| // Licensed to the Apache Software Foundation (ASF) under one |
| // or more contributor license agreements. See the NOTICE file |
| // distributed with this work for additional information |
| // regarding copyright ownership. The ASF licenses this file |
| // to you under the Apache License, Version 2.0 (the |
| // "License"); you may not use this file except in compliance |
| // with the License. You may obtain a copy of the License at |
| // |
| // http://www.apache.org/licenses/LICENSE-2.0 |
| // |
| // Unless required by applicable law or agreed to in writing, |
| // software distributed under the License is distributed on an |
| // "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
| // KIND, either express or implied. See the License for the |
| // specific language governing permissions and limitations |
| // under the License. |
| |
| #ifndef PARQUET_COLUMN_READER_H |
| #define PARQUET_COLUMN_READER_H |
| |
| #include <algorithm> |
| #include <cstdint> |
| #include <cstring> |
| #include <iostream> |
| #include <memory> |
| #include <unordered_map> |
| #include <vector> |
| |
| #include <arrow/util/bit-util.h> |
| |
| #include "parquet/column_page.h" |
| #include "parquet/encoding.h" |
| #include "parquet/exception.h" |
| #include "parquet/schema.h" |
| #include "parquet/types.h" |
| #include "parquet/util/memory.h" |
| #include "parquet/util/visibility.h" |
| |
| namespace parquet { |
| |
| class BitReader; |
| class RleDecoder; |
| |
| class PARQUET_EXPORT LevelDecoder { |
| public: |
| LevelDecoder(); |
| ~LevelDecoder(); |
| |
| // Initialize the LevelDecoder state with new data |
| // and return the number of bytes consumed |
| int SetData(Encoding::type encoding, int16_t max_level, int num_buffered_values, |
| const uint8_t* data); |
| |
| // Decodes a batch of levels into an array and returns the number of levels decoded |
| int Decode(int batch_size, int16_t* levels); |
| |
| private: |
| int bit_width_; |
| int num_values_remaining_; |
| Encoding::type encoding_; |
| std::unique_ptr<RleDecoder> rle_decoder_; |
| std::unique_ptr<BitReader> bit_packed_decoder_; |
| }; |
| |
| class PARQUET_EXPORT ColumnReader { |
| public: |
| ColumnReader(const ColumnDescriptor*, std::unique_ptr<PageReader>, |
| ::arrow::MemoryPool* pool = ::arrow::default_memory_pool()); |
| virtual ~ColumnReader(); |
| |
| static std::shared_ptr<ColumnReader> Make(const ColumnDescriptor* descr, |
| std::unique_ptr<PageReader> pager, |
| ::arrow::MemoryPool* pool = ::arrow::default_memory_pool()); |
| |
| // Returns true if there are still values in this column. |
| bool HasNext() { |
| // Either there is no data page available yet, or the data page has been |
| // exhausted |
| if (num_buffered_values_ == 0 || num_decoded_values_ == num_buffered_values_) { |
| if (!ReadNewPage() || num_buffered_values_ == 0) { return false; } |
| } |
| return true; |
| } |
| |
| Type::type type() const { return descr_->physical_type(); } |
| |
| const ColumnDescriptor* descr() const { return descr_; } |
| |
| protected: |
| virtual bool ReadNewPage() = 0; |
| |
| // Read multiple definition levels into preallocated memory |
| // |
| // Returns the number of decoded definition levels |
| int64_t ReadDefinitionLevels(int64_t batch_size, int16_t* levels); |
| |
| // Read multiple repetition levels into preallocated memory |
| // Returns the number of decoded repetition levels |
| int64_t ReadRepetitionLevels(int64_t batch_size, int16_t* levels); |
| |
| const ColumnDescriptor* descr_; |
| |
| std::unique_ptr<PageReader> pager_; |
| std::shared_ptr<Page> current_page_; |
| |
| // Not set if full schema for this field has no optional or repeated elements |
| LevelDecoder definition_level_decoder_; |
| |
| // Not set for flat schemas. |
| LevelDecoder repetition_level_decoder_; |
| |
| // The total number of values stored in the data page. This is the maximum of |
| // the number of encoded definition levels or encoded values. For |
| // non-repeated, required columns, this is equal to the number of encoded |
| // values. For repeated or optional values, there may be fewer data values |
| // than levels, and this tells you how many encoded levels there are in that |
| // case. |
| int64_t num_buffered_values_; |
| |
| // The number of values from the current data page that have been decoded |
| // into memory |
| int64_t num_decoded_values_; |
| |
| ::arrow::MemoryPool* pool_; |
| }; |
| |
| // API to read values from a single column. This is the main client facing API. |
| template <typename DType> |
| class PARQUET_EXPORT TypedColumnReader : public ColumnReader { |
| public: |
| typedef typename DType::c_type T; |
| |
| TypedColumnReader(const ColumnDescriptor* schema, std::unique_ptr<PageReader> pager, |
| ::arrow::MemoryPool* pool = ::arrow::default_memory_pool()) |
| : ColumnReader(schema, std::move(pager), pool), current_decoder_(NULL) {} |
| virtual ~TypedColumnReader() {} |
| |
| // Read a batch of repetition levels, definition levels, and values from the |
| // column. |
| // |
| // Since null values are not stored in the values, the number of values read |
| // may be less than the number of repetition and definition levels. With |
| // nested data this is almost certainly true. |
| // |
| // Set def_levels or rep_levels to nullptr if you want to skip reading them. |
| // This is only safe if you know through some other source that there are no |
| // undefined values. |
| // |
| // To fully exhaust a row group, you must read batches until the number of |
| // values read reaches the number of stored values according to the metadata. |
| // |
| // This API is the same for both V1 and V2 of the DataPage |
| // |
| // @returns: actual number of levels read (see values_read for number of values read) |
| int64_t ReadBatch(int64_t batch_size, int16_t* def_levels, int16_t* rep_levels, |
| T* values, int64_t* values_read); |
| |
| /// Read a batch of repetition levels, definition levels, and values from the |
| /// column and leave spaces for null entries on the lowest level in the values |
| /// buffer. |
| /// |
| /// In comparision to ReadBatch the length of repetition and definition levels |
| /// is the same as of the number of values read for max_definition_level == 1. |
| /// In the case of max_definition_level > 1, the repetition and definition |
| /// levels are larger than the values but the values include the null entries |
| /// with definition_level == (max_definition_level - 1). |
| /// |
| /// To fully exhaust a row group, you must read batches until the number of |
| /// values read reaches the number of stored values according to the metadata. |
| /// |
| /// @param batch_size the number of levels to read |
| /// @param[out] def_levels The Parquet definition levels, output has |
| /// the length levels_read. |
| /// @param[out] rep_levels The Parquet repetition levels, output has |
| /// the length levels_read. |
| /// @param[out] values The values in the lowest nested level including |
| /// spacing for nulls on the lowest levels; output has the length |
| /// values_read. |
| /// @param[out] valid_bits Memory allocated for a bitmap that indicates if |
| /// the row is null or on the maximum definition level. For performance |
| /// reasons the underlying buffer should be able to store 1 bit more than |
| /// required. If this requires an additional byte, this byte is only read |
| /// but never written to. |
| /// @param valid_bits_offset The offset in bits of the valid_bits where the |
| /// first relevant bit resides. |
| /// @param[out] levels_read The number of repetition/definition levels that were read. |
| /// @param[out] values_read The number of values read, this includes all |
| /// non-null entries as well as all null-entries on the lowest level |
| /// (i.e. definition_level == max_definition_level - 1) |
| /// @param[out] null_count The number of nulls on the lowest levels. |
| /// (i.e. (values_read - null_count) is total number of non-null entries) |
| int64_t ReadBatchSpaced(int64_t batch_size, int16_t* def_levels, int16_t* rep_levels, |
| T* values, uint8_t* valid_bits, int64_t valid_bits_offset, int64_t* levels_read, |
| int64_t* values_read, int64_t* null_count); |
| |
| // Skip reading levels |
| // Returns the number of levels skipped |
| int64_t Skip(int64_t num_rows_to_skip); |
| |
| private: |
| typedef Decoder<DType> DecoderType; |
| |
| // Advance to the next data page |
| virtual bool ReadNewPage(); |
| |
| // Read up to batch_size values from the current data page into the |
| // pre-allocated memory T* |
| // |
| // @returns: the number of values read into the out buffer |
| int64_t ReadValues(int64_t batch_size, T* out); |
| |
| // Read up to batch_size values from the current data page into the |
| // pre-allocated memory T*, leaving spaces for null entries according |
| // to the def_levels. |
| // |
| // @returns: the number of values read into the out buffer |
| int64_t ReadValuesSpaced(int64_t batch_size, T* out, int null_count, |
| uint8_t* valid_bits, int64_t valid_bits_offset); |
| |
| // Map of encoding type to the respective decoder object. For example, a |
| // column chunk's data pages may include both dictionary-encoded and |
| // plain-encoded data. |
| std::unordered_map<int, std::shared_ptr<DecoderType>> decoders_; |
| |
| void ConfigureDictionary(const DictionaryPage* page); |
| |
| DecoderType* current_decoder_; |
| }; |
| |
| template <typename DType> |
| inline int64_t TypedColumnReader<DType>::ReadValues(int64_t batch_size, T* out) { |
| int64_t num_decoded = current_decoder_->Decode(out, static_cast<int>(batch_size)); |
| return num_decoded; |
| } |
| |
| template <typename DType> |
| inline int64_t TypedColumnReader<DType>::ReadValuesSpaced(int64_t batch_size, T* out, |
| int null_count, uint8_t* valid_bits, int64_t valid_bits_offset) { |
| return current_decoder_->DecodeSpaced( |
| out, static_cast<int>(batch_size), null_count, valid_bits, valid_bits_offset); |
| } |
| |
| template <typename DType> |
| inline int64_t TypedColumnReader<DType>::ReadBatch(int64_t batch_size, |
| int16_t* def_levels, int16_t* rep_levels, T* values, int64_t* values_read) { |
| // HasNext invokes ReadNewPage |
| if (!HasNext()) { |
| *values_read = 0; |
| return 0; |
| } |
| |
| // TODO(wesm): keep reading data pages until batch_size is reached, or the |
| // row group is finished |
| batch_size = std::min(batch_size, num_buffered_values_ - num_decoded_values_); |
| |
| int64_t num_def_levels = 0; |
| int64_t num_rep_levels = 0; |
| |
| int64_t values_to_read = 0; |
| |
| // If the field is required and non-repeated, there are no definition levels |
| if (descr_->max_definition_level() > 0 && def_levels) { |
| num_def_levels = ReadDefinitionLevels(batch_size, def_levels); |
| // TODO(wesm): this tallying of values-to-decode can be performed with better |
| // cache-efficiency if fused with the level decoding. |
| for (int64_t i = 0; i < num_def_levels; ++i) { |
| if (def_levels[i] == descr_->max_definition_level()) { ++values_to_read; } |
| } |
| } else { |
| // Required field, read all values |
| values_to_read = batch_size; |
| } |
| |
| // Not present for non-repeated fields |
| if (descr_->max_repetition_level() > 0 && rep_levels) { |
| num_rep_levels = ReadRepetitionLevels(batch_size, rep_levels); |
| if (def_levels && num_def_levels != num_rep_levels) { |
| throw ParquetException("Number of decoded rep / def levels did not match"); |
| } |
| } |
| |
| *values_read = ReadValues(values_to_read, values); |
| int64_t total_values = std::max(num_def_levels, *values_read); |
| num_decoded_values_ += total_values; |
| |
| return total_values; |
| } |
| |
| inline void DefinitionLevelsToBitmap(const int16_t* def_levels, int64_t num_def_levels, |
| int16_t max_definition_level, int16_t max_repetition_level, int64_t* values_read, |
| int64_t* null_count, uint8_t* valid_bits, int64_t valid_bits_offset) { |
| int byte_offset = static_cast<int>(valid_bits_offset) / 8; |
| int bit_offset = static_cast<int>(valid_bits_offset) % 8; |
| uint8_t bitset = valid_bits[byte_offset]; |
| |
| // TODO(itaiin): As an interim solution we are splitting the code path here |
| // between repeated+flat column reads, and non-repeated+nested reads. |
| // Those paths need to be merged in the future |
| for (int i = 0; i < num_def_levels; ++i) { |
| if (def_levels[i] == max_definition_level) { |
| bitset |= (1 << bit_offset); |
| } else if (max_repetition_level > 0) { |
| // repetition+flat case |
| if (def_levels[i] == (max_definition_level - 1)) { |
| bitset &= ~(1 << bit_offset); |
| *null_count += 1; |
| } else { |
| continue; |
| } |
| } else { |
| // non-repeated+nested case |
| if (def_levels[i] < max_definition_level) { |
| bitset &= ~(1 << bit_offset); |
| *null_count += 1; |
| } else { |
| throw ParquetException("definition level exceeds maximum"); |
| } |
| } |
| |
| bit_offset++; |
| if (bit_offset == 8) { |
| bit_offset = 0; |
| valid_bits[byte_offset] = bitset; |
| byte_offset++; |
| // TODO: Except for the last byte, this shouldn't be needed |
| bitset = valid_bits[byte_offset]; |
| } |
| } |
| if (bit_offset != 0) { valid_bits[byte_offset] = bitset; } |
| *values_read = (bit_offset + byte_offset * 8 - valid_bits_offset); |
| } |
| |
| template <typename DType> |
| inline int64_t TypedColumnReader<DType>::ReadBatchSpaced(int64_t batch_size, |
| int16_t* def_levels, int16_t* rep_levels, T* values, uint8_t* valid_bits, |
| int64_t valid_bits_offset, int64_t* levels_read, int64_t* values_read, |
| int64_t* null_count_out) { |
| // HasNext invokes ReadNewPage |
| if (!HasNext()) { |
| *levels_read = 0; |
| *values_read = 0; |
| *null_count_out = 0; |
| return 0; |
| } |
| |
| int64_t total_values; |
| // TODO(wesm): keep reading data pages until batch_size is reached, or the |
| // row group is finished |
| batch_size = std::min(batch_size, num_buffered_values_ - num_decoded_values_); |
| |
| // If the field is required and non-repeated, there are no definition levels |
| if (descr_->max_definition_level() > 0) { |
| int64_t num_def_levels = ReadDefinitionLevels(batch_size, def_levels); |
| |
| // Not present for non-repeated fields |
| if (descr_->max_repetition_level() > 0) { |
| int64_t num_rep_levels = ReadRepetitionLevels(batch_size, rep_levels); |
| if (num_def_levels != num_rep_levels) { |
| throw ParquetException("Number of decoded rep / def levels did not match"); |
| } |
| } |
| |
| // TODO(itaiin): another code path split to merge when the general case is done |
| bool has_spaced_values; |
| if (descr_->max_repetition_level() > 0) { |
| // repeated+flat case |
| has_spaced_values = !descr_->schema_node()->is_required(); |
| } else { |
| // non-repeated+nested case |
| // Find if a node forces nulls in the lowest level along the hierarchy |
| const schema::Node* node = descr_->schema_node().get(); |
| has_spaced_values = false; |
| while (node) { |
| auto parent = node->parent(); |
| if (node->is_optional()) { |
| has_spaced_values = true; |
| break; |
| } |
| node = parent; |
| } |
| } |
| |
| int64_t null_count = 0; |
| if (!has_spaced_values) { |
| int values_to_read = 0; |
| for (int64_t i = 0; i < num_def_levels; ++i) { |
| if (def_levels[i] == descr_->max_definition_level()) { ++values_to_read; } |
| } |
| total_values = ReadValues(values_to_read, values); |
| for (int64_t i = 0; i < total_values; i++) { |
| ::arrow::BitUtil::SetBit(valid_bits, valid_bits_offset + i); |
| } |
| *values_read = total_values; |
| } else { |
| int16_t max_definition_level = descr_->max_definition_level(); |
| int16_t max_repetition_level = descr_->max_repetition_level(); |
| DefinitionLevelsToBitmap(def_levels, num_def_levels, max_definition_level, |
| max_repetition_level, values_read, &null_count, valid_bits, valid_bits_offset); |
| total_values = ReadValuesSpaced(*values_read, values, static_cast<int>(null_count), |
| valid_bits, valid_bits_offset); |
| } |
| *levels_read = num_def_levels; |
| *null_count_out = null_count; |
| |
| } else { |
| // Required field, read all values |
| total_values = ReadValues(batch_size, values); |
| for (int64_t i = 0; i < total_values; i++) { |
| ::arrow::BitUtil::SetBit(valid_bits, valid_bits_offset + i); |
| } |
| *null_count_out = 0; |
| *levels_read = total_values; |
| } |
| |
| num_decoded_values_ += *levels_read; |
| return total_values; |
| } |
| |
| template <typename DType> |
| inline int64_t TypedColumnReader<DType>::Skip(int64_t num_rows_to_skip) { |
| int64_t rows_to_skip = num_rows_to_skip; |
| while (HasNext() && rows_to_skip > 0) { |
| // If the number of rows to skip is more than the number of undecoded values, skip the |
| // Page. |
| if (rows_to_skip > (num_buffered_values_ - num_decoded_values_)) { |
| rows_to_skip -= num_buffered_values_ - num_decoded_values_; |
| num_decoded_values_ = num_buffered_values_; |
| } else { |
| // We need to read this Page |
| // Jump to the right offset in the Page |
| int64_t batch_size = 1024; // ReadBatch with a smaller memory footprint |
| int64_t values_read = 0; |
| |
| std::shared_ptr<PoolBuffer> vals = AllocateBuffer( |
| this->pool_, batch_size * type_traits<DType::type_num>::value_byte_size); |
| std::shared_ptr<PoolBuffer> def_levels = |
| AllocateBuffer(this->pool_, batch_size * sizeof(int16_t)); |
| |
| std::shared_ptr<PoolBuffer> rep_levels = |
| AllocateBuffer(this->pool_, batch_size * sizeof(int16_t)); |
| |
| do { |
| batch_size = std::min(batch_size, rows_to_skip); |
| values_read = ReadBatch(static_cast<int>(batch_size), |
| reinterpret_cast<int16_t*>(def_levels->mutable_data()), |
| reinterpret_cast<int16_t*>(rep_levels->mutable_data()), |
| reinterpret_cast<T*>(vals->mutable_data()), &values_read); |
| rows_to_skip -= values_read; |
| } while (values_read > 0 && rows_to_skip > 0); |
| } |
| } |
| return num_rows_to_skip - rows_to_skip; |
| } |
| |
| typedef TypedColumnReader<BooleanType> BoolReader; |
| typedef TypedColumnReader<Int32Type> Int32Reader; |
| typedef TypedColumnReader<Int64Type> Int64Reader; |
| typedef TypedColumnReader<Int96Type> Int96Reader; |
| typedef TypedColumnReader<FloatType> FloatReader; |
| typedef TypedColumnReader<DoubleType> DoubleReader; |
| typedef TypedColumnReader<ByteArrayType> ByteArrayReader; |
| typedef TypedColumnReader<FLBAType> FixedLenByteArrayReader; |
| |
| extern template class PARQUET_EXPORT TypedColumnReader<BooleanType>; |
| extern template class PARQUET_EXPORT TypedColumnReader<Int32Type>; |
| extern template class PARQUET_EXPORT TypedColumnReader<Int64Type>; |
| extern template class PARQUET_EXPORT TypedColumnReader<Int96Type>; |
| extern template class PARQUET_EXPORT TypedColumnReader<FloatType>; |
| extern template class PARQUET_EXPORT TypedColumnReader<DoubleType>; |
| extern template class PARQUET_EXPORT TypedColumnReader<ByteArrayType>; |
| extern template class PARQUET_EXPORT TypedColumnReader<FLBAType>; |
| |
| } // namespace parquet |
| |
| #endif // PARQUET_COLUMN_READER_H |