NewBestSub/it.uniud.newbestsub.dataset/DatasetView

DatasetView

class DatasetView

DatasetView

Composite façade that fans out output operations to two independent backends:

CSVView — streaming‑first CSV (owns buffering + canonical sort/rewrite).
ParquetView — streaming‑first Parquet (keeps its own buffers + final write).

Public surface

print — final snapshot writer (CSV + Parquet).
appendCardinality — streaming append of per‑generation cardinalities (FUN/VAR).
replaceTopBatch — streaming update of top‑solutions blocks (TOP).
closeStreams — finalize and write sorted/merged artifacts for both backends.

Path helpers

CSV path helpers are delegated to CSVView and preserve the original naming scheme.
Parquet helpers are provided here for logging/UX convenience and mirror CSV layouts.
All merged artifacts are emitted under /CSV/ and /Parquet/ subfolders.

Constructors

DatasetView

constructor()

Functions

appendCardinality

fun appendCardinality(model: DatasetModel, event: CardinalityResult)

Stream one cardinality event into both CSV and Parquet backends.

closeStreams

fun closeStreams(model: DatasetModel)

Finalize both backends:

getAggregatedDataFilePath

fun getAggregatedDataFilePath(model: DatasetModel, isTargetAll: Boolean = false): String

getAggregatedDataMergedFilePath

fun getAggregatedDataMergedFilePath(model: DatasetModel, isTargetAll: Boolean = false): String

getAggregatedDataMergedParquetPath

fun getAggregatedDataMergedParquetPath(model: DatasetModel, isTargetAll: Boolean = false): String

getAggregatedDataParquetPath

fun getAggregatedDataParquetPath(model: DatasetModel, isTargetAll: Boolean = false): String

Mirrors the CSV FINAL table location under /Parquet/.

getFunctionValuesFilePath

fun getFunctionValuesFilePath(model: DatasetModel): String

getFunctionValuesMergedFilePath

fun getFunctionValuesMergedFilePath(model: DatasetModel): String

getFunctionValuesMergedParquetPath

fun getFunctionValuesMergedParquetPath(model: DatasetModel): String

getFunctionValuesParquetPath

fun getFunctionValuesParquetPath(model: DatasetModel): String

getInfoFilePath

fun getInfoFilePath(model: DatasetModel, isTargetAll: Boolean = false): String

getInfoMergedFilePath

fun getInfoMergedFilePath(model: DatasetModel, isTargetAll: Boolean = false): String

getInfoMergedParquetPath

fun getInfoMergedParquetPath(model: DatasetModel, isTargetAll: Boolean = false): String

getInfoParquetPath

fun getInfoParquetPath(model: DatasetModel, isTargetAll: Boolean = false): String

Mirrors the CSV INFO table location under /Parquet/.

getTopSolutionsFilePath

fun getTopSolutionsFilePath(model: DatasetModel): String

getTopSolutionsMergedFilePath

fun getTopSolutionsMergedFilePath(model: DatasetModel): String

getTopSolutionsMergedParquetPath

fun getTopSolutionsMergedParquetPath(model: DatasetModel): String

getTopSolutionsParquetPath

fun getTopSolutionsParquetPath(model: DatasetModel): String

getVariableValuesFilePath

fun getVariableValuesFilePath(model: DatasetModel): String

getVariableValuesMergedFilePath

fun getVariableValuesMergedFilePath(model: DatasetModel): String

getVariableValuesMergedParquetPath

fun getVariableValuesMergedParquetPath(model: DatasetModel): String

getVariableValuesParquetPath

fun getVariableValuesParquetPath(model: DatasetModel): String

fun print(runResult: Triple<List<BinarySolution>, List<BinarySolution>, Triple<String, String, Long>>, model: DatasetModel)

Write final snapshot artifacts for a completed run to both CSV and Parquet.

replaceTopBatch

fun replaceTopBatch(model: DatasetModel, blocks: Map<Int, List<String>>)

Replace a batch of top solutions blocks (per K) in both CSV and Parquet backends.

writeCsv

fun writeCsv(csvRows: List<Array<String>>, resultCsvPath: String)

Thin CSV passthrough.

writeParquet

fun writeParquet(rows: List<Array<String>>, resultParquetPath: String)

Generic Parquet table writer for Final tables.