5976ef89b4
A suite is the root object: every script belongs to exactly one suite, the
suite holds the ordered member list by reference - the optional setup first,
the members, the optional teardown last - and runs them in one card session,
without resets and without resume.
- test_suites.py: the suite store (test_suites.json, --test-suites), validated
and atomic like the script store: roles normalised (setup first, teardown
last, at most one of each, a script appears once per suite), per-member
on_fail stop/continue and the suite-level require_adm.
- test_scripts.py schema v2: suite_id/require_adm; list_by_suite()/orphans();
the import reports the id_map a bundle import remaps with. v3.21.0 scripts
without a suite are attached once to an auto-created "Imported scripts"
suite (migrate_scripts_to_suites, called from __main__).
- API: /api/test/suites (+ update|delete|import - delete refused while
non-empty; the bundle import imports the scripts first, remaps the suite
references and skips unresolvable ones); the scripts endpoints gain
suite_id/require_adm plus move|copy (delete detaches from the suite first;
a suite update refuses a membership change).
- runner: /api/test/run {suite_id} resolves the members and the preset, runs
the ADM gate, then setup -> members (per-member on_fail) -> teardown. The
teardown always runs on a stop or a failure (skipped only when the card is
gone; it clears the stop flag for its duration); a card reset stops the
suite, skips the remaining members and is flagged. Members' live steps land
in the shared snapshot, their records (status/steps/note/log excerpt/
counters) in _TEST_RUN['suite']; the summary carries the counts, wall time,
preset, the SCP80 counters before/after per keyset and the session flags.
_test_log buffers the semantic lines (bounded) for the report's per-member
excerpts.
- ADM gate (_ensure_adm): a suite/script require_adm verifies the matched
preset's key once per card session (the /api/verify-adm badge sets the same
latch); a failure refuses the start, stamps the session (no automatic
retry) and needs a manual verify.
- PWA: the Test script pill becomes four views - the suite list (default),
the suite editor (the script list: role badges, on_fail, up/down, per-script
Run/Edit/Move/Copy/Delete, Add script with the templates), the script editor
(the step list; a single run shows its result under the steps and is not
preserved) and the suite report (per-member blocks with the step table,
notes and run log; summary; export JSON/Markdown). Suite bundles export/
import; the legacy script export imports into "Imported scripts". EN/RU.
- tests: the suite store/migration, the runner orchestration (order, on_fail,
setup failure, stop, card reset, teardown, per-member logs) and the ADM gate
(Python); the suite HTTP paths; the PWA report/Markdown/wiring (Node).
Tests: 833 frontend + 794 Python; sw cache simple-v397.
604 lines
30 KiB
Markdown
604 lines
30 KiB
Markdown
# Test script authoring guide
|
||
|
||
A **test script** is a JSON document that drives a deterministic dialogue with
|
||
the card over the local server: ordered **actions** (each sends something and
|
||
checks the response) and **expectations** (each fetches the proactive command
|
||
the card announced, checks it, and sends the scripted TERMINAL RESPONSE).
|
||
|
||
Scripts live in the **server-side store**
|
||
(`~/.pysim-simple-server/test_scripts.json`, `--test-scripts`): the PWA's Test
|
||
script pill manages them (Import/Export move JSON in and out, the old
|
||
localStorage export imports as-is), and any API client can create or run them
|
||
(`POST /api/test/run` with an inline `script` or a stored `script_id` - see
|
||
`docs/api.md`). While a run is active the card belongs to the script.
|
||
|
||
**Test suites** (v3.22.0) group the scripts: a suite is the root object (every
|
||
script belongs to exactly one suite, `suite_id` in the script store) and the
|
||
runner executes its scripts in order - the optional **setup** first, the
|
||
**member** scripts, the optional **teardown** last - in **one card session,
|
||
without resets and without resume**. Per member `on_fail` decides whether a
|
||
failure stops the suite (`stop`, default) or the next script runs
|
||
(`continue`); the **teardown always runs** on a stop or a failure (its job is
|
||
to leave the card ready for a new test) as long as the card is in the reader,
|
||
and a card reset (session change) stops the suite and is flagged in the
|
||
report. A suite (or a single script) may require the **ADM** verified: the
|
||
matched preset's key is tried **once** at the start - a failure refuses the
|
||
run and needs a manual verify (the ADM badge) before the next attempt. The
|
||
suite run's report (the PWA's report view, exportable as JSON or Markdown)
|
||
carries every member's step table, failing notes and run-log excerpt, plus the
|
||
summary (pass/warn/fail/skipped, wall time, the SCP80 counters before/after
|
||
per keyset and the card-session flag). The suites live in
|
||
`~/.pysim-simple-server/test_suites.json` (`--test-suites`).
|
||
|
||
**Format**: server ≥ 3.15.0 (`kvn`, `por` and `files` checks), scripts stored
|
||
server-side since 3.16.0, `menu-select` by text and the spec-order TERMINAL
|
||
RESPONSE codings since 3.17.0, the semantic `event` action since 3.18.0, the
|
||
`envelope` action's `src` override since 3.18.1, the inline `por` object, the
|
||
`alpha`/`sms` checks, `scp80.params.format` and the case-insensitive
|
||
`menu-select` text match since 3.19.0, the `proactive` cleanup action and the
|
||
ok-answering bounded cleanup since 3.21.0, the suites since 3.22.0.
|
||
|
||
## Writing a test for an applet
|
||
|
||
Collect these values before writing the JSON (from the applet's source/spec
|
||
and the card's preset):
|
||
|
||
| What | Where it goes | Notes |
|
||
|---|---|---|
|
||
| TAR to address (the RFM application or the applet's own) + MSL, or an explicit SPI1 | `scp80.params.tar` / `spi1` | the card preset must carry the TAR with its MSL, or the step passes `spi1`; RFM TARs: `B00000`/`B00001`/`B00010` compact, `B00120`/`B00140`/`B00130` expanded (see Worksheet 1) |
|
||
| Keyset number | `scp80.params.kvn` (1–15) | only when the applet's keys are not the preset's first keyset |
|
||
| Trigger C-APDUs (+ their format) | `scp80.params.apdu` / `format` | one secured packet per `scp80` step; the preset provides keys and counter, `format` wraps the APDU for an expanded-format application |
|
||
| Expected file paths + contents | `file-read.params.path` + `check.data` | a mask must have **exactly** the actual value's length |
|
||
| Expected proactive command(s) | `expect.command` + `qualifier`/`checks` | REFRESH `qualifier: "00"` = NAA init + full FCN, `files` asserts the changed EFs |
|
||
| PoR transport + expected PoR | `scp80.check.por`/`check.data` (inline) or `expect SEND SHORT MESSAGE` + `por` check | pick per the SPI2 the applet expects |
|
||
| Delivery timing | a `status` action when the card delivers on poll | the runner never polls by itself |
|
||
|
||
### The three status words (envelope SW · PoR status · R-APDU SW)
|
||
|
||
An SCP80 step carries three different "SW" values; keep them apart when
|
||
writing checks:
|
||
|
||
| Where it comes from | What it means | Script field | Report row |
|
||
|---|---|---|---|
|
||
| the **ENVELOPE** exchange | the card's answer to the command that was sent (`9000` executed, `91XX` a proactive command is pending, `62XX`/`63XX` a warning) | the action's `check.sw` | `SW` |
|
||
| the **PoR status** | the SCP80 security-processing verdict inside the response packet (`por_ok`, `cntr_low`, `rc_cc_ds_failed`, `tar_unknown`, …) | `por.status` (or `"por": "ok"`) | `PoR status` |
|
||
| the **R-APDU SW** | the response payload's own status word - the last executed command's / GET RESPONSE's status in the compact form (TS 102 226 §5.1.2 Table 5.1), or the last R-APDU's in the scripting (`AB`/`AF`) form | `por.sw` | `PoR SW` |
|
||
|
||
`por.sw` is a child of the **response payload**, not of the envelope: it
|
||
exists only when the addressed TAR is a **remote-management application**
|
||
(ISD / RFM). A third-party applet's **own TAR** answers with
|
||
application-defined bytes - the platform passes them through unchanged - so
|
||
there is no R-APDU SW to check: `por.sw` reports `(none)` and the whole
|
||
secured data is exposed as `por.data`. Assert such a response with `data`
|
||
and leave `sw` empty; the step's `check.sw` still asserts the envelope
|
||
exchange.
|
||
|
||
The form is decided by the command's **TAR** (v3.21.0): the preset's role
|
||
TARs and the standard ISD/RFM allocations (TS 101 220 Annex D) are decoded as
|
||
RM responses; any other TAR - an applet's own - is reported `raw` regardless
|
||
of whether its bytes happen to parse as a compact response, so no status word
|
||
is ever fabricated (a 2-byte reply cannot be one: Table 5.1 needs the command
|
||
count plus two status bytes - it is reported raw too).
|
||
|
||
A `62XX`/`63XX` envelope answer does not guarantee a response packet: an
|
||
application that refuses at the envelope level (e.g. an applet answering
|
||
`6200` to an unknown command) sends none, while other warning paths still
|
||
deliver one (the low-counter case arrives via SEND SHORT MESSAGE). For a
|
||
step expected to be refused, leave `por` at its `"any"` default or assert
|
||
`"por": "none"`; the run report's `PoR[…]` log line and `por` field show
|
||
which happened.
|
||
|
||
### Worksheet 1 - RFM update → file change → REFRESH (full FCN)
|
||
|
||
```json
|
||
{"name": "applet RFM update",
|
||
"steps": [
|
||
{"type": "action", "kind": "scp80",
|
||
"params": {"apdu": "<RFM C-APDU>", "tar": "B00001", "kvn": 1, "spi2": "01"},
|
||
"check": {"sw": {"mode": "mask", "value": "91??"}, "por": "none"},
|
||
"label": "RFM update"},
|
||
{"type": "expect", "command": "REFRESH", "qualifier": "00",
|
||
"checks": [{"kind": "files", "value": "<changed FIDs, e.g. 3F007F206F07>"}],
|
||
"respond": {"result": "ok"}},
|
||
{"type": "action", "kind": "file-read", "params": {"path": "<EF path>"},
|
||
"check": {"data": {"mode": "mask", "value": "<expected bytes, ? = wildcard>"}}}
|
||
]}
|
||
```
|
||
|
||
`tar` addresses the **RFM application that owns the file system**, not the
|
||
applet-under-test: `B00000` = UICC shared FS, `B00001` = ADF RFM, `B00010` =
|
||
SIM FS (compact), and `B00120`/`B00140`/`B00130` the expanded-format twins
|
||
(TS 101 220 Annex D; the probe list in the SCP80 tab carries the same names).
|
||
An applet that accepts secured packets under its **own** TAR - the TAR given
|
||
in its INSTALL [for install] parameters (TS 102 226 §8.2.1.3.2.7) - uses that
|
||
TAR instead. The preset must carry a TAR entry **with its MSL** or the step
|
||
passes an explicit `spi1`; an expanded-format RFM application also expects the
|
||
wrapped command template (`"format": "expanded"`/`"expanded-ae"`, see the
|
||
actions table).
|
||
|
||
A **scripted SELECT inside the secured packet** should ask for no response
|
||
data (`P2='0C'`, e.g. `00A4000C02A153`): a card can refuse the `P2='00'`
|
||
form (return the FCI) with `6A86` when it arrives inside a command script,
|
||
and the script has no use for an FCI anyway. The selects walk from the
|
||
current directory the RFM application starts in: MF for the UICC shared FS
|
||
RFM application (TAR `B00000`, TS 102 226 §7.2), the ADF for an ADF RFM
|
||
application (TAR `B00001`, §7.3).
|
||
|
||
### Worksheet 2 - incoming data → particular PoR, no other actions
|
||
|
||
Inline PoR:
|
||
|
||
```json
|
||
{"name": "applet data (inline PoR)",
|
||
"steps": [
|
||
{"type": "action", "kind": "scp80",
|
||
"params": {"apdu": "<data C-APDU>", "tar": "<applet TAR>", "spi2": "01"},
|
||
"check": {"sw": "9000", "por": "ok",
|
||
"data": {"mode": "mask", "value": "<expected PoR bytes>"}}}
|
||
]}
|
||
```
|
||
|
||
PoR-in-submit:
|
||
|
||
```json
|
||
{"name": "applet data (SEND SM PoR)",
|
||
"steps": [
|
||
{"type": "action", "kind": "scp80",
|
||
"params": {"apdu": "<data C-APDU>", "tar": "<applet TAR>", "spi2": "21"},
|
||
"check": {"sw": {"mode": "mask", "value": "91??"}, "por": "none"}},
|
||
{"type": "expect", "command": "SEND SHORT MESSAGE",
|
||
"checks": [{"kind": "por", "status": "por_ok", "sw": "<R-APDU SW>",
|
||
"data": {"mode": "mask", "value": "<expected response data>"}}],
|
||
"respond": {"result": "ok"}}
|
||
]}
|
||
```
|
||
|
||
### Validation checklist
|
||
|
||
1. **Client-side**: the PWA import (or `testScriptProblem`) rejects structural
|
||
mistakes; the server validates again with the same engine at run start.
|
||
2. **Import** the JSON (PWA → Simulator → Test script → Import, or
|
||
`POST /api/test/scripts`), select the matched preset, **Run**.
|
||
3. **Read the report**: every action/expectation shows expected vs actual per
|
||
check; ⚠ = warning (continues), ✗ = error (stops). A failed step names the
|
||
reason (`note`).
|
||
4. **Iterate**: a mismatch on `data` usually means the mask length differs from
|
||
the actual value, or the value changed between runs - mask the varying
|
||
bytes (counter, date-time) with `?`.
|
||
5. **Counters**: the run starts from the preset's stored counter and advances
|
||
it after every scp80 step the card answers (when SPI1 requests a counter
|
||
check); a refused packet moves it forward too - see the presets section.
|
||
|
||
### Reading the run log
|
||
|
||
Every step writes a semantic line to the server log under its step number,
|
||
request before response:
|
||
|
||
```
|
||
TEST-RUN step 1: SCP80 C-APDU=80E2900000
|
||
TEST-RUN step 1: SCP80 SECURED=0100…1A2B (N B) TAR=B00001 SPI1=16 SPI2=21 CNTR=0000000A
|
||
TEST-RUN step 1: SCP80 -> SW=9000
|
||
TEST-RUN step 1: PoR[inline] status=por_ok TAR=B00001 CNTR=0000000A raw=0271…aabb
|
||
TEST-RUN step 1: R-APDU SW=9000 data=AABB
|
||
TEST-RUN step 2: STATUS 1/2 -> 9102
|
||
TEST-RUN step 3: FETCH=8012000002 -> SW=9000 D03C8103012500…
|
||
TEST-RUN step 3: CMD 0x25 SET UP MENU qual=00
|
||
TEST-RUN step 3: TR=81030125008202828183020000 -> SW=9000
|
||
TEST-RUN step 4: MENU-SELECT ENVELOPE=80c2000009d30702020181900101 item=1 ('One') -> SW=9103
|
||
```
|
||
|
||
- A PoR delivered as SEND SHORT MESSAGE is logged on the step that fetches it:
|
||
`PoR[sms-submit] …` followed by the same `R-APDU` line.
|
||
- A SEND SHORT MESSAGE expectation also logs its parsed TPDU
|
||
(`SMS DA=… PID=… DCS=… UDL=… UD=…`) after the `CMD` line, so an applet's own
|
||
SMS can be read even without an `sms` check.
|
||
- A wrapped scp80 step logs both forms:
|
||
`SCP80 C-APDU=<plain> WRAPPED=<wrapped> (<format>)`; a pre-built packet logs
|
||
`SCP80 SP=…`.
|
||
- Other actions log their send/response pair too (`APDU TX=… -> SW=…`,
|
||
`ENVELOPE(Event Download) … -> SW=…`, `READ BINARY <path>`,
|
||
`READ RECORD <n> <path>`, `UPDATE BINARY <path>`, each with `-> SW=…`).
|
||
- A `proactive` action logs its rounds (`STATUS i/N -> …`, `FETCH=…`,
|
||
`CMD 0x…`, `TR=… -> SW=…`); the run cleanup logs
|
||
`TEST-RUN drain: <CMD> (0xNN) TR=… -> SW=…`.
|
||
- A failing step adds `TEST-RUN step N failed: …` with the reason.
|
||
- The shared secured-packet sender logs each segment's answer
|
||
(`OTA SEND: ENVELOPE i/N -> <SW>`).
|
||
- `--apdu-trace` adds the raw transport view
|
||
(`APDU-TRACE(…): <cmd> → SW: … RESP: …`) for every APDU, including the
|
||
file-manager reads and pySim's auto-handler traffic.
|
||
|
||
## Document shape
|
||
|
||
```json
|
||
{"name": "applet RFM update",
|
||
"steps": [
|
||
{"type": "action", "kind": "scp80", "params": {…}, "check": {…}, "label": "optional"},
|
||
{"type": "expect", "command": "REFRESH", "qualifier": "00",
|
||
"checks": [{"kind": "files", "value": "3F007F206F07"}],
|
||
"respond": {"result": "ok"}}
|
||
]}
|
||
```
|
||
|
||
Every step may carry `on_fail`: `"error"` (default - stops the run) or
|
||
`"warning"` (continues, the report shows ⚠). Each entry of `checks` may
|
||
override it with its own `on_fail`. The level applies to **every** step
|
||
failure, including an expectation that finds no pending command (v3.21.0) -
|
||
the tolerant "consume a command if there is one" form is an expectation with
|
||
`on_fail: "warning"`; to drain and confirm the card is idle, use the
|
||
`proactive` action.
|
||
|
||
## Presets: where the keys, TARs and counters come from
|
||
|
||
An SCP80 action never carries key material. The PWA sends the **stored
|
||
preset's id** with the run; the server resolves it and takes:
|
||
|
||
- the keyset - the step's `kvn` (1–15) or the preset's first keyset,
|
||
- the keys and the counter of that keyset,
|
||
- the TAR table with each TAR's **MSL** (Minimum SPI1).
|
||
|
||
So before a run the matched preset (ICCID) must contain the keyset and, for
|
||
every TAR the script addresses, either a TAR entry with its MSL or an explicit
|
||
`spi1` in the step. A step that omits them gets the preset's ISD role TAR as
|
||
`tar`, that TAR's MSL as `spi1` and `01` as `spi2`; `kvn` defaults to the
|
||
first keyset.
|
||
|
||
Counter handling: the run starts from the keyset's stored counter and advances
|
||
(and persists) its next value after **every** scp80 step the card answers, when
|
||
the packet requests a counter check (SPI1 b5b4 ≠ 00, TS 102 225 §5.1.1) - a
|
||
refused packet moves the local value forward too. The local counter therefore
|
||
only ever moves up, so a later packet is never below what the card has seen;
|
||
consecutive runs continue correctly.
|
||
|
||
## Actions
|
||
|
||
```json
|
||
{"type": "action", "kind": "…", "params": {…}, "check": {…}}
|
||
```
|
||
|
||
| kind | params | notes |
|
||
|---|---|---|
|
||
| `envelope` | `event` 0–255, `data` hex (optional), optional `src` | ENVELOPE(Event Download) with raw data; `src` overrides the device-identities source (`82` terminal / `83` network) |
|
||
| `event` | `event` (name or hex), `fields` (object), optional `src` | semantic ENVELOPE(Event Download) for 0x03/0x0B/0x12/0x1D (see below) |
|
||
| `menu-select` | `item_id` 1–255 **or** `text` + `mode` `exact`/`contains` (+ `case_sensitive: true` for an exact-case match — the default is case-insensitive; the step editor's **Match case** choice writes it), not both | ENVELOPE(Menu Selection); the text is resolved against the card's cached menu - refreshed whenever the card sends SET UP MENU, so it follows the applet's install parameters |
|
||
| `file-write` | `path`, `data`, `mode` `auto`/`binary`/`record`, `record` 1–255 | UPDATE BINARY/RECORD |
|
||
| `file-read` | `path`, `mode`, `record` | READ BINARY/RECORD; verify with `check.data` |
|
||
| `apdu` | `apdu` hex | raw transport, no auto-handler |
|
||
| `scp80` | `apdu` **xor** `sp` (optional `source`), `kvn`, `tar`, `spi1`, `spi2`, `format` | secured packet to the TAR; `format`: `compact` (default - the C-APDU verbatim), `expanded` (the `AA`-prefixed Command Scripting template) or `expanded-ae` (the `AE 80 … 00 00` form), TS 102 226 §5.2.1 |
|
||
| `status` | `attempts` 1–1000, `interval_ms` 0–10000 | with `attempts > 1` the default SW check is `91??` (poll until a command) |
|
||
| `proactive` | `respond`, `first`, `attempts` 1–1000 (default 3), `interval_ms` (default 200), `require` | drain the card's announced proactive commands and confirm it is idle (see below) |
|
||
|
||
`path` is `/`-separated: `MF` (or `3F00`) or an ADF name/AID first, then FIDs
|
||
or file names - e.g. `MF/7F20/6F07`, `ADF.USIM/EF.TEST`.
|
||
|
||
`file-write` sends a direct terminal-side UPDATE BINARY/RECORD - it does **not**
|
||
fire the card's EVENT_REMOTE_FILE_UPDATE (that event reports an update done by
|
||
a remote entity through OTA/RFM, TS 102 223 §7.5.9): assert the change with
|
||
`file-read` or the REFRESH `files` check.
|
||
|
||
### Semantic events (`kind: "event"`)
|
||
|
||
The server builds the event objects for the events a test needs most; an
|
||
event without a builder goes through the raw `envelope` action, whose data hex
|
||
the Phone tab's event form shows (**Event data (hex)**, copyable). Both
|
||
actions take the same optional `src` device-identities override.
|
||
|
||
```json
|
||
{"type": "action", "kind": "event",
|
||
"params": {"event": "location_status",
|
||
"fields": {"status": 0, "mcc": "250", "mnc": "01",
|
||
"lac": "00FF", "cell": "0001"}}}
|
||
```
|
||
|
||
| `event` | fields |
|
||
|---|---|
|
||
| `location_status` / `0x03` | `status` 0 normal / 1 limited / 2 no service; for normal service `mcc`, `mnc`, `lac`, `cell` |
|
||
| `access_tech` / `0x0B` | `tech` (TS 102 223 8.61 coding: 0 GSM, 3 UTRAN, 8 E-UTRAN, 10 NG-RAN, …) |
|
||
| `network_rejection` / `0x12` | `reg_type` 0–17, `access_tech`, `cause`; the location object follows the registration group - LU: `lac`, GPRS: `lac`+`rac`, EPS/5GS: `tac` (4–6 hex) + optional `ext_cause`; `mcc`/`mnc`; optional `ext_info_type` (1 CAG ID / 2 NID / 3 RedCap) + `ext_info` |
|
||
| `data_connection` / `0x1D` | `status` 0–2, `type` 0–2, optional `cause`, `ti` (hex byte), `datetime: "now"` (host clock, 8.39), `mcc`/`mnc`/`lac`/`cell`, `tech`, `loc_status`, `apn`, `pdp_type` |
|
||
|
||
`src` overrides the device-identities source (`82` terminal, `83` network);
|
||
the default is the terminal, matching the Phone tab's forms.
|
||
|
||
### Proactive cleanup (`kind: "proactive"`)
|
||
|
||
Consume whatever the card announces and leave it idle - at the start, in the
|
||
middle or at the end of a script (a leftover command from a previous run must
|
||
not derail the next step). Each round polls STATUS; a `91XX` answer fetches
|
||
the command and answers it (the first command with `first`, the rest with
|
||
`respond` - both `{"result": …}` like an expectation's response). The loop
|
||
ends when STATUS answers `9000` (idle), a non-`91XX` SW appears or `attempts`
|
||
rounds are spent; the **final SW is the step's `check.sw`** (default `9000`),
|
||
so a still-pending card fails the step and hands the pending command to the
|
||
next step.
|
||
|
||
```json
|
||
{"type": "action", "kind": "proactive",
|
||
"params": {"respond": {"result": "ok"},
|
||
"first": {"result": "cancel"},
|
||
"attempts": 5, "interval_ms": 200,
|
||
"require": {"command": "REFRESH", "qualifier": "00"}},
|
||
"check": {"sw": "9000"},
|
||
"label": "serve whatever is pending, leave the card idle"}
|
||
```
|
||
|
||
- An **empty drain passes** (that is the point) - unless `require` is given:
|
||
the step then fails with "nothing was pending".
|
||
- `require` asserts **at least one** consumed command matches (`command` name
|
||
or hex type, `ANY` allowed; optional `qualifier`, exact/mask).
|
||
- `first` lets a script test the card's refusal path for the first command
|
||
(e.g. `{"result": "cancel"}`) while the rest are answered `ok`.
|
||
- Every consumed command is reported (type, qualifier, the TR's SW) and
|
||
logged (`STATUS`/`FETCH`/`CMD`/`TR` lines under the step).
|
||
|
||
### `check` on actions
|
||
|
||
```json
|
||
"check": {"sw": {"mode": "mask", "value": "91??"},
|
||
"data": {"mode": "exact", "value": "0271…"},
|
||
"por": "ok"}
|
||
```
|
||
|
||
- `sw` - default `{"mode": "exact", "value": "9000"}` (a `status` action with
|
||
`attempts > 1` defaults to the `91??` mask).
|
||
- `data` - the response data (for SCP80: the inline PoR packet).
|
||
- `por` (SCP80 actions only) - `"any"` (default, no check), `"ok"`
|
||
(`response_status == por_ok`), `"none"` (no PoR expected), or an **object**
|
||
`{"status"?, "sw"?, "data"?}` asserting the decoded PoR of this very
|
||
exchange (at least one field; the same fields as the expectation's `por`
|
||
check, `status` accepting `ok` for `por_ok`) - it works for both the inline
|
||
and the SEND SHORT MESSAGE transport.
|
||
Example: `"por": {"status": "por_ok", "sw": "9000"}`.
|
||
- A plain hex string is exact; a value containing `?` is a mask. **A mask
|
||
must have exactly the same length as the actual value** (`?` is a per-nibble
|
||
wildcard) - e.g. `91??` matches `9102`, `027100????` matches a 4-byte
|
||
prefix.
|
||
|
||
## Expectations
|
||
|
||
```json
|
||
{"type": "expect", "command": "SEND SHORT MESSAGE",
|
||
"qualifier": "00",
|
||
"checks": [{"kind": "por", "status": "por_ok", "sw": "9000"}],
|
||
"respond": {"result": "ok"}}
|
||
```
|
||
|
||
The card must have announced a command (`91XX`) in the previous step - the
|
||
runner never polls by itself. `command` is a name (as shown in the proactive
|
||
log) or a hex type; `ANY`/`*` skips the type check. `qualifier` is the
|
||
command-details byte 3 (one byte, exact/mask).
|
||
|
||
When nothing is pending the step fails at its own `on_fail` level (the
|
||
default stops the run; `on_fail: "warning"` warns and continues - the
|
||
tolerant consume-if-pending form, v3.21.0). "Add a `status` action" applies
|
||
when the card delivers on poll: `status` with `attempts > 1` polls until a
|
||
command appears; it cannot make one appear when the card has none.
|
||
|
||
### Content checks
|
||
|
||
| kind | fields | checks |
|
||
|---|---|---|
|
||
| `text` | `mode` contains/exact, `value`, `case_sensitive` | the **Text string** (`8D`, DCS coded) of DISPLAY TEXT / GET INKEY / GET INPUT |
|
||
| `alpha` | `mode` contains/exact, `value`, `case_sensitive` | the command's **Alpha identifier** (tag `05`/`85`, TS 102 223 §8.2, Annex A coded) - distinct from `text` when a command carries both |
|
||
| `item` | `id`, `text`, `mode`, `case_sensitive` | items of SELECT ITEM / SET UP MENU |
|
||
| `raw` | `mode` exact/mask, `value` | the whole fetched command (hex, exact length needed) |
|
||
| `por` | `status`, `sw`, `data` | the PoR carried by a **SEND SHORT MESSAGE** (see below) |
|
||
| `sms` | `da`, `pid`, `dcs`, `udl`, `ud` | the TPDU of a SEND SHORT MESSAGE (see below) |
|
||
| `files` | `files` (list of path hex strings) or a comma/space separated string | the **File List** of a REFRESH, TS 102 223 §8.18; matched order-insensitively |
|
||
|
||
`por` needs an SCP80 step **before** the expectation: the TPDU of the SEND
|
||
SHORT MESSAGE is decoded with that step's keyset, SPI and counter. `status`
|
||
is the decoded response status name (`por_ok`, `cntr_low`,
|
||
`rc_cc_ds_failed`, …; `ok` is accepted for `por_ok`); `sw` is the R-APDU
|
||
status word **inside the response payload** - not the envelope's, see
|
||
*The three status words* in the authoring section - and `data` the R-APDU
|
||
response data (exact/mask); both come from
|
||
the decoded response (scripting `AB`/`AF`/compact forms). An applet's **own
|
||
TAR** answers with its application-defined bytes: when the data cannot be the
|
||
compact remote response (its command count exceeds the command script that
|
||
was sent), the whole secured data is reported as `data` and no `sw` is
|
||
decoded - assert such a response with `data` (the action's `por` object
|
||
follows the same rule). A PoR split over
|
||
several SMS parts is accumulated across SEND SHORT MESSAGE expectations: put
|
||
the `por` check on the step that completes the sequence (intermediate steps
|
||
should use `raw` or no PoR check).
|
||
|
||
`sms` parses the fetched SEND SHORT MESSAGE's TPDU (SMS-SUBMIT only - any other
|
||
MTI fails the check): `da` the destination digits as the TPDU carries them
|
||
(exact; whitespace and a leading `+` in the script are ignored - the
|
||
international flag lives in the TPDU's type-of-number),
|
||
`pid`/`dcs` single bytes and `ud` the raw user data (a UDH included) as hex
|
||
exact/mask, `udl` the user-data length in decimal (octets, or septets for the
|
||
7-bit alphabet - whatever the TPDU carries). Use it for an applet's own SMS;
|
||
the OTA PoR of the preceding scp80 step is what `por` decodes. At least one
|
||
field is required.
|
||
|
||
### `respond`
|
||
|
||
Optional - an omitted (or empty) object answers `ok` (`0x00`). The TERMINAL
|
||
RESPONSE follows the TS 102 223 §6.8.0 object order (Command details, Device
|
||
identities, **Result**, Duration, Text string, Item identifier, …):
|
||
|
||
- `result` - name or hex (`ok`, `partial`, `missing`, `refused`,
|
||
`not_understood`, `modified`, `cancel`, `back`, `timeout`, `no_response`).
|
||
- `text` + `dcs` - the Text string for a GET INKEY / GET INPUT answer. `dcs`
|
||
uses the spec's recommended codings: `"00"` GSM default alphabet 7 bits
|
||
**packed**, `"04"` GSM default alphabet 8 bits (one octet per character),
|
||
`"08"` **UCS2** (UTF-16-BE). An empty `text` sends the null text string
|
||
(Length `00`).
|
||
- `item_id` - the Item identifier for a successful SELECT ITEM answer.
|
||
- `raw` - extra COMPREHENSION-TLVs appended after the standard objects.
|
||
|
||
UCS2 answer to a GET INPUT:
|
||
|
||
```json
|
||
{"type": "expect", "command": "GET INPUT",
|
||
"respond": {"result": "ok", "text": "Пароль", "dcs": "08"}}
|
||
```
|
||
|
||
The `proactive` action's `respond`/`first` use the same shape (the command
|
||
type varies per drained command, so `text`/`item_id` apply where the command
|
||
carries them).
|
||
|
||
## The two PoR transports (SPI2)
|
||
|
||
1. **Inline** - the card answers the ENVELOPE `9000` with the PoR in the
|
||
response data. Verify on the action: `"por": "ok"` and/or
|
||
`"data": {"mode": "mask", "value": "…"}`, or assert the decoded PoR with
|
||
`"por": {"status": "por_ok", "sw": "9000"}`.
|
||
2. **PoR-in-submit** (SPI2 bit `20`, e.g. `spi2: "21"`) - the card announces
|
||
a **SEND SHORT MESSAGE** (`91XX`); the PoR is inside its SMS TPDU:
|
||
|
||
```json
|
||
{"type": "action", "kind": "scp80",
|
||
"params": {"apdu": "…", "tar": "B00001", "spi2": "21"},
|
||
"check": {"sw": {"mode": "mask", "value": "91??"}, "por": "none"}},
|
||
{"type": "expect", "command": "SEND SHORT MESSAGE",
|
||
"checks": [{"kind": "por", "status": "por_ok", "sw": "9000"}],
|
||
"respond": {"result": "ok"}}
|
||
```
|
||
|
||
If the card announces the SEND SHORT MESSAGE only on a later STATUS poll,
|
||
insert a `status` action (`attempts` > 1) before the expectation.
|
||
|
||
## "No actions" after a trigger
|
||
|
||
An exact SW check is the assertion: `"check": {"sw": "9000"}` fails if the
|
||
card answers `91XX` (a command was announced); the runner also errors on a
|
||
pending command the next step does not expect.
|
||
|
||
## Worked examples
|
||
|
||
### Event download before an applet action
|
||
|
||
```json
|
||
{"name": "applet event",
|
||
"steps": [
|
||
{"type": "action", "kind": "event",
|
||
"params": {"event": "location_status",
|
||
"fields": {"status": 0, "mcc": "250", "mnc": "01",
|
||
"lac": "00FF", "cell": "0001"}},
|
||
"check": {"sw": {"mode": "mask", "value": "91??"}}},
|
||
{"type": "expect", "command": "ANY", "respond": {"result": "ok"}}
|
||
]}
|
||
```
|
||
|
||
The event's default SW check is an exact `9000`; a mask (`91??`) is needed
|
||
when the applet answers with a proactive command, which the next expectation
|
||
then fetches.
|
||
|
||
### Menu selection by text (the id varies with the install parameters)
|
||
```json
|
||
{"name": "applet menu",
|
||
"steps": [
|
||
{"type": "action", "kind": "menu-select",
|
||
"params": {"text": "Settings", "mode": "exact"}},
|
||
{"type": "expect", "command": "ANY", "respond": {"result": "ok"}}
|
||
]}
|
||
```
|
||
|
||
The text is matched against the card's cached menu, **case-insensitively by
|
||
default** (`case_sensitive: true` forces an exact-case match); the cache is
|
||
refreshed whenever the card sends SET UP MENU (e.g. after an install +
|
||
REFRESH). If the
|
||
script must wait for the menu, add a `status` action and an
|
||
`expect SET UP MENU` before the selection.
|
||
|
||
### RFM update → file change → REFRESH (NAA init + full FCN)
|
||
|
||
```json
|
||
{"name": "applet RFM update",
|
||
"steps": [
|
||
{"type": "action", "kind": "scp80",
|
||
"params": {"apdu": "80E2900000", "tar": "B00001", "spi2": "01"},
|
||
"check": {"sw": {"mode": "mask", "value": "91??"}, "por": "none"},
|
||
"label": "RFM update"},
|
||
{"type": "expect", "command": "REFRESH", "qualifier": "00",
|
||
"checks": [{"kind": "files", "value": "3F007F206F07"}],
|
||
"respond": {"result": "ok"}},
|
||
{"type": "action", "kind": "file-read", "params": {"path": "MF/7F20/6F07"},
|
||
"check": {"data": {"mode": "mask", "value": "AA??CC"}}}
|
||
]}
|
||
```
|
||
|
||
`qualifier` `00` is "NAA Initialization and Full File Change Notification";
|
||
the `files` check asserts the exact set of changed files the applet reports.
|
||
For this mode the terminal must **not** reset the card
|
||
(TS 102 223 §6.4.7), so the run continues - the following `file-read`
|
||
verifies the updated content.
|
||
|
||
### Incoming data → particular PoR, no other actions
|
||
|
||
Inline PoR:
|
||
|
||
```json
|
||
{"name": "applet data (inline PoR)",
|
||
"steps": [
|
||
{"type": "action", "kind": "scp80",
|
||
"params": {"apdu": "80E2900000", "tar": "B00001", "spi2": "01"},
|
||
"check": {"sw": "9000", "por": "ok",
|
||
"data": {"mode": "mask", "value": "027100????0102"}}}
|
||
]}
|
||
```
|
||
|
||
PoR-in-submit (with the transport asserted by the expectation):
|
||
|
||
```json
|
||
{"name": "applet data (SEND SM PoR)",
|
||
"steps": [
|
||
{"type": "action", "kind": "scp80",
|
||
"params": {"apdu": "80E2900000", "tar": "B00001", "spi2": "21"},
|
||
"check": {"sw": {"mode": "mask", "value": "91??"}, "por": "none"}},
|
||
{"type": "expect", "command": "SEND SHORT MESSAGE",
|
||
"checks": [{"kind": "por", "status": "por_ok"}],
|
||
"respond": {"result": "ok"}}
|
||
]}
|
||
```
|
||
|
||
## Run errors worth knowing
|
||
|
||
- `SCP80 preset is incomplete: …` - the matched preset has no complete
|
||
keyset; `step N: keyset K is not defined in the preset` - the step's `kvn`
|
||
is outside the preset's keysets.
|
||
- `no MSL for TAR … - set it in the Cards tab or pass spi1` - the step's TAR
|
||
has no entry in the preset and the step carries no `spi1`.
|
||
- `no proactive command pending` - an expectation ran without a `91XX` from
|
||
the previous step; it fails at the step's `on_fail` level (the check row
|
||
names it). A `status` action with `attempts > 1` polls until a command
|
||
appears when the card delivers on poll; `on_fail: "warning"` tolerates the
|
||
empty case, and the `proactive` action drains and confirms idle.
|
||
- `menu-select: no menu item matches …` - the error lists the cached menu; a
|
||
miss on an explicit `case_sensitive: true` names the case-insensitive
|
||
candidates.
|
||
- `card verdict: no PoR (the card sent none for SPI2 …)` - an SCP80 operation
|
||
expected a PoR that never arrived (check SPI2 and the transport).
|
||
|
||
## Run cleanup
|
||
|
||
When a run ends (an error, a stop request, or an unexpected pending command)
|
||
with a command still announced, the runner fetches it and answers **`ok`
|
||
(0x00)** and keeps draining while the TERMINAL RESPONSE answers `91XX`
|
||
(bounded) - a cancel (`0x10`) is not sent automatically: it tells the card
|
||
the user aborted the proactive session, which can push assertive applications
|
||
onto an error path and make them queue a further command. The cleanup is
|
||
reported like a normal step (`TR ok …`, `pending REFRESH (0x01) answered with
|
||
ok`); a script that wants to test the refusal path answers explicitly (an
|
||
expectation's `respond`, or the proactive action's `first`).
|
||
|
||
## Known limitations
|
||
|
||
- The runner executes one command at a time and never polls by itself; a
|
||
proactive command that arrives only on a STATUS poll needs an explicit
|
||
`status` action.
|
||
- A `por` check only decodes SEND SHORT MESSAGE commands; the inline transport
|
||
is checked on the action (`por`/`data`).
|
||
- An expanded-format command (`"format": "expanded"`/`"expanded-ae"`) has no
|
||
in-packet GET RESPONSE (TS 102 226 §5.2.1.1): if the card answers `61xx`,
|
||
send the GET RESPONSE as a follow-up `scp80` step (its own secured packet,
|
||
same format).
|
||
- `files` matches the File List as a set (order-insensitive); duplicate paths
|
||
are kept.
|
||
- Semantic `event` actions cover four events (0x03/0x0B/0x12/0x1D); every
|
||
other event uses the raw `envelope` action with the data hex from the
|
||
Phone tab's event form.
|