zedda.collect_warnings()

zedda.collect_warnings(
    path,
    sample_size=None,
) -> list[dict]

Programmatic equivalent of warnings(). Returns a structured list of dicts instead of printing. This is the right function for CI/CD pipelines.

# Arguments

Argument	Type	Default	Description
path	str / Path / DataFrame	(required)	Input dataset
sample_size	int or None	None	Sample size in rows

# Returns

list[dict]. Each dict has these keys:

Key	Type	Description
icon	str	Severity icon (emoji-style, used by the terminal renderer)
column	str	Column name (or dataset-level label)
message	str	Human-readable description
category	str	Issue category (e.g. high_nulls, outliers)
severity	str	critical / warning / info
fix_code	str	Inline pandas fix snippet (may be empty)
fix_action	str	Short description of the fix action (may be empty)
auto_fixable	bool	Whether clean() would fix this automatically

# Example

import zedda as zd

issues = zd.collect_warnings("data.csv")

# Fail CI on any critical issue
critical = [i for i in issues if i["severity"] == "critical"]
assert not critical, f"{len(critical)} critical issues: {[i['message'] for i in critical]}"

# Count auto-fixable
auto_fixable = [i for i in issues if i["auto_fixable"]]
print(f"{len(auto_fixable)} of {len(issues)} issues are auto-fixable")

# CI/CD pattern

import sys
import zedda as zd

issues = zd.collect_warnings(sys.argv[1])
critical = [i for i in issues if i["severity"] == "critical"]

if critical:
    for i in critical:
        print(f"[CRITICAL] {i['column']}: {i['message']}")
    sys.exit(1)

# See also

• warnings() — the print-first variant.
• clean() — apply every auto-fixable warning.
• fix() — generate pandas code for every warning.