Transition Table and Data Tools

Transition Table & Data Tools

Beyond the visual canvas, AutomataLab gives you a tabular editor and a set of "data‑out" tools aimed at coursework, papers, and grading.

The transition (δ) table

The δ side‑panel tab shows every transition grouped by source state and lets you edit the machine as a table — a fast alternative to the canvas:

• Add, delete, retarget, and relabel transitions inline (finite automata, PDAs, and single‑tape TMs).
• Click‑to‑locate — jump from any row to the matching state or edge on the canvas.
• The table reflects the active machine type's notation (symbols for FA, read, pop → push for PDA, read → write, dir for TM).

It's the quickest way to audit a machine for missing or duplicate moves, and it pairs naturally with the Validation checks.

Data export

Open Export… from the toolbar's ⋯ overflow menu. You can export:

Artifact	Formats	Use
Transition table	CSV, LaTeX	Drop the δ‑table straight into a spreadsheet or a paper/assignment.
Execution trace	CSV, JSON	The step‑by‑step run of the current input.
Computation tree	JSON	The full explored tree/trellis of a nondeterministic run.
Machine definition	JSON	The .autolab.json machine itself (see File Format).

The LaTeX transition table is ready to paste into a tabular environment; the CSV variants open cleanly in any spreadsheet tool.

Batch / test‑suite runner

Click Batch… (near the input bar) to run many inputs at once instead of typing them one by one:

• Enter one input per line. Optionally prefix a line with accept: or reject: to declare the expected verdict.
• The runner executes each input headlessly and shows a pass/fail table (actual verdict vs. expectation, with mismatches highlighted).
• Export the results as CSV for a grading record or a regression checklist.

This is ideal for verifying a machine against a language spec: paste a list of strings that should be accepted and a list that should be rejected, and confirm every row passes.

The batch runner and exporters use the same engine as the interactive simulator (via a headless run‑to‑completion path), so results always match what you'd see by stepping through manually.