Importing and exporting

Test cases move in and out of Sekee as files, in the formats other test management tools read and write. Both buttons sit on Test Cases, next to Watch folder.

Export takes the selected folder and everything beneath it, or the whole project when All test cases is selected. Import offers that folder as the destination but lets you pick another in the dialog; with All test cases selected it offers the top level, and then every case in the file has to name a folder of its own.

An import never changes or removes anything that is already here. A folder whose name matches is reused rather than replaced, and a test case whose name is already in the destination folder is skipped, or brought in under a new name if you ask for that. There is no mode that overwrites.

Exporting

Press Export. The dialog lists the folder you are exporting, its subfolders and every test case in them, all ticked. Untick anything you do not want. The count at the bottom says how many cases the file will hold.

The file is named after what you exported, with the extension the format needs: exporting Checkout as a spreadsheet gives you checkout.csv, and as a PDF, checkout.pdf.

Format	Keeps the folder tree	Read by
CSV	As a Folder column	TestRail, Zephyr Scale, Xray, Qase, Testmo, Testiny, Tuskr, TestCollab, Azure DevOps
Excel	As a Folder column	qTest, PractiTest, and anything that opens a workbook
JSON	Yes	Sekee. Keeps the rich text the others flatten
TestRail XML	Yes	TestRail, Testiny, Qase, Testmo
TestLink XML	Yes	TestLink, qTest, Qase, QA Touch
PDF	Shown as a heading per case	People. Nothing can import it

A single test case can also be exported from its own screen, under Export — all six of the formats above, plus that screen's own Markdown export, which is the only export of any kind that carries the screenshots.

Choose JSON to move a library between two Sekee projects. The spreadsheet and XML formats flatten rich text to plain text, because a cell cannot hold formatting; JSON keeps it exactly. No format carries attachments, comments or run history — see what comes across.

The words in an exported file are spelled out for a reader who is not in Sekee: the automation status stored as can_be_done is written Can be automated. An import reads either spelling back.

What a spreadsheet export looks like

One row per step, which is the convention every other tool follows. The first row of a test case carries its fields and its first step; each further step is a row with the case columns left empty.

ID,Title,Folder,State,Automation,Estimate,Tags,Preconditions,Description,Scenario,Steps,Expected Result,Additional Data
86,Apply a promotional code,Checkout,In progress,Can be automated,5-10 minutes,payments,,A valid code reduces the total.,,Open the basket,Both items are listed,
,,,,,,,,,,Enter a valid code,The total drops,
,,,,,,,,,,Remove the code,The total returns to the subtotal,

Importing

Press Import and drop in a file. Sekee reads CSV, TSV, Excel, JSON, and TestRail or TestLink XML — up to 20 MB, and up to 10,000 test cases and steps counted together. A larger library imports in parts. There is no PDF import; a PDF is a picture of a table, and reading one back would produce test cases that quietly differ from what was exported.

There is no interchange standard for test cases, so nothing is assumed. What there is instead is a handful of conventions nearly every tool follows, and Sekee recognises them:

• The header row names the columns. Sekee matches them against the names other tools use — Section Hierarchy, Steps (Step), Test Script (Step-by-Step) - Expected Result, Action, Preconditions and several dozen more — so a file exported straight out of another tool usually maps itself.
• A test case usually spans several rows, one per step. A row with a title starts a new case; rows without one add a step to it.
• Some tools put every step in one cell instead, numbered or on separate lines.

Every one of those is a guess, and every guess is shown back to you as a control you can change before anything is written.

What the dialog asks

• Sheet — which sheet of a workbook, when it has more than one.
• Rows — one row per step, or one row per test case.
• Steps in a cell are separated by — numbering, new lines, or nothing. Only asked when a case is one row.
• Columns — what each column of the file is. Detected from the header, with a sample value under each name so you can check. A title is required; anything set to Don't import is left out.
• Import into — the destination folder. Folders named in the file are created underneath it. Defaults to the folder you were looking at.
• If a name is already there — skip it, or import it under a new name. A renamed case gets (imported) appended.

Under those, a summary says how many test cases, steps and folders are about to be created, and names the first few. Nothing is written until you press Import.

Folder paths

A folder column may hold a path — Checkout > Payments, Checkout/Payments and Checkout -> Payments all work. Missing folders are created; existing ones are reused, matched on name without regard to case.

The two XML formats carry the tree in their nesting instead, so no column is needed.

What comes across, and what does not

Names, folders, preconditions, descriptions, scenarios, steps with their expected results and additional data, tags, state, automation status and estimate. Values other tools spell differently are translated — a TestRail estimate of 420s becomes 5-10 minutes, an Azure DevOps state of Design becomes In progress.

Not imported: attachments and screenshots, comments, run history, custom step fields, requirement links, and the source tool's own ID. The Source ID column is recognised so you can see it was read, but nothing is stored under it — Sekee gives every imported case its own identity.

An imported step is a plain step. Nothing is linked to a shared step, even where the text matches one, because the file cannot say which steps its tool considered shared.

Everything imported is recorded as created by you, whatever the file says.

Who can do what

Anyone who can see the tests can export them; it takes a copy and changes nothing. Importing needs the same permission as creating a test case, so owners, admins and members can do it and viewers cannot.

People following any folder the tests landed in — or any folder above it, since a subscription covers the whole subtree — get one email saying how many tests arrived. Not one per test, which is what a three-hundred-case import from another tool would otherwise send.