Skip to main content

PSV Format (Pipe-Separated Values)

Description​

PSV (Pipe-Separated Values) is a variant of CSV where fields are separated by pipe characters (|). It's useful when data contains commas that shouldn't be treated as delimiters. PSV extends the CSV implementation with a fixed pipe delimiter.

File Extensions​

  • .psv - Pipe-separated values files

Implementation Details​

Reading​

The PSV implementation:

  • Extends CSV implementation
  • Uses pipe (|) as delimiter
  • Supports all CSV features (encoding detection, quote handling)
  • Converts each row to a dictionary

Writing​

Writing support:

  • Uses pipe delimiter
  • Supports all CSV writing features
  • Requires field names specification

Key Features​

  • Pipe delimiter: Fixed pipe character separator
  • CSV compatibility: Inherits all CSV features
  • Encoding support: Automatic encoding detection
  • Quote handling: Properly handles quoted fields
  • Totals support: Can count total rows

Usage​

from iterable import open_iterable

# Basic reading
source = open_iterable('data.psv')
for row in source:
    print(row)
source.close()

# Writing
dest = open_iterable('output.psv', mode='w', iterableargs={
    'keys': ['id', 'name', 'value']
})
dest.write({'id': '1', 'name': 'John', 'value': 'test'})
dest.close()

Parameters​

  • keys (list[str]): Column names (required for writing)
  • encoding (str): File encoding (default: auto-detected)
  • quotechar (str): Quote character (default: ")

Limitations​

  1. Flat data only: Only supports tabular data
  2. Schema required for writing: Must specify field names
  3. No type preservation: All values are strings
  4. Pipe in data: If data contains pipe characters, they must be quoted

Compression Support​

PSV files can be compressed with all supported codecs:

  • GZip (.psv.gz)
  • BZip2 (.psv.bz2)
  • LZMA (.psv.xz)
  • LZ4 (.psv.lz4)
  • ZIP (.psv.zip)
  • Brotli (.psv.br)
  • ZStandard (.psv.zst)

Use Cases​

  • Data with commas: When data contains commas that shouldn't be delimiters
  • Unix pipelines: Compatible with Unix pipe operations
  • Database exports: Some databases export in PSV format
  • CSV - Comma-separated format
  • SSV - Semicolon-separated format
  • TSV - Tab-separated format (uses CSV with tab delimiter)