Aller au contenu principal

OMOP Export

Category: Data
Node id: omop-export

Standards: OMOP-CDM-5.4, EHDS

Exports OMOP CDM tables to CSV (tab-separated) or Parquet files, one file per table

What it does​

  • OMOP CDM table export to CSV (tab-separated, OMOP convention)
  • OMOP CDM table export to Parquet with compression (snappy, gzip, zstd)
  • Dual-format export (CSV + Parquet simultaneously)
  • Per-table file generation with row counts and file sizes
  • Selective table export via include_tables filter
  • Path traversal security validation

Use cases​

  • Export OMOP CDM tables for data warehouse loading
  • Generate Parquet files for analytics platforms (Spark, BigQuery)
  • Create OMOP data snapshots for regulatory submissions
  • Selective table export for research datasets

Runtime​

  • Performance tier: medium

Input ports​

NameTypeRequiredDescription
omop_tablesjsonYesOMOP table map from fhir-to-omop node
datajsonNoAlternative input: OMOP table map

Output ports​

NameTypeRequiredDescription
resultjsonNoExport result containing files_written, table_stats, format, and output_directory
successbooleanNoWhether the export succeeded

Settings​

Basic​

SettingNameTypeRequiredDefaultAllowed valuesDescription
OperationoperationenumYesexport_csvexport_csv, export_parquet, export_allExport format
Output Directoryoutput_directorystringYesDirectory to write exported files
CompressioncompressionenumNozstdsnappy, gzip, zstd, noneCompression for Parquet files

Advanced​

SettingNameTypeRequiredDefaultAllowed valuesDescription
Include Tablesinclude_tablesstringNoComma-separated list of table names to export (empty exports all)
Base Directorybase_directorystringNoBase directory for path security validation

Errors​

CodeDescription
OUTPUT_DIR_REQUIREDOutput directory must be specified
PATH_TRAVERSALOutput path is outside allowed base directory
DIR_CREATE_FAILEDFailed to create output directory
EXPORT_FAILEDFailed to write table to file
UNSUPPORTED_OPERATIONOperation not recognized

Runtime notes​

  • CSV export uses tab-separated format (OMOP convention), not comma-separated
  • The node recursively searches nested input maps for OMOP CDM table keys (person, condition_occurrence, etc.)
  • export_all writes both CSV and Parquet for each table; files_written values become {csv, parquet} objects
  • Empty tables are silently skipped and do not appear in files_written or table_stats

See Build any workflow for how to put nodes together and Ports, types and data flow for how port types connect.

Quelque chose n'est pas clair ? Envoyez vos commentaires sur la documentation par courriel.