skipprd df

skipprd df prints every row of one namespace, or of all namespaces in a pipeline, as a single JSON document. It is SELECT * without writing SQL, and the CLI twin of Python Session.df(). Use it for a quick look at small datasets, in tests, or to hand data to another tool. It reads the same tables as skipprd query: live WAL data, combined with the Iceberg table when the destination is SkipprLake, AthenaIceberg, or DuckDB. For large tables, use skipprd query --plain with a LIMIT instead.

Skippr, your data department

Speak to an expert. Upfront insights in days, not months.

Book an intro call →

20 minutes · No sales pitch

Usage

Python

import skippr

session = skippr.Session(skippr.Config.discover().get_pipeline("bikehire"))
everything = session.df()           # all namespaces in the pipeline
rides = session.df("rides")         # one namespace
print(rides.num_rows)

CLI

skipprd df --pipeline <name> [--namespace <namespace>]
skipprd df --namespace <pipeline>.<namespace>

Options

Flag	Default	Description
-p, --pipeline <NAME>	PIPELINE_NAME env var	Pipeline to read. Required unless --namespace is written as <pipeline>.<namespace>.
--namespace <NAME>	All namespaces	One namespace, as rides (uses --pipeline) or bikehire.rides. Leave it out to read every namespace in the pipeline.
--config <PATH>	./skippr.yml	Config file to read.
--log [LEVEL]	Off	Print logs to stderr. Errors are only shown on screen with this flag.

See CLI overview for the other global flags.

The output has the same shape as skipprd query --plain: {"header": [...], "rows": [[...]]}, with every value as a string.

Examples

One namespace

skipprd df --pipeline bikehire --namespace rides

{"header":["ride_id","station","duration_s","member"],"rows":[["1","Waterloo","420","true"],["2","Euston","615","false"]]}

Fully qualified name

skipprd df --namespace bikehire.rides

Every namespace in a pipeline

skipprd df --pipeline bikehire

Rows from all namespaces are combined into one document. This only works when every namespace has the same columns; if they differ, the command fails, so read them one at a time with --namespace.

Pipe into another tool

skipprd df --namespace bikehire.rides | jq '.rows | length'

Troubleshooting

Symptom	Cause	Fix
Exits 1 with no output	The error is only printed with --log.	Rerun with --log.
df() requires a pipeline when listing all namespaces	No --pipeline, no PIPELINE_NAME, and no --namespace.	Pass --pipeline.
df("namespace") requires Session.pipeline	--namespace rides was given without a pipeline.	Pass --pipeline, or write --namespace bikehire.rides.
df name must be namespace or pipeline.namespace	The name has more than one ..	Use rides or bikehire.rides.
{"header":[],"rows":[]}	The pipeline has no namespaces or no data yet.	Run skipprd discover and skipprd sync first.

Next steps

• skipprd query
• Python