 37 | if (*format.format_idx_to_column_idx[i] != 0) |
| 38 | format.throwInvalidFormat("Invalid input part" , i); |
| 39 | if (has_data) |
| 40 | format.throwInvalidFormat("${data} can occur only once" , i); |
| 41 | if (format.formats[i] != ColumnFormat::None) |
| 42 | format.throwInvalidFormat("${data} must have empty or None deserialization type" , i); |
| 43 | has_data = true; |
| 44 | format_data_idx = i; |
| 45 | } |
| 46 | else |
| 47 | { |
| 48 | if (format.formats[i] == ColumnFormat::Xml || format.formats[i] == ColumnFormat::Raw) |
| 49 | format.throwInvalidFormat("XML and Raw deserialization is not supported" , i); |
| 50 | } |
| 51 | } |
| 52 | |
| 53 | /// Validate format string for rows |
| 54 | std::vector<UInt8> column_in_format(header_.columns(), false); |
| 55 | for (size_t i = 0; i < row_format.columnsCount(); ++i) |
| 56 | { |
| 57 | if (row_format.formats[i] == ColumnFormat::Xml || row_format.formats[i] == ColumnFormat::Raw) |
| 58 | row_format.throwInvalidFormat("XML and Raw deserialization is not supported" , i); |
| 59 | |
| 60 | if (row_format.format_idx_to_column_idx[i]) |
| 61 | { |
| 62 | if (header_.columns() <= *row_format.format_idx_to_column_idx[i]) |
| 63 | row_format.throwInvalidFormat("Column index " + std::to_string(*row_format.format_idx_to_column_idx[i]) + |
| 64 | " must be less then number of columns (" + std::to_string(header_.columns()) + ")" , i); |
| 65 | if (row_format.formats[i] == ColumnFormat::None) |
| 66 | row_format.throwInvalidFormat("Column is not skipped, but deserialization type is None" , i); |
| 67 | |
| 68 | size_t col_idx = *row_format.format_idx_to_column_idx[i]; |
| 69 | if (column_in_format[col_idx]) |
| 70 | row_format.throwInvalidFormat("Duplicate column" , i); |
| 71 | column_in_format[col_idx] = true; |
| 72 | } |
| 73 | } |
| 74 | |
| 75 | for (size_t i = 0; i < header_.columns(); ++i) |
| 76 | if (!column_in_format[i]) |
| 77 | always_default_columns.push_back(i); |
| 78 | } |
| 79 | |
| 80 | void TemplateRowInputFormat::readPrefix() |
| 81 | { |
| 82 | size_t last_successfully_parsed_idx = 0; |
| 83 | try |
| 84 | { |
| 85 | tryReadPrefixOrSuffix<void>(last_successfully_parsed_idx, format_data_idx); |
| 86 | } |
| 87 | catch (Exception & e) |
| 88 | { |
| 89 | format.throwInvalidFormat(e.message() + " While parsing prefix" , last_successfully_parsed_idx); |
| 90 | } |
| 91 | } |
| 92 | |
| 93 | /// Asserts delimiters and skips fields in prefix or suffix. |
| 94 | /// tryReadPrefixOrSuffix<bool>(...) is used in checkForSuffix() to avoid throwing an exception after read of each row |
| 95 | /// (most likely false will be returned on first call of checkString(...)) |
| 96 | template <typename ReturnType> |
| 97 | ReturnType TemplateRowInputFormat::tryReadPrefixOrSuffix(size_t & input_part_beg, size_t input_part_end) |
| 98 | { |
| 99 | static constexpr bool throw_exception = std::is_same_v<ReturnType, void>; |
| 100 | |
| 101 | skipSpaces(); |
| 102 | if constexpr (throw_exception) |
| 103 | assertString(format.delimiters[input_part_beg], buf); |
| 104 | else |
| 105 | { |
| 106 | if (likely(!checkString(format.delimiters[input_part_beg], buf))) |
| 107 | return ReturnType(false); |
| 108 | } |
| 109 | |
| 110 | while (input_part_beg < input_part_end) |
| 111 | { |
| 112 | skipSpaces(); |
| 113 | if constexpr (throw_exception) |
| 114 | skipField(format.formats[input_part_beg]); |
| 115 | else |
| 116 | { |
| 117 | try |
| 118 | { |
| 119 | skipField(format.formats[input_part_beg]); |
| 120 | } |
| 121 | catch (const Exception & e) |
| 122 | { |
| 123 | if (e.code() != ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF && |
| 124 | e.code() != ErrorCodes::CANNOT_PARSE_ESCAPE_SEQUENCE && |
| 125 | e.code() != ErrorCodes::CANNOT_PARSE_QUOTED_STRING) |
| 126 | throw; |
| 127 | /// If it's parsing error, then suffix is not found |
| 128 | return ReturnType(false); |
| 129 | } |
| 130 | } |
| 131 | ++input_part_beg; |
| 132 | |
| 133 | skipSpaces(); |
| 134 | if constexpr (throw_exception) |
| 135 | assertString(format.delimiters[input_part_beg], buf); |
| 136 | else |
| 137 | { |
| 138 | if (likely(!checkString(format.delimiters[input_part_beg], buf))) |
| 139 | return ReturnType(false); |
| 140 | } |
| 141 | } |
| 142 | |
| 143 | if constexpr (!throw_exception) |
| 144 | return ReturnType(true); |
| 145 | } |
| 146 | |
| 147 | bool TemplateRowInputFormat::readRow(MutableColumns & columns, RowReadExtension & ) |
| 148 | { |
| 149 | /// This function can be called again after it returned false |
| 150 | if (unlikely(end_of_stream)) |
| 151 | return false; |
| 152 | |
| 153 | skipSpaces(); |
| 154 | |
| 155 | if (unlikely(checkForSuffix())) |
| 156 | { |
| 157 | end_of_stream = true; |
| 158 | return false; |
| 159 | } |
| 160 | |
| 161 | updateDiagnosticInfo(); |
| 162 | |
| 163 | if (likely(row_num != 1)) |
| 164 | assertString(row_between_delimiter, buf); |
| 165 | |
| 166 | extra.read_columns.assign(columns.size(), false); |
| 167 | |
| 168 | for (size_t i = 0; i < row_format.columnsCount(); ++i) |
| 169 | { |
| 170 | skipSpaces(); |
| 171 | assertString(row_format.delimiters[i], buf); |
| 172 | skipSpaces(); |
| 173 | if (row_format.format_idx_to_column_idx[i]) |
| 174 | { |
| 175 | size_t col_idx = *row_format.format_idx_to_column_idx[i]; |
| 176 | extra.read_columns[col_idx] = deserializeField(data_types[col_idx], *columns[col_idx], i); |
| 177 | } |
| 178 | else |
| 179 | skipField(row_format.formats[i]); |
| 180 | |
| 181 | } |
| 182 | |
| 183 | skipSpaces(); |
| 184 | assertString(row_format.delimiters.back(), buf); |
| 185 | |
| 186 | for (const auto & idx : always_default_columns) |
| 187 | data_types[idx]->insertDefaultInto(*columns[idx]); |
| 188 | |
| 189 | return true; |
| 190 | } |
| 191 | |
| 192 | bool TemplateRowInputFormat::deserializeField(const DataTypePtr & type, IColumn & column, size_t file_column) |
| 193 | { |
| 194 | ColumnFormat col_format = row_format.formats[file_column]; |
| 195 | bool read = true; |
| 196 | bool parse_as_nullable = settings.null_as_default && !type->isNullable(); |
| 197 | try |
| 198 | { |
| 199 | switch (col_format) |
| 200 | { |
| 201 | case ColumnFormat::Escaped: |
| 202 | if (parse_as_nullable) |
| 203 | read = DataTypeNullable::deserializeTextEscaped(column, buf, settings, type); |
| 204 | else |
| 205 | type->deserializeAsTextEscaped(column, buf, settings); |
| 206 | break; |
| 207 | case ColumnFormat::Quoted: |
| 208 | if (parse_as_nullable) |
| 209 | read = DataTypeNullable::deserializeTextQuoted(column, buf, settings, type); |
| 210 | else |
| 211 | type->deserializeAsTextQuoted(column, buf, settings); |
| 212 | break; |
| 213 | case ColumnFormat::Csv: |
| 214 | /// Will read unquoted string until settings.csv.delimiter |
| 215 | settings.csv.delimiter = row_format.delimiters[file_column + 1].empty() ? default_csv_delimiter : |
| 216 | row_format.delimiters[file_column + 1].front(); |
| 217 | if (parse_as_nullable) |
| 218 | read = DataTypeNullable::deserializeTextCSV(column, buf, settings, type); |
| 219 | else |
| 220 | type->deserializeAsTextCSV(column, buf, settings); |
| 221 | break; |
| 222 | case ColumnFormat::Json: |
| 223 | if (parse_as_nullable) |
| 224 | read = DataTypeNullable::deserializeTextJSON(column, buf, settings, type); |
| 225 | else |
| 226 | type->deserializeAsTextJSON(column, buf, settings); |
| 227 | break; |
| 228 | default: |
| 229 | __builtin_unreachable(); |
| 230 | } |
| 231 | } |
| 232 | catch (Exception & e) |
| 233 | { |
| 234 | if (e.code() == ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF) |
| 235 | throwUnexpectedEof(); |
| 236 | throw; |
| 237 | } |
| 238 | return read; |
| 239 | } |
| 240 | |
| 241 | void TemplateRowInputFormat::skipField(TemplateRowInputFormat::ColumnFormat col_format) |
| 242 | { |
| 243 | String tmp; |
| 244 | constexpr const char * field_name = "<SKIPPED COLUMN>" ; |
| 245 | constexpr size_t field_name_len = 16; |
| 246 | try |
| 247 | { |
| 248 | switch (col_format) |
| 249 | { |
| 250 | case ColumnFormat::None: |
| 251 | /// Empty field, just skip spaces |
| 252 | break; |
| 253 | case ColumnFormat::Escaped: |
| 254 | readEscapedString(tmp, buf); |
| 255 | break; |
| 256 | case ColumnFormat::Quoted: |
| 257 | readQuotedString(tmp, buf); |
| 258 | break; |
| 259 | case ColumnFormat::Csv: |
| 260 | readCSVString(tmp, buf, settings.csv); |
| 261 | break; |
| 262 | case ColumnFormat::Json: |
| 263 | skipJSONField(buf, StringRef(field_name, field_name_len)); |
| 264 | break; |
| 265 | default: |
| 266 | __builtin_unreachable(); |
| 267 | } |
| 268 | } |
| 269 | catch (Exception & e) |
| 270 | { |
| 271 | if (e.code() == ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF) |
| 272 | throwUnexpectedEof(); |
| 273 | throw; |
| 274 | } |
| 275 | } |
| 276 | |
| 277 | /// Returns true if all rows have been read i.e. there are only suffix and spaces (if ignore_spaces == true) before EOF. |
| 278 | /// Otherwise returns false |
| 279 | bool TemplateRowInputFormat::checkForSuffix() |
| 280 | { |
| 281 | PeekableReadBufferCheckpoint checkpoint{buf}; |
| 282 | bool suffix_found = false; |
| 283 | size_t last_successfully_parsed_idx = format_data_idx + 1; |
| 284 | try |
| 285 | { |
| 286 | suffix_found = tryReadPrefixOrSuffix<bool>(last_successfully_parsed_idx, format.columnsCount()); |
| 287 | } |
| 288 | catch (const Exception & e) |
| 289 | { |
| 290 | if (e.code() != ErrorCodes::ATTEMPT_TO_READ_AFTER_EOF && |
| 291 | e.code() != ErrorCodes::CANNOT_PARSE_ESCAPE_SEQUENCE && |
| 292 | e.code() != ErrorCodes::CANNOT_PARSE_QUOTED_STRING) |
| 293 | throw; |
| 294 | } |
| 295 | |
| 296 | if (unlikely(suffix_found)) |
| 297 | { |
| 298 | skipSpaces(); |
| 299 | if (buf.eof()) |
| 300 | return true; |
| 301 | } |
| 302 | |
| 303 | buf.rollbackToCheckpoint(); |
| 304 | return false; |
| 305 | } |
| 306 | |
| 307 | bool TemplateRowInputFormat::parseRowAndPrintDiagnosticInfo(MutableColumns & columns, WriteBuffer & out) |
| 308 | { |
| 309 | out << "Suffix does not match: " ; |
| 310 | size_t last_successfully_parsed_idx = format_data_idx + 1; |
| 311 | const ReadBuffer::Position row_begin_pos = buf.position(); |
| 312 | bool caught = false; |
| 313 | try |
| 314 | { |
| 315 | PeekableReadBufferCheckpoint checkpoint{buf, true}; |
| 316 | tryReadPrefixOrSuffix<void>(last_successfully_parsed_idx, format.columnsCount()); |
| 317 | } |
| 318 | catch (Exception & e) |
| 319 | { |
| 320 | out << e.message() << " Near column " << last_successfully_parsed_idx; |
| 321 | caught = true; |
| 322 | } |
| 323 | if (!caught) |
| 324 | { |
| 325 | out << " There is some data after suffix (EOF expected, got " ; |
| 326 | verbosePrintString(buf.position(), std::min(buf.buffer().end(), buf.position() + 16), out); |
| 327 | out << "). " ; |
| 328 | } |
| 329 | out << " Format string (from format_schema): \n" << format.dump() << "\n" ; |
| 330 | |
| 331 | if (row_begin_pos != buf.position()) |
| 332 | { |
| 333 | /// Pointers to buffer memory were invalidated during checking for suffix |
| 334 | out << "\nCannot print more diagnostic info." ; |
| 335 | return false; |
| 336 | } |
| 337 | |
| 338 | out << "\nUsing format string (from format_schema_rows): " << row_format.dump() << "\n" ; |
| 339 | out << "\nTrying to parse next row, because suffix does not match:\n" ; |
| 340 | try |
| 341 | { |
| 342 | if (likely(row_num != 1)) |
| 343 | assertString(row_between_delimiter, buf); |
| 344 | } |
| 345 | catch (const DB::Exception &) |
| 346 | { |
| 347 | writeErrorStringForWrongDelimiter(out, "delimiter between rows" , row_between_delimiter); |
| 348 | |
| 349 | return false; |
| 350 | } |
| 351 | for (size_t i = 0; i < row_format.columnsCount(); ++i) |
| 352 | { |
| 353 | skipSpaces(); |
| 354 | try |
| 355 | { |
| 356 | assertString(row_format.delimiters[i], buf); |
| 357 | } |
| 358 | catch (const DB::Exception &) |
| 359 | { |
| 360 | writeErrorStringForWrongDelimiter(out, "delimiter before field " + std::to_string(i), row_format.delimiters[i]); |
| 361 | return false; |
| 362 | } |
| 363 | |
| 364 | skipSpaces(); |
| 365 | if (row_format.format_idx_to_column_idx[i]) |
| 366 | { |
| 367 | auto & = getPort().getHeader(); |
| 368 | size_t col_idx = *row_format.format_idx_to_column_idx[i]; |
| 369 | if (!deserializeFieldAndPrintDiagnosticInfo(header.getByPosition(col_idx).name, data_types[col_idx], |
| 370 | *columns[col_idx], out, i)) |
| 371 | { |
| 372 | out << "Maybe it's not possible to deserialize field " + std::to_string(i) + |
| 373 | " as " + ParsedTemplateFormatString::formatToString(row_format.formats[i]); |
| 374 | return false; |
| 375 | } |
| 376 | } |
| 377 | else |
| 378 | { |
| 379 | static const String skipped_column_str = "<SKIPPED COLUMN>" ; |
| 380 | static const DataTypePtr skipped_column_type = std::make_shared<DataTypeNothing>(); |
| 381 | static const MutableColumnPtr skipped_column = skipped_column_type->createColumn(); |
| 382 | if (!deserializeFieldAndPrintDiagnosticInfo(skipped_column_str, skipped_column_type, *skipped_column, out, i)) |
| 383 | return false; |
| 384 | } |
| 385 | } |
| 386 | |
| 387 | skipSpaces(); |
| 388 | try |
| 389 | { |
| 390 | assertString(row_format.delimiters.back(), buf); |
| 391 | } |
| 392 | catch (const DB::Exception &) |
| 393 | { |
| 394 | writeErrorStringForWrongDelimiter(out, "delimiter after last field" , row_format.delimiters.back()); |
| 395 | return false; |
| 396 | } |
| 397 | |
| 398 | return true; |
| 399 | } |
| 400 | |
| 401 | void TemplateRowInputFormat::writeErrorStringForWrongDelimiter(WriteBuffer & out, const String & description, const String & delim) |
| 402 | { |
| 403 | out << "ERROR: There is no " << description << ": expected " ; |
| 404 | verbosePrintString(delim.data(), delim.data() + delim.size(), out); |
| 405 | out << ", got " ; |
| 406 | if (buf.eof()) |
| 407 | out << "<End of stream>" ; |
| 408 | else |
| 409 | verbosePrintString(buf.position(), std::min(buf.position() + delim.size() + 10, buf.buffer().end()), out); |
| 410 | out << '\n'; |
| 411 | } |
| 412 | |
| 413 | void TemplateRowInputFormat::tryDeserializeFiled(const DataTypePtr & type, IColumn & column, size_t file_column, |
| 414 | ReadBuffer::Position & prev_pos, ReadBuffer::Position & curr_pos) |
| 415 | { |
| 416 | prev_pos = buf.position(); |
| 417 | if (row_format.format_idx_to_column_idx[file_column]) |
| 418 | deserializeField(type, column, file_column); |
| 419 | else |
| 420 | skipField(row_format.formats[file_column]); |
| 421 | curr_pos = buf.position(); |
| 422 | } |
| 423 | |
| 424 | bool TemplateRowInputFormat::isGarbageAfterField(size_t, ReadBuffer::Position) |
| 425 | { |
| 426 | /// Garbage will be considered as wrong delimiter |
| 427 | return false; |
| 428 | } |
| 429 | |
| 430 | bool TemplateRowInputFormat::allowSyncAfterError() const |
| 431 | { |
| 432 | return !row_format.delimiters.back().empty() || !row_between_delimiter.empty(); |
| 433 | } |
| 434 | |
| 435 | void TemplateRowInputFormat::syncAfterError() |
| 436 | { |
| 437 | bool at_beginning_of_row_or_eof = false; |
| 438 | while (!at_beginning_of_row_or_eof) |
| 439 | { |
| 440 | skipToNextDelimiterOrEof(row_format.delimiters.back()); |
| 441 | if (buf.eof()) |
| 442 | { |
| 443 | end_of_stream = true; |
| 444 | return; |
| 445 | } |
| 446 | buf.ignore(row_format.delimiters.back().size()); |
| 447 | |
| 448 | skipSpaces(); |
| 449 | if (checkForSuffix()) |
| 450 | return; |
| 451 | |
| 452 | bool last_delimiter_in_row_found = !row_format.delimiters.back().empty(); |
| 453 | |
| 454 | if (last_delimiter_in_row_found && checkString(row_between_delimiter, buf)) |
| 455 | at_beginning_of_row_or_eof = true; |
| 456 | else |
| 457 | skipToNextDelimiterOrEof(row_between_delimiter); |
| 458 | |
| 459 | if (buf.eof()) |
| 460 | at_beginning_of_row_or_eof = end_of_stream = true; |
| 461 | } |
| 462 | /// It can happen that buf.position() is not at the beginning of row |
| 463 | /// if some delimiters is similar to row_format.delimiters.back() and row_between_delimiter. |
| 464 | /// It will cause another parsing error. |
| 465 | } |
| 466 | |
| 467 | /// Searches for delimiter in input stream and sets buffer position to the beginning of delimiter (if found) or EOF (if not) |
| 468 | void TemplateRowInputFormat::skipToNextDelimiterOrEof(const String & delimiter) |
| 469 | { |
| 470 | if (delimiter.empty()) |
| 471 | return; |
| 472 | |
| 473 | while (!buf.eof()) |
| 474 | { |
| 475 | void * pos = memchr(buf.position(), delimiter[0], buf.available()); |
| 476 | if (!pos) |
| 477 | { |
| 478 | buf.position() += buf.available(); |
| 479 | continue; |
| 480 | } |
| 481 | |
| 482 | buf.position() = static_cast<ReadBuffer::Position>(pos); |
| 483 | |
| 484 | PeekableReadBufferCheckpoint checkpoint{buf}; |
| 485 | if (checkString(delimiter, buf)) |
| 486 | return; |
| 487 | |
| 488 | buf.rollbackToCheckpoint(); |
| 489 | ++buf.position(); |
| 490 | } |
| 491 | } |
| 492 | |
| 493 | void TemplateRowInputFormat::throwUnexpectedEof() |
| 494 | { |
| 495 | throw Exception("Unexpected EOF while parsing row " + std::to_string(row_num) + ". " |
| 496 | "Maybe last row has wrong format or input doesn't contain specified suffix before EOF." , |
| 497 | ErrorCodes::CANNOT_READ_ALL_DATA); |
| 498 | } |
| 499 | |
| 500 | void TemplateRowInputFormat::resetParser() |
| 501 | { |
| 502 | RowInputFormatWithDiagnosticInfo::resetParser(); |
| 503 | end_of_stream = false; |
| 504 | } |
| 505 | |
| 506 | void registerInputFormatProcessorTemplate(FormatFactory & factory) |
| 507 | { |
| 508 | for (bool ignore_spaces : {false, true}) |
| 509 | { |
| 510 | factory.registerInputFormatProcessor(ignore_spaces ? "TemplateIgnoreSpaces" : "Template" , [=]( |
| 511 | ReadBuffer & buf, |
| 512 | const Block & sample, |
| 513 | IRowInputFormat::Params params, |
| 514 | const FormatSettings & settings) |
| 515 | { |
| 516 | ParsedTemplateFormatString resultset_format; |
| 517 | if (settings.template_settings.resultset_format.empty()) |
| 518 | { |
| 519 | /// Default format string: "${data}" |
| 520 | resultset_format.delimiters.resize(2); |
| 521 | resultset_format.formats.emplace_back(ParsedTemplateFormatString::ColumnFormat::None); |
| 522 | resultset_format.format_idx_to_column_idx.emplace_back(0); |
| 523 | resultset_format.column_names.emplace_back("data" ); |
| 524 | } |
| 525 | else |
| 526 | { |
| 527 | /// Read format string from file |
| 528 | resultset_format = ParsedTemplateFormatString( |
| 529 | FormatSchemaInfo(settings.template_settings.resultset_format, "Template" , false, |
| 530 | settings.schema.is_server, settings.schema.format_schema_path), |
| 531 | [&](const String & partName) -> std::optional<size_t> |
| 532 | { |
| 533 | if (partName == "data" ) |
| 534 | return 0; |
| 535 | throw Exception("Unknown input part " + partName, |
| 536 | ErrorCodes::SYNTAX_ERROR); |
| 537 | }); |
| 538 | } |
| 539 | |
| 540 | ParsedTemplateFormatString row_format = ParsedTemplateFormatString( |
| 541 | FormatSchemaInfo(settings.template_settings.row_format, "Template" , false, |
| 542 | settings.schema.is_server, settings.schema.format_schema_path), |
| 543 | [&](const String & colName) -> std::optional<size_t> |
| 544 | { |
| 545 | return sample.getPositionByName(colName); |
| 546 | }); |
| 547 | |
| 548 | return std::make_shared<TemplateRowInputFormat>(sample, buf, params, settings, ignore_spaces, resultset_format, row_format, settings.template_settings.row_between_delimiter); |
| 549 | }); |
| 550 | } |
| 551 | |
| 552 | for (bool ignore_spaces : {false, true}) |
| 553 | { |
| 554 | factory.registerInputFormatProcessor(ignore_spaces ? "CustomSeparatedIgnoreSpaces" : "CustomSeparated" , [=]( |
| 555 | ReadBuffer & buf, |
| 556 | const Block & sample, |
| 557 | IRowInputFormat::Params params, |
| 558 | const FormatSettings & settings) |
| 559 | { |
| 560 | ParsedTemplateFormatString resultset_format = ParsedTemplateFormatString::setupCustomSeparatedResultsetFormat(settings.custom); |
| 561 | ParsedTemplateFormatString row_format = ParsedTemplateFormatString::setupCustomSeparatedRowFormat(settings.custom, sample); |
| 562 | |
| 563 | return std::make_shared<TemplateRowInputFormat>(sample, buf, params, settings, ignore_spaces, resultset_format, row_format, settings.custom.row_between_delimiter); |
| 564 | }); |
| 565 | } |
| 566 | } |
| 567 | |
| 568 | } |
| 569 | |