hardistest-cases:init

Description

Command Behavior

Writes the test cases of a ticket into a notebook a tester reviews and fills in: an Excel workbook, a CSV or a markdown table.

The test cases usually come from a NormalizedTestCase[] JSON file written by an AI agent that read the specification and the code. This command turns them into a file a human can read and correct, then hardistest-cases:upsert sends the corrected file to the test management tool.

One column set per kind of notebook. Functional notebooks have the steps and a helper SOQL query, technical notebooks have the Apex class and method under test, maintenance notebooks have no module nor priority. The kind comes from the ID column, or from --kind.
Tester columns: Actual result, Comment and Status (a value list: Passed, Failed, Blocked, N/A). They are written empty for new cases, and kept when an existing notebook is converted to another format.
A summary of the cases per module and priority, in its own sheet or under the CSV rows.
Readable back: every file written here is read by upsert, after a tester corrected it.
A blank notebook when no test case is given, with identifiers already filled.

Input

The format is read from the file extension:

Input	Flag	Typical producer
JSON test cases	`--testsjsonfile`	An AI agent that read the ticket and the code
An existing notebook (.md, .xlsx, .csv)	`--notebook`	A previous run of this command, or a notebook a tester filled in
Nothing	neither flag	Blank notebook

Output

Format: --format (xlsx by default, both writes the xlsx and the CSV), or the extension of --outputfile.
File: --outputfile, or hardis-report/test-cases-<TICKET>-<KIND>-<DATE>.<ext>. The file always gets the extension of its format: --outputfile notes.txt writes notes.xlsx.
A blank notebook holds at most 999 cases (modules x --rows), the most a test case id counter can number.
An existing file is never overwritten silently: the command asks first, and refuses in CI and in agent mode.

Configuration

None: this command reads a file and writes a file, without org, provider nor secret.

Technical explanations

JSON contract: id, title and expected are required. module, priority (1, 2 or 3), preconditions, target, soql and steps ([{ "action": "...", "expected": "..." }]) are optional. Every problem of an entry is reported at once, with its index.
Identifiers: <TICKET>-F01 functional, <TICKET>-T01 technical, <TICKET>-01 maintenance, with a 2 or 3 digit counter. Letters, digits, _, ., # and - only. The ticket and the kind are derived from the identifier, and an unreadable one is refused.
Steps cell: one numbered line per step, 1. Open the record → The record page is displayed. The first → or -> separates the action from its expected result, and a step without arrow has no expected result. An arrow inside an action is written =>, so it is not read as the separator. Steps are separated by a line break in the xlsx, and by <br> in the CSV and in markdown.
CSV: , delimiter, UTF-8 with BOM so Excel reads the accents, CRLF line endings. A cell starting with =, +, - or @ is prefixed with an apostrophe, so a spreadsheet does not run it as a formula. The reader detects , or ;.
Headers: the reader also accepts the French headers of earlier notebooks (Cas de test, Étapes, Résultat attendu...).

Agent Mode

Use --agent to disable all interactive prompts. Without test cases, --kind and --ticket-number are required. An existing output file makes the command fail instead of asking.

sf hardis:ticket:test-cases:init --testsjsonfile cases.json --agent

The same applies in CI.

Parameters

Name	Type	Description	Default	Options
agent	boolean	Run in non-interactive mode for agents and automation
debug -d	boolean	Activate debug mode (more logs)
flags-dir	option	Import flag values from a directory.
format	option	Output format: xlsx (default), csv, md, or both for xlsx and csv. Defaults to the extension of --outputfile		xlsx csv md both
json	boolean	Format output as json.
kind -k	option	Column set to write. Defaults to the kind derived from the ID column		functional technical maintenance
modules	option	Module names of a blank notebook, one group of rows each
notebook -n	option	Existing notebook to read the test cases from: .md, .xlsx or .csv
outputfile -f	option	Path of the generated notebook
rows	option	Rows per module of a blank notebook	3
testsjsonfile -j	option	NormalizedTestCase[] JSON file holding the test cases to write
ticket-number	option	Ticket key of a blank notebook. With test cases, sets their carrier ticket
websocket	option	Websocket host:port for VsCode SFDX Hardis UI integration

Examples

$ sf hardis:ticket:test-cases:init --testsjsonfile cases.json

$ sf hardis:ticket:test-cases:init --testsjsonfile cases.json --format both --outputfile ./PROJ-123.xlsx

$ sf hardis:ticket:test-cases:init --notebook docs/tests/PROJ-123.xlsx --outputfile docs/tests/PROJ-123.md

$ sf hardis:ticket:test-cases:init --kind functional --ticket-number PROJ-123 --modules Opportunity --rows 5 --agent

$ sf hardis:ticket:test-cases:init --testsjsonfile cases.json --agent