Release notes

Tabalyst is in beta. Interfaces and JSON formats may change between releases; update with pip install --upgrade tabalyst. Each JSON format has its own changelog: profile, scan and Inspect. The full release notes of each version are on GitHub.

0.6.1

Makes the choice of a table or collection easier, and reorganizes the documentation.

• When a JSON file or workbook has several equally plausible collections, report, scan and inspect print one command per choice, ready to copy: tabalyst report shop.xlsx --collection Costs. The short form of --collection is used whenever the name is plain.
• tabalyst report --all-collections reports every visible collection of a JSON file or workbook, as <name>.<collection>.html and .json. -d is optional for one file and required for several.
• An option that a kind of file cannot use, such as --delimiter on a workbook, is ignored with a warning.
• The report header says “Report” and the report links to the documentation.
• The documentation has a page per command in the command line reference and per area in the Python API reference, and the first pages moved under get-started/ (the old addresses redirect). No JSON format changed.

0.6.0

Adds Excel workbooks and relicenses the code under MPL-2.0.

• tabalyst inspect, scan and report read .xlsx and .xlsm workbooks. Inspect lists the sheets and named tables as candidates and selects one when it is clearly the largest; dataset_path ($.Sheet or $.Sheet.Table) and header_row choose the table.
• Profile format revision 11 and scan format revision 6: stored scans of an older revision are replaced by tabalyst scan and tabalyst report. The Inspect format gains the excel kind without a revision change.
• The code is now licensed under MPL-2.0; the documentation stays CC BY 4.0.

0.5.1

Chooses a JSON collection from report, and from inspect’s own output.

• tabalyst report gains --collection, as tabalyst scan has it. It outranks the Inspect file and the configuration, and cannot be used with --scan.
• --collection accepts a short form: data.items stands for $.data.items[], . for the root array $[], and a key made of digits, such as groups.121, is accepted as it is.
• When several collections are equally plausible, tabalyst inspect prints one ready-to-run tabalyst report command per collection.
• The report theme button cycles auto (the system preference, now the default), light and dark.

0.5.0

Adds Tabalyst Inspect and JSON Lines support.

• tabalyst inspect finds the collection of records of a .json, .jsonl or .ndjson file and writes an editable <source>-inspect.json. Its config sets the collection, the flattening of nested objects, the arrays mode and the error policy. Scan and Report apply it.
• When several collections are equally plausible, or none is usable, scan and report stop with exit code 2 and list the candidates.
• JSONL sources can be scanned and reported. A bad line is excluded, counted and located by default; errors.policy: strict stops at the first one.
• A stored scan is reused only when the full source content, the effective configuration and the Tabalyst version are unchanged.
• New Python functions tabalyst.inspect() and tabalyst.generate_inspections(). On Windows, Tabalyst expands wildcards itself.
• Compatibility. tabalyst scan without -o or -d no longer writes <stem>.scan.json beside a JSON source. A JSON file with a single object or a scalar root has no dataset. Formats: scan revision 5, profile revision 10, Inspect revision 1; documents of 0.4.4 must be generated again.

0.4.4

Restores fast default CSV scans. tabalyst scan stores a reusable scan.json without building a DuckDB database, and tabalyst report reuses it. Adds tabalyst report --details, one standalone HTML page per column.

0.4.3

First beta release. Adds tabalyst cache info and tabalyst cache clean for the disposable query caches of DuckDB projects.

0.4.2

Several times faster scans on large files: batched detection, and worker processes for files of 16 MiB or more (--workers).

0.4.1

Adaptive detection (scan.detection.warmup_values, scan.detection.probe_interval) for columns with many distinct values, and a report time that includes the scan.

0.4.0

Introduces Tabalyst Scan, a streaming analysis engine for CSV and JSON, and rebuilds Tabalyst Report on it: reports for JSON files, bounded memory, no pandas. Adds tabalyst scan, the detectors and the scan configuration section.

0.3.0

Introduces Tabalyst Sample: tabalyst sample with first, last, random and stratified methods, --rows, --percent and --seed.

0.2.0

Command-oriented toolkit: tabalyst report INPUT... [OPTIONS], automatic report names, several inputs and wildcards, -d/--output-dir, batch preflight checks, progress on standard error and the generate_reports() API.

0.1.x

First packaged releases: tabalyst.analyze(), the HTML report and the JSON profile, executions.json, and the refreshed report design of 0.1.2.