autors-cli

autors-cli is a CANoe-inspired terminal engineering workbench for the whole autors workspace. It uses Ratatui to place workspace discovery, network descriptions, traces, calibration data, diagnostics, simulation, and platform integration behind one keyboard-driven interface.

Current workbench surfaces

• The capability center discovers every autors-* package through Cargo metadata. Its descriptions, features, internal dependencies, paths, and engineering workflows therefore track workspace changes instead of being a fixed package list.
• Native A2L inspection summarizes projects, modules, measurements, characteristics, functions, conversions, layouts, and type definitions. A retained A2L session also drives a measurement/calibration workbench: module-qualified objects expose addresses, types, shapes, units, limits, read/write policy, physical values, editable virtual ECU memory, DAQ selection, and timestamped 10 Hz acquisition. Physical/raw conversion uses autors-values; DAQ packing and sample buffers use autors-comm. Captured physical samples are retained beyond the small recent-value display; w writes them through autors-mdf as a round-trippable MDF 3.30 recording with timestamps, module-qualified channels, units, limits, and descriptions.
• Native DBC inspection shows messages, transmitters, DLCs, signal details, units, scaling, and plausibility results.
• Native CANoe ASC inspection shows classic CAN, CAN FD, errors, events, timestamp metadata, and per-frame details.
• A DBC remains attached to the workbench session: opening an ASC or BLF trace afterwards adds symbolic message names and decoded physical signal values, including multiplex-branch selection and value-table labels.
• The CAN/CAN FD workbench provides manual transmit, a bounded live trace, bus-load statistics, DBC symbolic decoding, and remaining-bus simulation. Its always-available virtual adapter adds receive injection and loopback; optional Vector, Kvaser, and PEAK drivers run the same scheduler and trace pipeline on real channels. Scheduled messages can be enabled, triggered once, and edited for payload and cycle time while simulation is running. w exports the bounded live trace as a CANoe/CANalyzer-compatible ASC file.
• The LIN workbench loads LDF frame and schedule definitions, executes master/slave remaining-bus behavior (including header-only subscriber requests), switches schedule tables at runtime, edits and triggers frames, and decodes received application signals into the live trace. The same feature-gated Vector, Kvaser, and PEAK families can replace the virtual LIN channel without changing the scheduling workflow. w records full frames and header-only slave requests in a round-trippable PLIN-View LTRC file.
• The protocol lab builds and decodes UDS/KWP PDUs, wraps and validates DoIP diagnostic frames, and exposes typed CCP and XCP command/response codecs. Named command templates and raw hexadecimal input share one inspectable history. A separate live action sends UDS through the real ISO-TP state machine, CCP/XCP through native CAN request-response frames on the selected adapter, or UDS over a routed DoIP TCP client. Command/response frames remain visible in both protocol history and the CAN trace; the virtual CAN adapter supplies deterministic ECU responses for hardware-free workflow tests. The DoIP page broadcasts vehicle-identification requests, deduplicates every answering entity, shows VIN/EID/GID/action/synchronization data, and can use the selected responder's endpoint and logical address for routing activation.
• Native LDF and LTRC views cover LIN nodes, unconditional frames, signals, schedule inventory, trace frames, checksums, missing bytes, and bus events.
• Native BLF inspection inventories every typed/raw object and decodes common CAN, CAN FD, and LIN rows into trace columns.
• ASC, BLF, and LTRC tables have timestamp-driven play/pause, reset, and 0.125×-16× playback controls for offline trace review.
• Native MDF, CDF, and ODX views expose measurement channels, calibration instances, diagnostic variants, services, protocols, and DTC totals.
• Opening an ODX file also retains a database-driven diagnostic session (d). It merges inherited ECU/base variants, expands identifier-specific services, generates SID/subfunction/identifier request prefixes, lists dynamic request parameters, switches between ISO-TP/CAN and DoIP, and decodes live positive responses back into named physical values. ODX communication parameters can populate the physical CAN request/response IDs without retyping them.
• DCM, CANape PAR, and MATLAB calibration files gain an A2L-aware typed view after an A2L is retained. The workbench tries every module, selects the best object match, and shows scalar/array shape, physical preview, units, function ownership, axes, and explicit reasons for skipped values.
• Native PRM/CNF views expose flash-procedure command graphs, protocol mode, branches, controller addresses, transport settings, and source/erase/ destination memory segments. The retained PRM preflight workbench uses the shared autors-prm script interpreter to traverse command sets, procedures, state branches, and instruction semantics while explicitly suppressing waits, transport I/O, flashing, and Seed & Key calls.
• Native program-image views use content-first detection for Intel HEX, Motorola S-record, VBF, Ford I-HEX, TI-TXT, UF2, hexadecimal ASCII, mixed record streams, and raw binary files. They expose sparse address segments, holes, initialized byte counts, detection confidence, previews, and CRC-32.
• Native ELF and linker MAP views list sections and symbols with addresses, sizes, types, bindings, and the metadata needed for the shared A2L address synchronization workflow. After retaining an A2L, opening either source computes typed object matches in the symbol workbench (y): safe address-only changes are preselected, size mismatches require an explicit policy opt-in, bit-mask handling is visible, address multipliers can be recomputed, and s writes selected updates to an explicit A2L copy.
• Every other text or binary format can be opened in a safe universal line or hex inspector while its dedicated engineering view is being integrated.

Run

cargo run -p autors-cli
cargo run -p autors-cli -- network.dbc
cargo run -p autors-cli -- --workspace C:\work\auto_rs capture.asc
cargo run -p autors-cli --features hardware-all

Native drivers are opt-in and loaded dynamically at runtime. Use hardware-vector, hardware-kvaser, or hardware-peak for one adapter family, or hardware-all for all three. A missing DLL is reported in the TUI instead of preventing the normal virtual-only build from starting.

Use arrow keys or h/j/k/l to navigate. Press o to open a file, / to filter, Enter to open a selected crate's README, F5 to rescan the workspace, b to enter the CAN workbench, n to enter the LIN workbench, Esc to go back, and ? for the complete keyboard reference. Press a to enter the A2L measurement/calibration workbench after opening an A2L file; Tab changes the object view, Enter arms a measurement, Space runs DAQ, and e edits a writable physical value. Press w in this workbench to save the captured physical samples as MDF. In either bus workbench, use c to connect, s to transmit, i to inject a received frame, and Space to run or pause scheduling. Use a to cycle through compiled adapters, e to enumerate CAN channels when the selected backend exposes discovery, , and . to select a channel, and v to edit the zero-based channel and vendor hardware-type number directly. Receive injection is intentionally restricted to virtual adapters. The message table also supports Enter (enable), t (one-shot trigger), p (payload), and m (CAN cycle time). Use [ and ] to switch LDF schedule tables.

Press g for the protocol lab. Use Tab or Shift-Tab (or keys 1-4) to select UDS/KWP, DoIP, CCP, or XCP, then press e to enter a named template or raw wire bytes. The left panel documents the commands accepted by the selected codec. Press c to configure CAN IDs/FD mode or a DoIP endpoint and r to execute the template on that live transport. CAN protocol execution uses the connection selected in the CAN workbench. On DoIP, d discovers entities over UDP and [/] selects the endpoint used by the next routing activation.

After opening an ODX file, press d for the diagnostic database workbench. Tab selects an ECU/base variant, arrow keys select an expanded service row, t switches CAN/DoIP, and c applies resolvable ODX physical CAN IDs. e encodes the selected service for offline inspection; r executes it live. Both prompts accept optional raw bytes for dynamic request parameters after the ODX-generated fixed prefix.

After opening a PRM file, press f for the flash preflight workbench. Tab selects the entry command set and Space/Enter runs the no-I/O interpreter; the report shows each executed semantic step, suppressed I/O step, state, and branch target.

For address synchronization, open the A2L and then its ELF/AXF or linker MAP, and press y. Enter selects individual changes, m changes the address multiplier, z explicitly allows size-mismatch updates, p controls bit-mask preservation, and s writes a separate synchronized A2L target.

Development

cargo test -p autors-cli
cargo test -p autors-cli --features hardware-all
cargo clippy -p autors-cli --all-targets --no-deps -- -D warnings