Yuxuan Yan
f4b7ef3ae6
A standalone tools/pqcat.py (head/tail/columns/info) wired as `cli.py pqcat` and runnable directly via a PATH symlink, plus README docs. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
51 lines
1.7 KiB
Python
Executable File
51 lines
1.7 KiB
Python
Executable File
#!/usr/bin/env python3
|
|
"""CLI helper to inspect parquet files, like `cat` for `.pq`."""
|
|
|
|
import click
|
|
import pandas as pd
|
|
|
|
|
|
@click.command(name="pqcat")
|
|
@click.argument("path", type=click.Path(exists=True))
|
|
@click.option("-n", "--head", "head", type=int, default=None, help="Show only the first N rows")
|
|
@click.option("-t", "--tail", "tail", type=int, default=None, help="Show only the last N rows")
|
|
@click.option("-c", "--columns", default=None, help="Comma-separated subset of columns to show")
|
|
@click.option("--info", is_flag=True, help="Show shape + dtypes instead of the rows")
|
|
def pqcat(path, head, tail, columns, info):
|
|
"""Dump a parquet file (or partitioned dataset dir) to stdout.
|
|
|
|
PATH is a single .pq file or a Hive-partitioned dataset directory.
|
|
"""
|
|
df = pd.read_parquet(path)
|
|
|
|
if columns:
|
|
cols = [c.strip() for c in columns.split(",") if c.strip()]
|
|
missing = [c for c in cols if c not in df.columns]
|
|
if missing:
|
|
raise click.ClickException(f"Columns not found: {', '.join(missing)}")
|
|
df = df[cols]
|
|
|
|
if info:
|
|
click.echo(f"path: {path}")
|
|
click.echo(f"shape: {df.shape[0]:,} rows x {df.shape[1]} cols")
|
|
click.echo("dtypes:")
|
|
for name in df.columns:
|
|
click.echo(f" {name}: {df[name].dtype}")
|
|
return
|
|
|
|
if head is not None:
|
|
df = df.head(head)
|
|
elif tail is not None:
|
|
df = df.tail(tail)
|
|
|
|
with pd.option_context(
|
|
"display.max_rows", None if (head is None and tail is None) else len(df),
|
|
"display.max_columns", None,
|
|
"display.width", None,
|
|
):
|
|
click.echo(df.to_string(index=False))
|
|
|
|
|
|
if __name__ == "__main__":
|
|
pqcat()
|