Compare commits

...
Author SHA1 Message Date
dc 59e09e26a5 fix(deploy): drop the deprecated service-pack deploy from the setup screen
Build / Build-and-ng-test (pull_request) Successful in 4m34s
Lighthouse Checks / lighthouse (pull_request) Successful in 17m54s
Build / Build-and-test-development (pull_request) Successful in 29m50s
The Viya setup screen still carried a path that deployed the backend as a
service pack from a viya.json (`deployServicePack`). That is no longer how a
deployment works: the frontend and services arrive with the sasjs deploy, and the
only thing left to do is create the control library and its configuration data,
which `makedata` does - and then deletes itself.

The path was already unreachable from the UI (its button had been commented out
"for reference"), so the wizard only ever ran `makedata` in practice while its
copy still claimed to deploy services. Removed:

- `executeJson()` and the `viya.json` upload plumbing: `readFile`,
  `clearUploadInput`, and the fields and status flag that fed them
- the "Deploy SAS Jobs" row in the completion modal, which could never show
- the commented-out button and file input that kept the dead path visible

The screen now says what it does: it runs the one-off job that creates the
control library and its configuration data, and the status row reads "Create
control library" rather than "Create database".
2026-10-29 18:34:35 +00:00
allan eaa3b3e619 Merge pull request 'fix(embed): a rejected mapping must not block the page' (#406) from fix/va-parmap-column into release/7.22
Sites / deploy (push) Successful in 2m51s
Build / Build-and-ng-test (pull_request) Successful in 4m55s
Lighthouse Checks / lighthouse (pull_request) Successful in 17m45s
Build / Build-and-test-development (pull_request) Successful in 28m56s
Reviewed-on: #406
2026-10-29 18:29:30 +00:00
dc d845b9e9b7 fix(embed): a rejected mapping must not block the page
Build / Build-and-ng-test (pull_request) Successful in 4m55s
Lighthouse Checks / lighthouse (pull_request) Successful in 16m59s
Build / Build-and-test-development (pull_request) Successful in 27m58s
The deferred VA load builds the filter from the report's parameters before the
first getdata, which is what makes that first load already filtered. A va_parmap
entry is a column the report author typed, so the request can legitimately fail
when the table does not have it - validatefilter aborts on an unknown variable -
and that abort surfaced as a modal over an editor which then loaded fine
(dc/dc#404).

Validating the mapping first would cost a SAS call on every load of a mapped
object, which is not affordable, so the attempt is simply made quietly:
saveQuery takes wrapper options and the deferred load suppresses the error
modal. A mapping that is right still filters the very first load; a wrong one
falls back to the normal load, and the metadata pass reports it in the banner
and reconfigures once, at setup time rather than on every page load.
2026-10-29 18:22:41 +00:00
semantic-release-bot 428e70a4c1 chore(release): 7.21.0 [skip ci]
# [7.21.0](https://git.datacontroller.io/dc/dc/compare/v7.20.0...v7.21.0) (2026-10-29)

### Bug Fixes

* **client:** parse word-date values from VA parameters ([405b28d](405b28dccc))
* **client:** parse word-date values from VA parameters ([ce5b1aa](ce5b1aa961))
* **makedata:** name the sasjs server group in the estate seed ([eeffa92](eeffa92f9e))

### Features

* **derived:** apply the core field scan in the MPE_DERIVED hook ([a4a0f9c](a4a0f9cb2f))
2026-10-29 16:32:45 +00:00
allan 94e040c006 Merge pull request 'feat: Release/7.21' (#401) from release/7.21 into main
Release / Build-production-and-ng-test (push) Successful in 3m30s
Release / Build-and-test-development (push) Successful in 20m56s
Release / release (push) Successful in 5m15s
Reviewed-on: #401
2026-10-29 16:05:27 +00:00
allan cdc0ab2a91 Merge pull request 'fix(client): parse word-date values, and seed demo dates that repeat (release/7.21)' (#405) from fix/va-word-date-release into release/7.21
Sites / deploy (push) Successful in 2m56s
Build / Build-and-ng-test (pull_request) Successful in 4m58s
Lighthouse Checks / lighthouse (pull_request) Successful in 18m9s
Build / Build-and-test-development (pull_request) Successful in 27m56s
Reviewed-on: #405
2026-10-29 15:08:00 +00:00
dc 8449ab3987 test(data): seed the demo table with dates that repeat
Build / Build-and-ng-test (pull_request) Successful in 4m51s
Lighthouse Checks / lighthouse (pull_request) Successful in 17m26s
Build / Build-and-test-development (pull_request) Canceled after 18m20s
The demo rows had a random date each, so a date filter over the demo table
matched at most a couple of rows - which is not enough to show a date filter
working, in a report or in a test. The generated rows now cycle through 24
first-of-month dates over two years, about 20 rows each.

No test asserts a specific generated date; the fixed rows at the top of the
table are unchanged.
2026-10-29 14:42:33 +00:00
dc 405b28dccc fix(client): parse word-date values from VA parameters
A date parameter whose display format is not DATE9 or a numeric date reached
DC as text like "January 1, 2026" and the value was dropped: the clause
builder produced nothing for that parameter, so with no mapping the editor
filtered on the other parameters only, with no banner to say the date had been
ignored. Reproduced inside a report - a SOME_DATE parameter set to
"January 1, 2026" left the FILTER line reading
(( SOME_DROPDOWN = 'Option 1' )) and the grid at the unfiltered row count.

parseMonthNameDate now accepts a month name spelled out or abbreviated, with
the day before or after it, comma optional and "." or "-" as separators, so
WORDDATE-style values ("January 1, 2026", "Jan 1, 2026", "1 January 2026",
"1-Jan-2026") convert like DATE9 does. Only the first three letters of the
month name are matched, case-insensitively.

A month and year with no day ("January 2026") is still rejected rather than
being read as the first of the month - inventing a day would silently filter
on rows the author did not choose.

Unit tests cover each accepted shape, the case-insensitive match, DATE9 still
parsing as before, and the month-without-day rejection.
2026-10-29 14:42:33 +00:00
allan 559221fad4 Merge pull request 'fix(client): parse word-date values, and seed demo dates that repeat' (#403) from fix/va-date-worddate into main
Release / Build-production-and-ng-test (push) Canceled after 10s
Release / Build-and-test-development (push) Canceled after 0s
Release / release (push) Canceled after 0s
Reviewed-on: #403
2026-10-29 14:39:39 +00:00
dc 999e16b8fc test(data): seed the demo table with dates that repeat
Build / Build-and-ng-test (pull_request) Successful in 4m6s
Lighthouse Checks / lighthouse (pull_request) Successful in 17m45s
Build / Build-and-test-development (pull_request) Successful in 26m22s
The demo rows had a random date each, so a date filter over the demo table
matched at most a couple of rows - which is not enough to show a date filter
working, in a report or in a test. The generated rows now cycle through 24
first-of-month dates over two years, about 20 rows each.

No test asserts a specific generated date; the fixed rows at the top of the
table are unchanged.
2026-10-29 14:05:29 +00:00
dc ce5b1aa961 fix(client): parse word-date values from VA parameters
Build / Build-and-ng-test (pull_request) Successful in 3m53s
Lighthouse Checks / lighthouse (pull_request) Successful in 17m17s
Build / Build-and-test-development (pull_request) Failing after 27m2s
A date parameter whose display format is not DATE9 or a numeric date reached
DC as text like "January 1, 2026" and the value was dropped: the clause
builder produced nothing for that parameter, so with no mapping the editor
filtered on the other parameters only, with no banner to say the date had been
ignored. Reproduced inside a report - a SOME_DATE parameter set to
"January 1, 2026" left the FILTER line reading
(( SOME_DROPDOWN = 'Option 1' )) and the grid at the unfiltered row count.

parseMonthNameDate now accepts a month name spelled out or abbreviated, with
the day before or after it, comma optional and "." or "-" as separators, so
WORDDATE-style values ("January 1, 2026", "Jan 1, 2026", "1 January 2026",
"1-Jan-2026") convert like DATE9 does. Only the first three letters of the
month name are matched, case-insensitively.

A month and year with no day ("January 2026") is still rejected rather than
being read as the first of the month - inventing a day would silently filter
on rows the author did not choose.

Unit tests cover each accepted shape, the case-insensitive match, DATE9 still
parsing as before, and the month-without-day rejection.
2026-10-29 13:43:47 +00:00
allan f74100a818 Merge pull request 'feat(derived): apply the core field scan in the MPE_DERIVED hook' (#402) from refactor/derivedvarcheck into release/7.21
Sites / deploy (push) Successful in 3m23s
Build / Build-and-ng-test (pull_request) Successful in 6m10s
Lighthouse Checks / lighthouse (pull_request) Successful in 18m15s
Build / Build-and-test-development (pull_request) Successful in 29m46s
Reviewed-on: #402
2026-10-29 11:25:26 +00:00
dc a4a0f9cb2f feat(derived): apply the core field scan in the MPE_DERIVED hook
Build / Build-and-ng-test (pull_request) Successful in 6m6s
Lighthouse Checks / lighthouse (pull_request) Successful in 20m15s
Build / Build-and-test-development (pull_request) Successful in 32m40s
@sasjs/core 6.1.3 moved the derived variable field rules out of
mp_derivedvars into mp_derivedvarcheck, so the same scan can run before a
definition is used.  DC carried its own copy of that list in
mpe_derived_check.sas, which is now deleted: the post edit hook calls the
core macro instead, and the hook keeps only the rules that belong to the
control table rather than the expression - LIBREF/DSN required, and ACTIVE
within the picker pair.

The scan is stricter than the copy it replaces, in three ways:

- VAR_NM may not be _N_, _ERROR_ or _IORC_ (automatic variables of the data
  step)
- VAR_LOGIC may not exceed 4000 characters
- VAR_LOGIC may not call a function that submits code or reaches the file
  system, the environment or the macro symbol table - a function name cannot
  be assembled at run time, so this is a control the expression cannot evade

and it relaxes one: VAR_LABEL may carry a line break, which is flattened to a
space when the label is used, so only the remaining control characters are
refused.

The mock hook mirror is updated to the same rules, and stagedata.test.4
gains the two submit-time cases that pin the new wiring.
2026-10-29 10:51:00 +00:00
allan e6a5cce475 Merge pull request 'fix(makedata): name the sasjs server group in the estate seed' (#400) from fix/estate-seed-admin into release/7.21
Sites / deploy (push) Successful in 4m20s
Build / Build-and-ng-test (pull_request) Successful in 6m53s
Lighthouse Checks / lighthouse (pull_request) Successful in 20m47s
Build / Build-and-test-development (pull_request) Successful in 33m53s
Reviewed-on: #400
2026-10-29 08:33:52 +00:00
semantic-release-bot 97b4cd4d61 chore(release): 7.20.0 [skip ci]
# [7.20.0](https://git.datacontroller.io/dc/dc/compare/v7.19.0...v7.20.0) (2026-10-28)

### Bug Fixes

* **ci:** install doxygen for the SAS docs job ([f58925b](f58925b299))
* **ci:** keep foreign release assets, install jq first, pin the CLI ([27686d3](27686d3aad))
* **ci:** publish against the tag this job cuts, not releases/latest ([8888734](888873421c))
* **ci:** publish docs after the assets, with the webdoc app password ([dd482c4](dd482c4069))
* **ci:** publish the code documentation with surfer 7 and app passwords ([25ffa71](25ffa71add))
* **ci:** publish the typedoc to the webdoc site, not the app default ([c5808a1](c5808a16eb))
* **ci:** rebuild and upload release assets on demand ([6486c80](6486c80ed4))
* **ci:** use the lockfile and the pinned CLI in the docs job ([af1813f](af1813f829))
* **ci:** verify the forge certificate on the release API calls ([cf7b21a](cf7b21a8b6))
* **cypress:** hold download out too, and retry once in CI ([c97e886](c97e8866db)), closes [#395](#395) [#398](#398)
* **cypress:** register the deleteFolder task, restore the spec order, hold metanav out ([c803773](c803773d3c)), closes [#395](#395)
* **deploy:** do not offer the deploy wizard inside a report ([80e8911](80e8911af8))
* **deploy:** keep the app query when opening the standalone URL ([e51953c](e51953ca8a))
* **deploy:** one definition of the standalone URL, and test the guards ([459ff2d](459ff2da37)), closes [#TCS](https://git.datacontroller.io/dc/dc/issues/TCS)
* **deploy:** strip the embed parameters by name, and scope the timezone runs to the VA specs ([f7e7538](f7e753863b))
* **embed:** build quarter ends from the calendar, not minus 24 hours ([7e3045a](7e3045a21b))
* **embed:** defer the column check until metadata loads, report AND-ed pairs ([2623a3f](2623a3fb4b))
* **embed:** log a parse-time mapping problem once ([0011f7e](0011f7e933))
* **embed:** make the va_parmap diagnostics say what they mean ([541a61b](541a61ba08))
* **embed:** match a mapping on the parameter label only, and warn only when the values differ ([e9216b0](e9216b0895))
* **embed:** never build a clause from a value that did not convert, and say so in the banner ([bd3c220](bd3c220af2))
* **embed:** one parameter maps one column, and report typos on both sides ([f83dc35](f83dc359e2))
* **embed:** record the parameter label only, so a name-keyed va_parmap is reported ([aa91daf](aa91daf9f7))
* **embed:** resolve transform values as local dates and give ranges whole days ([10157c1](10157c1159))
* **embed:** warn when a va_parmap entry is ignored ([411f6c3](411f6c3354))
* **makedata:** pair the %else of the keepme block so it parses in open code ([c0ea810](c0ea810766))

### Features

* **embed:** map VA parameters to DC columns from the URL ([15a1f34](15a1f34d52))
* **embed:** resolve va_parmap values through named date anchors ([88010c3](88010c3980))
2026-10-28 23:11:50 +00:00
dc eeffa92f9e fix(makedata): name the sasjs server group in the estate seed
Build / Build-and-ng-test (pull_request) Successful in 6m57s
Lighthouse Checks / lighthouse (pull_request) Successful in 20m54s
Build / Build-and-test-development (pull_request) Successful in 34m49s
makeDataServer.json seeded ADMIN=AllUsers, a Viya-style group that no account on
the shared estate belongs to. The seed still reports SUCCESS with a group the
caller is not in, and then every getdata/stagedata call aborts with
'<user> is not authorised to edit <libds> in the MPE_SECURITY table', which
reads like a broken deployment. The estate's accounts are in all-users, the
SASjs Server group every logged-in user has, so seed with that and describe the
rule in the estate-testing skill rather than telling the reader to override the
shipped value.
2026-10-28 22:48:02 +00:00
allan dd25be8e75 Merge pull request 'Release/7.20' (#387) from release/7.20 into main
Release / Build-production-and-ng-test (push) Successful in 4m19s
Release / Build-and-test-development (push) Successful in 28m0s
Release / release (push) Successful in 7m31s
Reviewed-on: #387
2026-10-28 22:35:17 +00:00
allan 2e40749052 Merge pull request 'docs(skills): add the estate deploy-and-test procedure' (#397) from docs/estate-testing-skill into release/7.20
Build / Build-and-ng-test (pull_request) Canceled after 0s
Build / Build-and-test-development (pull_request) Canceled after 0s
Lighthouse Checks / lighthouse (pull_request) Canceled after 0s
Reviewed-on: #397
2026-10-28 22:34:37 +00:00
dc ad061946c3 docs(skills): add the estate deploy-and-test procedure
Build / Build-and-ng-test (pull_request) Successful in 6m56s
Lighthouse Checks / lighthouse (pull_request) Canceled after 19m45s
Build / Build-and-test-development (pull_request) Canceled after 16m16s
The shared-estate cycle is two npm scripts and sasjs test, but the seed in
npm run server is load-bearing: without it testsetup 404s settings.sas and
59 of the 64 test programs fail to complete, which reads like a broken
deployment. Record the targets, the token-pair credentials, the ADMIN group
the estate actually uses, how to read the results, and the frontend
passphrase steps as a repo-local skill, and point the testing doc at it.
2026-10-28 22:03:34 +00:00
24 changed files with 513 additions and 478 deletions

No files matched your search

+2 -1
View File
@@ -1,10 +1,11 @@
# Testing
Backend (SAS) tests are run with the sasjs CLI from the `sas/` directory.
Backend (SAS) tests are run with the sasjs CLI from the `sas/` directory. For a deploy-and-run on the shared state see the `dc-estate-testing` skill (`.agents/skills/dc-estate-testing/SKILL.md`), which covers the credentials, the `makedata` seed the suite needs and the estate's group name.
## Commands
- Full deploy + test cycle: `npm run 4gl` (compiles, deploys to the 4gl target, runs makedata) then `sasjs test -t 4gl`.
- The shared estate under its `server` target: `npm run server` (compile, deploy, seed) then `sasjs test -t server`.
- Run tests only (no rebuild/redeploy): `sasjs test -t 4gl`.
- Run a subset of tests: `sasjs test -t 4gl SOMESTRING` — the positional argument filters to matching tests (e.g. `sasjs test -t 4gl stagedata.test.3`).
+79
View File
@@ -0,0 +1,79 @@
---
name: dc-estate-testing
description: >
Use when deploying Data Controller to a shared SASjs Server estate (sas.4gl.io) or running the SAS
test suite against it. Covers the target and credential setup, the makedata seed the suite needs on
a fresh appLoc and the group it must name, how to read the results, and how to deploy the frontend
so the streamed app is the build you just made.
---
# Deploying and testing on the shared SASjs Server estate
The whole cycle is two npm scripts plus `sasjs test`. Everything below explains what those scripts do and the steps that go wrong.
cd sas && npm ci # macro dependencies
cd ../client && npm ci && npm run build # frontend only - see "Deploying the frontend"
cd ../sas && npm run server # cpfavicon + sasjs cbd -t server + makedata
npx sasjs test -t server
`npm run server` is `sasjs cbd -t server && npm run serverdata`, and the seed in the second half is not optional. `sasjs cbd` on its own deploys the services and the test programs but leaves the appLoc with no Data Controller configuration, and `testsetup` reads `<appLoc>/services/public/settings.sas` through the drive API. `makedata` is what creates that file, so without the seed `testsetup` gets a 404, aborts, and 59 of the 64 test programs fail to complete on setup alone - which reads like a broken deployment rather than a missing seed.
## The targets
`server` is the shared estate: `serverType: SASJS` on https://sas.4gl.io, `appLoc` `/Public/app/dc`, a real SAS 9 binary behind SASjs Server (`GET /SASjsApi/info` reports `mode: server`, `runTimes: ["sas"]`, `localLoginEnabled: true`). `4gl` is the same estate with a slightly different config, hence the older `npm run 4gl` / `4glmakedata` pair.
A run against `server` exercises the SASjs Server branch of every macro and the SAS 9 backend, and nothing else: no Viya branch, no SAS 9 STP branch. Say that when reporting, rather than implying the matrix is covered.
To deploy your own branch without replacing the app everyone else uses, add a target that copies `server` and changes `appLoc` and `streamServiceName` (for example `/Public/app/<you>/dc`), keep it out of the commit, and pass it with `-t`. `sasjs cbd` overwrites the whole appLoc it is pointed at.
## Credentials
The CLI logs in with a registered client, not a username and password: `.env.<target>` (gitignored, mode 600) needs `CLIENT`, `ACCESS_TOKEN` and `REFRESH_TOKEN`. With `CLIENT` alone it prints a `/SASjsLogon?client_id=...` URL and waits for an authorization code, which cannot be answered headless.
Mint the pair from a local login, with the password read from a file rather than argv:
1. `GET /` with a cookie jar - it sets `XSRF-TOKEN`, which every POST must echo in an `X-XSRF-TOKEN` header.
2. `POST /SASLogon/login` with `{"username","password"}`.
3. `GET /SASjsApi/client` on the same session - the registered client ids.
4. `POST /SASLogon/authorize` with `{"clientId"}` -> `{"code"}`.
5. `POST /SASjsApi/auth/token` with `{"clientId","code"}` -> `accessToken` / `refreshToken`.
An unmatched `/SASjsApi/...` path answers `200` with the Swagger page, so check the content type; HTML means the route does not exist. Delete the token file when the run is over.
## The makedata seed
`npm run serverdata` sends `deploy/makeDataServer.json`: one `fromjs` row with `ADMIN`, `DCPATH` and `OUTFMT`. Two things about it are worth knowing before you trust its result.
- **`ADMIN` must name a group the calling account is actually in.** The shipped value is `all-users`, the SASjs Server group that every logged-in user is a member of, which is what the shared estate needs. A name the account is not in - a Viya-style group such as `AllUsers`, or one estate's own group carried into another - still answers `SUCCESS` from the seed, and then every later `getdata`/`stagedata` call aborts with `<user> is not authorised to edit <lib>.<ds> in the <lib>.MPE_SECURITY table`. Check what the account belongs to with a call to `services/public/getgroups` before changing the value.
- **Each run creates a new library.** The response carries `STATUS`, `DCPATH` and `LIBREF` (for example `DC621835`), and the LIBREF is clock-derived, so a libref from an earlier run answers `dataset <lib>.<table> does not exist` - take the current one from the response.
- **`KEEPME` controls whether the seed can be repeated.** Without it `makedata` deletes itself and the configurator after a successful run (deliberately, so it cannot be executed twice over real data), and a second seed fails with `The Program at (<appLoc>/services/admin/makedata) does not exist` until you run `sasjs cbd` again. A harness that reseeds between runs sets `KEEPME` in the input row.
- **Seed once per deploy, not before every test run.** The suite builds its own `dctest` library from the seeded control tables; reseeding in between is wasted work, and there is no need to run `makedata` and `sasjs test` in the opposite order.
## Reading the results
Results land in `sas/sasjsresults/`: `testResults.json` (the counts worth quoting: `tests_with_results`, `tests_that_pass`, `failed_to_complete`, `completed_with_failures`), `testResults.csv` (one row per assertion, lower-case headers `test_target`, `test_loc`, `sasjs_test_id`, `test_suite_result`, `test_description`, `test_url`), and one SAS log per program under `logs/`. The CLI's closing lines are not the source of truth when they disagree with the JSON.
A current green run is 64 of 64 test programs with 241 assertions passing and nothing failing to complete.
`ERROR` lines in the per-program logs are expected and are not a failed run: the specs that assert a refusal get `ERROR: Execution canceled by an %ABORT CANCEL statement`, and the ones that feed bad payloads get the rejection message they are asserting on (`Invalid libds:...`, `Invalid libref:...`). Judge the run by the JSON counts, then by whether the errors you find are the ones the tests asked for.
## Deploying the frontend
The streamed web app is a separate artefact and `sasjs cbd` only streams it if `client/dist` exists:
1. `client/libraries/sheet-crypto.tgz.gpg` is an encrypted dependency (`file:libraries/sheet-crypto.tgz` in `client/package.json`) that is not on the public registry. Decrypt it with the passphrase the maintainer holds: `printf '%s' "$SHEET_PWD" | gpg --batch --yes --passphrase-fd 0 --output ./libraries/sheet-crypto.tgz --decrypt ./libraries/sheet-crypto.tgz.gpg`.
2. `client/.npmrc` is gitignored and needs `legacy-peer-deps=true` (the committed lock is written with it) plus the `@sheet` registry lines.
3. `cd client && npm ci && npm run build`.
4. The build carries the repo default appLoc, so point it at the deployment before deploying: rewrite `apploc="/Public/app/dc"` in `dist/index.html` to the target appLoc.
5. `cd ../sas && npx sasjs cbd -t <target>` - with `client/dist` present the log says `Copying web app files` and `Web app is available at <url>`.
Without step 1 the deploy prints `ERROR An error has occurred when compiling web app services ... webSourcePath: .../client/dist doesn't exist` and then deploys the services anyway, so the deploy looks successful while the app streams nothing. The AppStream URL is session-protected: an anonymous `curl` gets a `401`, so verify it with a logged-in cookie jar and fetch the `main.<hash>.js` the page names.
Never commit `client/.npmrc` or the decrypted tarball; both are ignored, and `git status` staying clean is the check.
## Reading a failed run
- `ERROR An error occurred while executing the request. Job location: <appLoc>/tests/...` means the response was not JSON, so the program broke before it could write results. Open `sasjsresults/logs/<suite>.log` and read the first `ERROR`; a single compile break cascades into hundreds of `ERROR 180-322: Statement is not valid` lines and the tail is noise.
- `WARNING: Apparent symbolic reference TESTLOC not resolved` in `testsetup.log`, next to a `404 Not Found` for `services/public/settings.sas`, is the missing seed described above, not a macro problem.
- `sasjs test` runs the DEPLOYED program, so after editing a service, hook or test, run `sasjs cbd` again before testing or you will debug code that is no longer in the tree.
- A deployment can succeed and still be missing its web app, its data or its configuration - check the thing you changed rather than the exit code.
+1 -1
View File
@@ -43,7 +43,7 @@ Data Controller must run entirely locally (offline / on-prem, no internet access
Agent-related content lives in `.agents/`: technical/agent-facing documentation goes in `.agents/docs/` (not `docs/`), and skills in `.agents/skills/`. When writing explanatory or technical docs about the codebase, put them in `.agents/docs/`.
Skills installed via `npx skills add` are tracked in `skills-lock.json` and live under `.agents/skills/`. Repo-local custom skills (e.g. `dc-sas` for Data Controller SAS development, `handsontable`, `hyperformula`) are kept here too. Avoid editing npx-managed skill files directly — use `npx skills update` instead.
Skills installed via `npx skills add` are tracked in `skills-lock.json` and live under `.agents/skills/`. Repo-local custom skills (e.g. `dc-sas` for Data Controller SAS development, `dc-estate-testing` for estate deploys and test runs, `handsontable`, `hyperformula`) are kept here too. Avoid editing npx-managed skill files directly — use `npx skills update` instead.
## Code comments and test names
+52
View File
@@ -1,3 +1,55 @@
# [7.21.0](https://git.datacontroller.io/dc/dc/compare/v7.20.0...v7.21.0) (2026-10-29)
### Bug Fixes
* **client:** parse word-date values from VA parameters ([405b28d](https://git.datacontroller.io/dc/dc/commit/405b28dcccfe5994521bf63f0270d4877f544294))
* **client:** parse word-date values from VA parameters ([ce5b1aa](https://git.datacontroller.io/dc/dc/commit/ce5b1aa96175d0fd1f048fea6d127cbc743804f2))
* **makedata:** name the sasjs server group in the estate seed ([eeffa92](https://git.datacontroller.io/dc/dc/commit/eeffa92f9e973b323171bbefe2ef1cbc27e3aee7))
### Features
* **derived:** apply the core field scan in the MPE_DERIVED hook ([a4a0f9c](https://git.datacontroller.io/dc/dc/commit/a4a0f9cb2f7c0b6252405b71df64aae912d1b075))
# [7.20.0](https://git.datacontroller.io/dc/dc/compare/v7.19.0...v7.20.0) (2026-10-28)
### Bug Fixes
* **ci:** install doxygen for the SAS docs job ([f58925b](https://git.datacontroller.io/dc/dc/commit/f58925b29956f1f753f0a367fdcd0afe7f3f51a0))
* **ci:** keep foreign release assets, install jq first, pin the CLI ([27686d3](https://git.datacontroller.io/dc/dc/commit/27686d3aad5e34ace59ca1b735a17cca1a4025e1))
* **ci:** publish against the tag this job cuts, not releases/latest ([8888734](https://git.datacontroller.io/dc/dc/commit/888873421c3aeaf312d5ec45e3c805317c66976a))
* **ci:** publish docs after the assets, with the webdoc app password ([dd482c4](https://git.datacontroller.io/dc/dc/commit/dd482c4069202a0c6b9983129c19efc09bd160a7))
* **ci:** publish the code documentation with surfer 7 and app passwords ([25ffa71](https://git.datacontroller.io/dc/dc/commit/25ffa71adddba5e334f4a8c166f284d1be498fcd))
* **ci:** publish the typedoc to the webdoc site, not the app default ([c5808a1](https://git.datacontroller.io/dc/dc/commit/c5808a16eba0d29562fcc76f13ac4a33ab8da0b7))
* **ci:** rebuild and upload release assets on demand ([6486c80](https://git.datacontroller.io/dc/dc/commit/6486c80ed4f368bbf116c0ecdecc0e5e5f02075f))
* **ci:** use the lockfile and the pinned CLI in the docs job ([af1813f](https://git.datacontroller.io/dc/dc/commit/af1813f829861df41c89b512c824e8d3ebac7a4b))
* **ci:** verify the forge certificate on the release API calls ([cf7b21a](https://git.datacontroller.io/dc/dc/commit/cf7b21a8b61ec3bf7ebe21968ccf13ade264bc64))
* **cypress:** hold download out too, and retry once in CI ([c97e886](https://git.datacontroller.io/dc/dc/commit/c97e8866db9ee404b2118d24baf53fc7058debef)), closes [#395](https://git.datacontroller.io/dc/dc/issues/395) [#398](https://git.datacontroller.io/dc/dc/issues/398)
* **cypress:** register the deleteFolder task, restore the spec order, hold metanav out ([c803773](https://git.datacontroller.io/dc/dc/commit/c803773d3c36476d8263f01fc3b8344aa3503d44)), closes [#395](https://git.datacontroller.io/dc/dc/issues/395)
* **deploy:** do not offer the deploy wizard inside a report ([80e8911](https://git.datacontroller.io/dc/dc/commit/80e8911af8857fa2c417f028cd59b64a87233b9e))
* **deploy:** keep the app query when opening the standalone URL ([e51953c](https://git.datacontroller.io/dc/dc/commit/e51953ca8a6734e1649bad68cf4ad2057d677334))
* **deploy:** one definition of the standalone URL, and test the guards ([459ff2d](https://git.datacontroller.io/dc/dc/commit/459ff2da371375e996839d69deb40f681c356f1d)), closes [#TCS](https://git.datacontroller.io/dc/dc/issues/TCS)
* **deploy:** strip the embed parameters by name, and scope the timezone runs to the VA specs ([f7e7538](https://git.datacontroller.io/dc/dc/commit/f7e753863b61d8545bd0777d80267661b3c58670))
* **embed:** build quarter ends from the calendar, not minus 24 hours ([7e3045a](https://git.datacontroller.io/dc/dc/commit/7e3045a21b67053b76a81bfe836e31571ccbffbf))
* **embed:** defer the column check until metadata loads, report AND-ed pairs ([2623a3f](https://git.datacontroller.io/dc/dc/commit/2623a3fb4ba3bf19c00636b1717b53d5a2d54f3a))
* **embed:** log a parse-time mapping problem once ([0011f7e](https://git.datacontroller.io/dc/dc/commit/0011f7e933d85be8e33debe3a6ecb9000edaabe3))
* **embed:** make the va_parmap diagnostics say what they mean ([541a61b](https://git.datacontroller.io/dc/dc/commit/541a61ba08bc7256a911c50de7078859cc46c84e))
* **embed:** match a mapping on the parameter label only, and warn only when the values differ ([e9216b0](https://git.datacontroller.io/dc/dc/commit/e9216b08956d5599662259a2b05cc1524c458df5))
* **embed:** never build a clause from a value that did not convert, and say so in the banner ([bd3c220](https://git.datacontroller.io/dc/dc/commit/bd3c220af2678a2a3f8394fe3084ee03c23d0b15))
* **embed:** one parameter maps one column, and report typos on both sides ([f83dc35](https://git.datacontroller.io/dc/dc/commit/f83dc359e258f5d1cb9fc49f0eaab9748995d823))
* **embed:** record the parameter label only, so a name-keyed va_parmap is reported ([aa91daf](https://git.datacontroller.io/dc/dc/commit/aa91daf9f78c00cea486d0f885f8e634366bd1a0))
* **embed:** resolve transform values as local dates and give ranges whole days ([10157c1](https://git.datacontroller.io/dc/dc/commit/10157c11595685efdc206fe7b16a5cafa098f26d))
* **embed:** warn when a va_parmap entry is ignored ([411f6c3](https://git.datacontroller.io/dc/dc/commit/411f6c3354b7fbfec11fef3b174ae1f8066c92ba))
* **makedata:** pair the %else of the keepme block so it parses in open code ([c0ea810](https://git.datacontroller.io/dc/dc/commit/c0ea810766ebfda5ae5358ab14146fbcc07fcc8e))
### Features
* **embed:** map VA parameters to DC columns from the URL ([15a1f34](https://git.datacontroller.io/dc/dc/commit/15a1f34d5263503cac83b4fe4e52482777dc2f1f))
* **embed:** resolve va_parmap values through named date anchors ([88010c3](https://git.datacontroller.io/dc/dc/commit/88010c3980c2ea1c39bbb024ed32d63e495e61f0))
# [7.19.0](https://git.datacontroller.io/dc/dc/compare/v7.18.0...v7.19.0) (2026-10-28)
+1 -1
View File
@@ -8,7 +8,7 @@ This file is the shared glossary and orientation for the repo. When your output
- `client/` - the Angular frontend (TypeScript). Uses [Handsontable](.agents/skills/handsontable/SKILL.md) for the editable grid and [HyperFormula](.agents/skills/hyperformula/SKILL.md) for Excel-formula support. Lint with `npm run lint:check` from `client/`.
- `sas/` - the SAS backend. Services, macros, hooks and tests live under `sas/sasjs/**`; `sas/sasjsbuild/` is generated output (never hand-edit). Lint `.sas` files with `sasjs lint` from `sas/`.
- `.agents/docs/` - technical deep-dives (see below). `.agents/skills/` - task skills (handsontable, hyperformula, dc-sas for Data Controller SAS development, plus npx-installed sasjs/ sas skills).
- `.agents/docs/` - technical deep-dives (see below). `.agents/skills/` - task skills (handsontable, hyperformula, dc-sas for Data Controller SAS development, dc-estate-testing for deploying to and testing on the shared estate, plus npx-installed sasjs/ sas skills).
## Roles
@@ -8,20 +8,6 @@
@if (autodeployDone) {
<p class="m-0 align-self-start">Done</p>
<hr class="w-100" />
@if (autoDeployStatus.deployServicePack !== null) {
<div class="deploy-status-row">
@if (autoDeployStatus.deployServicePack === true) {
<clr-icon
class="deploy-success"
shape="success-standard"
></clr-icon>
}
@if (!autoDeployStatus.deployServicePack === false) {
<clr-icon class="deploy-error" shape="times-circle"></clr-icon>
}
<p>Deploy SAS Jobs</p>
</div>
}
@if (autoDeployStatus.runMakeData !== null) {
<div class="deploy-status-row">
@if (autoDeployStatus.runMakeData) {
@@ -33,16 +19,13 @@
@if (autoDeployStatus.runMakeData === false) {
<clr-icon class="deploy-error" shape="times-circle"></clr-icon>
}
<p>Create database</p>
<p>Create control library</p>
</div>
}
<hr class="w-100" />
<div class="buttons">
<button (click)="navigateToHome()" class="btn btn-primary mt-15 mr-0">
@if (
autoDeployStatus.deployServicePack === false ||
autoDeployStatus.runMakeData === false
) {
@if (autoDeployStatus.runMakeData === false) {
<clr-icon class="deploy-error" shape="times-circle"></clr-icon>
}
LAUNCH
@@ -78,9 +61,9 @@
<h4 class="text-center my-15">Viya Deploy</h4>
<p class="deploy-intro">
Configure and deploy Data Controller to your Viya environment. Deployment
creates the backend SAS services and the control library used by the
application.
Configure Data Controller for your Viya environment. The frontend and services
are already in place; this runs the one-off job that creates the control
library and its configuration data, then removes itself.
</p>
<hr />
@@ -248,52 +231,6 @@
</div>
}
<!-- Keeping this for a reference in case future VIYA changes and starts allowing separate backend and frontend) -->
<!-- <clr-checkbox-wrapper>
<input
clrCheckbox
[(ngModel)]="recreateDatabase"
(click)="recreateDatabaseClicked($event)"
type="checkbox"
checked
/>
<label>Recreate database</label>
</clr-checkbox-wrapper> -->
<hr />
<button
(click)="runAutoDeploy()"
class="btn-autodeploy btn btn-primary d-inline-block mr-10"
>
Deploy
</button>
<!-- Keeping this for a reference in case future VIYA changes and starts allowing separate backend and frontend) -->
<!-- <button
(click)="executeJson()"
class="btn-autodeploy btn btn-primary d-inline-block mr-10"
[disabled]="!jsonFile"
>
Deploy {{ !jsonFile ? '(json file is not available)' : '' }}
</button> -->
<!-- <button
(click)="uploadJsonAuto.click()"
class="btn-autodeploy btn btn-primary d-inline-block mr-10"
>
Upload different file to deploy
</button>
<input
#uploadJsonAuto
type="file"
hidden
(click)="clearUploadInput($event)"
(change)="onJsonFileChange($event)"
/> -->
<clr-modal [(clrModalOpen)]="recreateDatabaseModal" [clrModalClosable]="false">
<h3 class="modal-title">Warning</h3>
<div class="modal-body">
@@ -51,16 +51,10 @@ export class AutomaticComponent implements OnInit {
public selectedComputeContext: string = ''
public makeDataResponse: string = ''
public jsonFile: any = null
public autodeploying: boolean = false
public autodeployDone: boolean = false
public recreateDatabaseModal: boolean = false
public isSubmittingJson: boolean = false
public isJsonSubmitted: boolean = false
/**
* Default was `false` when deploy was done with frontend and backend separately.
* Now we are using only streaming app, so we always want to recreate database (makedata)
*/
/** Whether `makedata` runs as part of the deploy. */
public recreateDatabase: boolean = true
public createDatabaseLoading: boolean = false
public adminGroupsLoading: boolean = false
@@ -86,18 +80,14 @@ export class AutomaticComponent implements OnInit {
public currentUserInfo: ViyaApiCurrentUser | null = null
public computeContexts: ComputeContextItem[] = []
/** autoDeployStatus
* This object presents the status for two steps that we have for deploy.
* `deployServicePack` - Creating services based on `viya.json`
* `runMakeData` - Running `makedata` service
* If any of them is `null` or `false` that means step failed
* and will be shown to user on deploy done modal.
/**
* Status of the one step this screen runs: `makedata`, which creates the
* control library and its configuration data and then deletes itself.
* `null` means it has not run yet, `false` that it failed.
*/
public autoDeployStatus: {
deployServicePack: any
runMakeData: any
} = {
deployServicePack: null,
runMakeData: null
}
@@ -420,56 +410,14 @@ export class AutomaticComponent implements OnInit {
}
/**
* Executes sas.json file to deploy the backend
* Method will first try to run the `auto deploy`
* If that fails the rest of the code is ignored.
* If request is successfull, method will continue to try
* to create database if checkbox is toggled on
* Runs the deployment. The control library and its configuration data are
* created by `makedata`, which self-deletes once it has run; the frontend and
* services arrived with the sasjs deploy, so there is nothing else to deploy
* from here.
*/
public async executeJson() {
this.isSubmittingJson = true
try {
let uploadJsonFile = await this.sasJs.deployServicePack(
this.jsonFile,
this.dcAdapterSettings?.appLoc,
undefined,
undefined,
true
)
this.autoDeployStatus.deployServicePack = true
this.isJsonSubmitted = true
} catch (ex: any) {
let textEx = ''
if (typeof ex.message !== 'string') {
textEx = JSON.stringify(ex).replace(/\\/gm, '')
} else {
textEx = ex.message
}
this.autoDeployStatus.deployServicePack = false
this.eventService.showInfoModal(
'Deploy error',
`Exception: \n ${textEx !== '' ? textEx : ex}`
)
this.autodeploying = false
this.autodeployDone = false
return
}
this.isSubmittingJson = false
}
public async runAutoDeploy(executeJson: boolean = false) {
public async runAutoDeploy() {
if (!this.deployInNewWindow) this.autodeploying = true
if (executeJson) {
this.executeJson()
}
if (this.recreateDatabase) {
this.createDatabase()
} else {
@@ -708,22 +656,12 @@ export class AutomaticComponent implements OnInit {
this.deployService.downloadFile(content, filename, extension)
}
public async onJsonFileChange(event: any) {
let file = event.target.files[0]
this.jsonFile = await this.deployService.readFile(file)
}
public recreateDatabaseClicked(event: Event) {
;(<HTMLInputElement>event.target).checked === true
? (this.recreateDatabaseModal = true)
: ''
}
public clearUploadInput(event: Event) {
this.deployService.clearUploadInput(event)
}
public openSasRequestsModal() {
this.eventService.openRequestsModal()
}
+10 -1
View File
@@ -3647,7 +3647,16 @@ export class EditorComponent
if (clauses.length === 0) return false
try {
const res: any = await this.sasStoreService.saveQuery(libds, clauses)
// A `va_parmap` entry is a column the report author typed, so this
// request can legitimately fail: the table may not have that column, and
// validatefilter aborts on an unknown variable (dc/dc#404). Attempting it
// is still worth it - when the mapping is right the very first getdata is
// already filtered - but the failure must not stop the page. An abort
// modal here covers an editor that is about to load fine; the mapping
// banner names the bad entry once the real metadata arrives.
const res: any = await this.sasStoreService.saveQuery(libds, clauses, {
suppressErrorAbortModal: true
})
const id = res?.result?.[0]?.FILTER_RK
const table = res?.result?.[0]?.FILTER_TABLE
if (id === undefined || table === undefined) return false
-18
View File
@@ -26,22 +26,4 @@ export class DeployService {
document.body.removeChild(downloadLink)
}
}
public readFile(file: File): Promise<any> {
return new Promise((resolve, reject) => {
let JsonFileReader = new FileReader()
JsonFileReader.onload = () => {
if (JsonFileReader.result) {
resolve(JSON.parse(JsonFileReader.result.toString()))
}
}
JsonFileReader.readAsText(file)
})
}
public clearUploadInput(event: Event) {
;(<HTMLInputElement>event.target).value = ''
}
}
@@ -116,6 +116,53 @@ describe('SasStoreService', () => {
expect(seen.length).toBe(0)
})
})
describe('saveQuery', () => {
const capture = (adapterResponse: any = { result: [] }) => {
const calls: any[][] = []
const store = new SasStoreService(
{
request: (...args: any[]) => {
calls.push(args)
return Promise.resolve({ adapterResponse })
}
} as any,
{} as any,
{} as any
)
return { store, calls }
}
it('posts the filter to public/validatefilter', async () => {
const { store, calls } = capture()
await store.saveQuery('DC.MY_TABLE', [
{ VARIABLE_NM: 'A', OPERATOR_NM: '=', RAW_VALUE: "'x'" } as any
])
expect(calls[0][0]).toBe('public/validatefilter')
expect(calls[0][1].iwant[0].filter_table).toBe('DC.MY_TABLE')
expect(calls[0][1].filterquery.length).toBe(1)
})
it('passes wrapper options through, so a rejected filter can stay quiet', async () => {
// The VA deferred load attempts a filter built from a va_parmap entry the
// table may not have, and must not cover a page that loads fine with an
// abort modal (dc/dc#404).
const { store, calls } = capture()
await store.saveQuery('DC.MY_TABLE', [], {
suppressErrorAbortModal: true
})
expect(calls[0][3]).toEqual({ suppressErrorAbortModal: true })
})
it('sends no wrapper options when the caller gives none', async () => {
const { store, calls } = capture()
await store.saveQuery('DC.MY_TABLE', [])
expect(calls[0][3]).toBeUndefined()
})
})
})
/**
+20 -2
View File
@@ -8,6 +8,7 @@ import { HelperService } from './helper.service'
import { QueryClause } from '../models/TableData'
import { globals } from '../_globals'
import { FilterClause, FilterGroup, FilterQuery } from '../models/FilterQuery'
import { RequestWrapperOptions } from '../models/request-wrapper/RequestWrapperOptions'
import {
$DataFormats,
EditorsGetDataSASResponse,
@@ -395,13 +396,30 @@ export class SasStoreService {
).adapterResponse
}
public async saveQuery(libds: string, filterQuery: QueryClause[]) {
/**
* Persists a filter and returns its FILTER_RK.
*
* `wrapperOptions` reaches the request wrapper, so a caller that can live with
* a rejected filter (the VA deferred load, whose `va_parmap` entry may name a
* column the table does not have) can suppress the abort modal that would
* otherwise cover a page which is about to load perfectly well.
*/
public async saveQuery(
libds: string,
filterQuery: QueryClause[],
wrapperOptions?: RequestWrapperOptions
) {
let tables = {
iwant: [{ filter_table: libds }],
filterquery: filterQuery
}
const res = await this.sasService.request('public/validatefilter', tables)
const res = await this.sasService.request(
'public/validatefilter',
tables,
undefined,
wrapperOptions
)
this.filter.next(res)
@@ -131,6 +131,41 @@ describe('VaFilterService', () => {
expect(build([{ label: 'SOME_DATE', value: '2025/03/31' }])).toHaveSize(1)
})
it('parses word dates, in either order and with an optional comma', () => {
const expected = "'01JAN2026'd"
// The reported shape: a date parameter displayed as WORDDATE.
expect(
build([{ label: 'SOME_DATE', value: 'January 1, 2026' }])[0].RAW_VALUE
).toBe(expected)
expect(
build([{ label: 'SOME_DATE', value: 'Jan 1, 2026' }])[0].RAW_VALUE
).toBe(expected)
expect(
build([{ label: 'SOME_DATE', value: 'Jan. 1, 2026' }])[0].RAW_VALUE
).toBe(expected)
expect(
build([{ label: 'SOME_DATE', value: '1 January 2026' }])[0].RAW_VALUE
).toBe(expected)
expect(
build([{ label: 'SOME_DATE', value: '1-Jan-2026' }])[0].RAW_VALUE
).toBe(expected)
})
it('matches a spelled-out month case-insensitively, and keeps DATE9 working', () => {
expect(
build([{ label: 'SOME_DATE', value: 'DECEMBER 31, 2025' }])[0].RAW_VALUE
).toBe("'31DEC2025'd")
expect(
build([{ label: 'SOME_DATE', value: '01AUG1962' }])[0].RAW_VALUE
).toBe("'01AUG1962'd")
})
it('still drops a month-name value that has no day', () => {
// A month and year alone is not a date - accepted shapes always carry a
// day, so this must not silently become the first of the month.
expect(build([{ label: 'SOME_DATE', value: 'January 2026' }])).toEqual([])
})
it('resolves ambiguous slash dates day-first (never US), disambiguating by value', () => {
const iso = build([{ label: 'SOME_DATE', value: '2025-03-31' }])[0]
.RAW_VALUE
+35 -7
View File
@@ -707,14 +707,42 @@ export class VaFilterService {
return this.parseMonthNameDate(value) ?? this.parseNumericDate(value, order)
}
/** DATE9 "ddMMM[yy]yy" (e.g. 01AUG1962, 01AUG62) -> local Date, or null. */
/**
* Month-name date shapes -> local Date, or null. Three families:
* - DATE9 "ddMMM[yy]yy" (e.g. 01AUG1962, 01AUG62);
* - word date with the month first, comma optional: "January 1, 2026",
* "Jan 1 2026", "Jan. 1, 2026";
* - word date with the day first: "1 January 2026", "1-Jan-2026".
*
* A date parameter's display format decides which of these the report author
* sees, and VA sends the formatted text, so a report using WORDDATE (or a
* long month name in any order) has to parse here rather than being dropped.
* The month name may be abbreviated or spelled out; only its first three
* letters are matched.
*/
private parseMonthNameDate(value: string): Date | null {
const m = /^(\d{1,2})([A-Za-z]{3})(\d{2}|\d{4})$/.exec(value)
if (!m) return null
const [, day, monthAbbr, year] = m
const monthIndex = VaFilterService.SAS_MONTHS.indexOf(
monthAbbr.toUpperCase()
)
const monthByName = (name: string): number =>
VaFilterService.SAS_MONTHS.indexOf(
name.replace(/\./g, '').slice(0, 3).toUpperCase()
)
const date9 = /^(\d{1,2})([A-Za-z]{3})(\d{2}|\d{4})$/.exec(value)
if (date9) {
const [, day, monthAbbr, year] = date9
const monthIndex = monthByName(monthAbbr)
if (monthIndex < 0) return null
return this.buildDate(this.fullYear(year), monthIndex + 1, Number(day))
}
const monthFirst =
/^([A-Za-z]{3,9})\.?[\s-]+(\d{1,2}),?[\s-]+(\d{2}|\d{4})$/.exec(value)
const dayFirst =
/^(\d{1,2})[\s-]+([A-Za-z]{3,9})\.?,?[\s-]+(\d{2}|\d{4})$/.exec(value)
const name = monthFirst ? monthFirst[1] : dayFirst ? dayFirst[2] : null
if (!name) return null
const day = monthFirst ? monthFirst[2] : dayFirst![1]
const year = monthFirst ? monthFirst[3] : dayFirst![3]
const monthIndex = monthByName(name)
if (monthIndex < 0) return null
return this.buildDate(this.fullYear(year), monthIndex + 1, Number(day))
}
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "dcfrontend",
"version": "7.19.0",
"version": "7.21.0",
"description": "Data Controller",
"devDependencies": {
"@nogoo9/gitleaks": "8.30.1-post.2",
+1 -1
View File
@@ -1,7 +1,7 @@
{
"fromjs": [
{
"ADMIN": "AllUsers",
"ADMIN": "all-users",
"DCPATH": "/tmp/dcdata",
"OUTFMT": "JSON"
}
@@ -13,10 +13,12 @@
// The rules mirror services/hooks/mpe_derived_postedit.sas, in its order:
// normalise the incoming rows, reject a duplicate name and a bad field, then
// check the dependency graph that the resulting configuration would have (the
// incoming rows plus the current rows they do not replace). As on the server,
// the collision between VAR_NM and a physical column is left to the read-time
// engine, which runs with the target library assigned - the hook does not have
// it here either.
// incoming rows plus the current rows they do not replace). The field rules
// are core's mp_derivedvarcheck, which the server hook calls and the read-time
// engine runs too - this file spells them out a second time, so a change there
// has to be mirrored here. As on the server, the collision between VAR_NM and
// a physical column is left to the read-time engine, which runs with the target
// library assigned - the hook does not have it here either.
// ─────────────────────────────────────────────────────────────────────────────
;(function mpeDerivedPostedit() {
@@ -83,9 +85,45 @@
seen.add(key)
}
// ─── Field level checks, the rules the engine applies at read time ─────────
// C0 controls and DEL: what a single-line label cannot carry
// ─── Field level checks ────────────────────────────────────────────────────
// Two passes, in the order the SAS hook runs them. The first is the fields
// this hook owns - the control table key, which the loader normalises but
// does not require, and ACTIVE, the picker pair the EDIT screen offers.
for (const row of rows) {
if (isDeleted(row)) continue
const name = field(row, 'VAR_NM')
let reason = ''
if (!field(row, 'LIBREF') || !field(row, 'DSN'))
reason = 'LIBREF and DSN are required'
// ACTIVE is a picker in the EDIT screen (HARDSELECT over the
// MPE_SELECTBOX pair) - the same pair the hook enforces
else if (!['0', '1'].includes(String(colVal(row, 'ACTIVE')).trim()))
reason = 'ACTIVE must be 1 or 0'
if (reason) return reject(`MPE_DERIVED: ${name}: ${reason}`)
}
// The second pass is the rules for the derived variable itself, which are
// core's (mp_derivedvarcheck): the engine runs the same macro at read time,
// so a definition is measured against one rule list wherever it is checked.
// C0 controls and DEL: what a label cannot carry once a line break has been
// flattened to a space, which is what the engine does with it
const CONTROL_CHARS = /[\u0000-\u001F\u007F]/
const flatten = (v) => v.replace(/[\r\n]+/g, ' ')
const AUTOMATIC_VARS = ['_N_', '_ERROR_', '_IORC_']
// the capabilities refused by name - a function name cannot be assembled at
// run time, so naming them is a control the expression cannot evade
const DENY_FUNCS = new RegExp(
'\\b(DOSUBL|JAVAOBJ|JAVAMETA|SYMGET|SYMPUTX?|SYSGET|GETOPTION' +
'|PATHNAME|FILEEXIST|FOPEN|FREAD|FGET|FCLOSE' +
'|DOPEN|DREAD|DNUM|MOPEN|SLEEP' +
'|FILENAME|FDELETE|FCOPY|DCREATE|DLCREATE)\\s*\\('
)
// quoted literals are data, not calls
const unquoted = (v) =>
v.replace(/'([^']|'')*'/g, ' ').replace(/"([^"]|"")*"/g, ' ')
// true when the value carries a character outside the allowed set
const hasInvalidChar = (value, chars) =>
@@ -93,8 +131,6 @@
for (const row of rows) {
if (isDeleted(row)) continue
const libref = field(row, 'LIBREF')
const dsn = field(row, 'DSN')
const name = field(row, 'VAR_NM')
const type = field(row, 'VAR_TYPE')
const len = Number(colVal(row, 'VAR_LEN'))
@@ -103,19 +139,16 @@
const logic = strip(colVal(row, 'VAR_LOGIC'))
let reason = ''
if (!libref || !dsn) reason = 'LIBREF and DSN are required'
else if (!name) reason = 'VAR_NM is required'
if (!name) reason = 'VAR_NM is required'
else if (hasInvalidChar(name, NAME_CHARS))
reason = 'VAR_NM must be a valid SAS name'
else if ('0123456789'.includes(name[0]))
reason = 'VAR_NM must not start with a digit'
else if (AUTOMATIC_VARS.includes(name))
reason = 'VAR_NM is an automatic variable of the data step'
else if (name === DELETE_MARK) reason = 'VAR_NM is reserved'
else if (type !== 'C' && type !== 'N')
reason = 'VAR_TYPE must be C or N'
// ACTIVE is a picker in the EDIT screen (HARDSELECT over the
// MPE_SELECTBOX pair) - the same pair mpe_derived_check.sas enforces
else if (!['0', '1'].includes(String(colVal(row, 'ACTIVE')).trim()))
reason = 'ACTIVE must be 1 or 0'
else if (!Number.isInteger(len) || len < 1)
reason = 'VAR_LEN must be a positive integer'
else if (type === 'C' && len > 32767)
@@ -132,20 +165,24 @@
reason = 'VAR_FMT must be a format name, without ; & % or /* */'
else if (hasInvalidChar(fmt, FMT_CHARS))
reason = 'VAR_FMT must be a format name'
// free text, in any script - what is refused is anything that cannot be
// a single-line column label, and a value the engine would cut to 256
else if (CONTROL_CHARS.test(label))
reason =
'VAR_LABEL must be plain text - no line breaks or control characters'
// free text, in any script - what is refused is a control character,
// and a value the engine would cut to 256
else if (CONTROL_CHARS.test(flatten(label)))
reason = 'VAR_LABEL must be plain text - no control characters'
else if ([...label].length > 256)
reason = 'VAR_LABEL exceeds 256 characters'
else if (!logic) reason = 'VAR_LOGIC is required'
else if (logic.length > 4000)
reason = 'VAR_LOGIC exceeds 4000 characters'
else if (logic.includes(';'))
reason = 'VAR_LOGIC must be a single expression - remove the ;'
else if (logic.includes('&') || logic.includes('%'))
reason = 'VAR_LOGIC must not contain macro triggers'
else if (logic.includes('/*') || logic.includes('*/'))
reason = 'VAR_LOGIC must not contain comments'
else if (DENY_FUNCS.test(unquoted(logic).toUpperCase()))
reason =
'VAR_LOGIC may not call a function that submits code or reaches the file system, the environment or the macro symbol table'
if (reason) return reject(`MPE_DERIVED: ${name}: ${reason}`)
}
+4 -4
View File
@@ -7,7 +7,7 @@
"name": "dc-sas",
"dependencies": {
"@sasjs/cli": "6.2.0",
"@sasjs/core": "6.1.0"
"@sasjs/core": "6.1.4"
}
},
"node_modules/@asamuzakjp/css-color": {
@@ -235,9 +235,9 @@
}
},
"node_modules/@sasjs/core": {
"version": "6.1.0",
"resolved": "https://registry.npmjs.org/@sasjs/core/-/core-6.1.0.tgz",
"integrity": "sha512-U0Ekq/TXVMquYV+AG6urBQSUgqUvV37yduufy68QQVgDj3PJiLUJDcOXKOGbhWGD+Nv877VuPCviGYqrqbjXvg==",
"version": "6.1.4",
"resolved": "https://registry.npmjs.org/@sasjs/core/-/core-6.1.4.tgz",
"integrity": "sha512-vIF2LIftxPWg4CvO3i5raEfrHFHY/6S+MqVGpjRR71dgjLRJEYNyPJgGqD6BCZwHv5V6/2nDnBi/BIIMiLXfzA==",
"license": "MIT"
},
"node_modules/@sasjs/lint": {
+1 -1
View File
@@ -34,7 +34,7 @@
"private": true,
"dependencies": {
"@sasjs/cli": "6.2.0",
"@sasjs/core": "6.1.0"
"@sasjs/core": "6.1.4"
},
"overrides": {
"nanoid": "3.3.18"
+13 -3
View File
@@ -49,9 +49,19 @@
being maintained. A cycle is rejected.
Semicolons, macro triggers and comments are rejected, so a row can only ever
be an expression. A derived variable may not reuse the name of a physical
column, and may not reference itself: an intra-row calculation would imply a
retained value, which depends on the order rows happen to arrive in.
be an expression, and an expression that would let the row submit code of its
own - DOSUBL, or a function reaching the file system, the environment or the
macro symbol table - is refused by name.
The field level rules themselves are core's: mp_derivedvarcheck scans the
definitions and mp_derivedvars calls it, which is the same macro the MPE_DERIVED
post edit hook runs at submit time (services/hooks/mpe_derived_postedit). A
row is therefore measured against one rule list, whether it is being staged
or read.
A derived variable may not reuse the name of a physical column, and may not
reference itself: an intra-row calculation would imply a retained value,
which depends on the order rows happen to arrive in.
<h4> SAS Macros </h4>
@li mp_derivedvars.sas
-75
View File
@@ -1,75 +0,0 @@
/**
@file
@brief Field level checks for a staged MPE_DERIVED submission
@details Writes one row per staged row that breaks a rule, with the reason in
REASON, so a caller can report the first failure and abort. Rows marked for
deletion are skipped - they are being removed, so their fields are not loaded.
The rules are the ones %mpe_derived applies at read time, so a submission
cannot stage a row the engine would then refuse.
@param inds the staged rows, already normalised by the caller
@param outds the dataset to create: one row per failing row, plus REASON
@author 4GL Apps Ltd
@copyright 4GL Apps Ltd. This code may only be used within Data Controller
and may not be re-distributed or re-sold without the express permission of
4GL Apps Ltd.
**/
%macro mpe_derived_check(inds=work.staging_ds, outds=work.derived_check);
data &outds;
set &inds;
length reason $200;
if upcase(_____DELETE__THIS__RECORD_____)='YES' then delete;
call missing(reason);
if missing(libref) or missing(dsn) then reason='LIBREF and DSN are required';
else if var_nm='' then reason='VAR_NM is required';
else if verify(strip(var_nm),'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_')>0
then reason='VAR_NM must be a valid SAS name';
else if verify(substr(var_nm,1,1),'0123456789')=0
then reason='VAR_NM must not start with a digit';
else if var_nm='_____DELETE__THIS__RECORD_____'
then reason='VAR_NM is reserved';
else if var_type not in ('C','N')
then reason='VAR_TYPE must be C or N';
/* ACTIVE is a picker in the EDIT screen (HARDSELECT over the MPE_SELECTBOX
pair) - the same pair is enforced here, so a submission that bypasses the
grid cannot land a value the screen would never offer. */
else if missing(active) or active not in (0,1)
then reason='ACTIVE must be 1 or 0';
else if missing(var_len) or var_len ne int(var_len) or var_len<1
then reason='VAR_LEN must be a positive integer';
else if var_type='C' and var_len>32767
then reason='VAR_LEN exceeds 32767 for a character variable';
else if var_type='N' and var_len>8
then reason='VAR_LEN exceeds 8 for a numeric variable';
else if index(var_fmt,';')>0 or indexc(var_fmt,'&%')>0
or index(var_fmt,'/*')>0 or index(var_fmt,'*/')>0
then reason='VAR_FMT must be a format name, without ; & % or /* */';
else if verify(strip(var_fmt),'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_$.')>0
then reason='VAR_FMT must be a format name';
/* VAR_LABEL is free text: any character is allowed, in any script, and the
quotes in it are escaped by %mp_derivedvars when the attrib statement is
built. What is refused is anything that cannot be a single-line column
label - a control character or a line break - and a value long enough to
be cut down silently, since the engine truncates the label to 256. */
else if anycntrl(var_label)>0
then reason='VAR_LABEL must be plain text - no line breaks or control characters';
else if klength(var_label)>256
then reason='VAR_LABEL exceeds 256 characters';
else if missing(var_logic)
then reason='VAR_LOGIC is required';
else if index(var_logic,';')>0
then reason='VAR_LOGIC must be a single expression - remove the ;';
else if indexc(var_logic,'&%')>0
then reason='VAR_LOGIC must not contain macro triggers';
else if index(var_logic,'/*')>0 or index(var_logic,'*/')>0
then reason='VAR_LOGIC must not contain comments';
if not missing(reason) then output;
run;
%mend mpe_derived_check;
-194
View File
@@ -1,194 +0,0 @@
/**
@file
@brief testing the mpe_derived_check macro
@details Checks each field level rule the MPE_DERIVED post edit hook applies
to a staged submission. The macro is exercised directly so the rules can be
probed with values a service cannot carry: VAR_LABEL is varchar(256) in the
table and a staging dataset declares the same length, so a longer label is
truncated before any check would see it.
<h4> SAS Macros </h4>
@li mf_nobs.sas
@li mp_assert.sas
@li mpe_derived_check.sas
@author 4GL Apps Ltd
@copyright 4GL Apps Ltd. This code may only be used within Data Controller
and may not be re-distributed or re-sold without the express permission of
4GL Apps Ltd.
**/
/**
* Test 1 - one row per failing rule, and only failing rows
*/
data work.staging_ds;
length libref $8 dsn $32 var_nm $32 var_type $1 var_len 8
var_fmt $32 var_label $4000 var_deps $4000 var_logic $4000
active 8 _____DELETE__THIS__RECORD_____ $3;
libref='DCTEST';
dsn='DC_DERIV';
/* valid unless a row says otherwise - ACTIVE is mandatory in the screen and
the check enforces the same pair here */
active=1;
_____DELETE__THIS__RECORD_____='No';
/* meets every rule */
var_nm='GOOD';
var_type='N';
var_len=8;
var_fmt='';
var_label='a plain label';
var_deps='';
var_logic='1';
output;
/* delete marked rows are being removed, so their fields are not loaded */
var_nm='DELME';
var_label='first line'||'0A'x||'second line';
_____DELETE__THIS__RECORD_____='Yes';
output;
_____DELETE__THIS__RECORD_____='No';
/* free text is allowed, including quotes and punctuation */
var_nm='FREETXT';
var_label='It''s "quoted", plain text: 100 percent';
output;
/* a line break cannot be a single line column label */
var_nm='LBLBREAK';
var_label='first line'||'0A'x||'second line';
output;
/* longer than the table holds, so it would be cut down silently */
var_nm='LBLLONG';
var_label=repeat('x',300);
output;
/* a format name, not an expression */
var_nm='BADFMT';
var_fmt='DATE9;';
var_label='fmt';
output;
/* one expression, no statement separator, no macro triggers */
var_nm='BADLOGIC';
var_fmt='';
var_logic='x; y';
output;
/* C or N */
var_nm='BADTYPE';
var_logic='1';
var_type='X';
output;
/* ACTIVE is a picker in the screen: only 1 or 0 */
var_nm='BADACTIVE';
var_type='C';
var_logic='1';
active=2;
output;
stop;
run;
%mpe_derived_check(inds=work.staging_ds, outds=work.derived_check)
/**
* Test 2 - the reported rows, rule by rule
*/
data work.expected;
length var_nm $32 reason $200;
var_nm='LBLBREAK';
reason='VAR_LABEL must be plain text - no line breaks or control characters';
output;
var_nm='LBLLONG';
reason='VAR_LABEL exceeds 256 characters';
output;
var_nm='BADFMT';
reason='VAR_FMT must be a format name, without ; & % or /* */';
output;
var_nm='BADLOGIC';
reason='VAR_LOGIC must be a single expression - remove the ;';
output;
var_nm='BADTYPE';
reason='VAR_TYPE must be C or N';
output;
var_nm='BADACTIVE';
reason='ACTIVE must be 1 or 0';
output;
run;
proc sql;
create table work.cmp as
select e.var_nm
,case when a.reason=e.reason then 1 else 0 end as same
from work.expected e
left join work.derived_check a
on upcase(strip(a.var_nm))=upcase(e.var_nm);
quit;
data _null_;
set work.cmp;
call symputx(cats('rule_',var_nm),same,'l');
run;
%mp_assert(
iftrue=(%mf_nobs(work.derived_check)=6),
desc=%str(Checking only the six failing rows are reported),
outds=work.test_results
)
%mp_assert(
iftrue=(&rule_LBLBREAK=1),
desc=%str(Checking a label with a line break is rejected),
outds=work.test_results
)
%mp_assert(
iftrue=(&rule_LBLLONG=1),
desc=%str(Checking a label over 256 characters is rejected),
outds=work.test_results
)
%mp_assert(
iftrue=(&rule_BADFMT=1),
desc=%str(Checking a format carrying a statement separator is rejected),
outds=work.test_results
)
%mp_assert(
iftrue=(&rule_BADLOGIC=1),
desc=%str(Checking a logic value with a statement separator is rejected),
outds=work.test_results
)
%mp_assert(
iftrue=(&rule_BADTYPE=1),
desc=%str(Checking a VAR_TYPE other than C or N is rejected),
outds=work.test_results
)
%mp_assert(
iftrue=(&rule_BADACTIVE=1),
desc=%str(Checking an ACTIVE outside 1 or 0 is rejected),
outds=work.test_results
)
/**
* Test 3 - a row that meets every rule, a free text label, and a delete
* marked row are all left alone
*/
proc sql noprint;
select count(*) into: notchecked
from work.derived_check
where upcase(strip(var_nm)) in ('GOOD','FREETXT','DELME');
quit;
%mp_assert(
iftrue=(&notchecked=0),
desc=%str(Checking a valid row, a free text label and a delete marked row are not reported),
outds=work.test_results
)
+3 -1
View File
@@ -2355,7 +2355,9 @@ insert into &lib..MPE_VALIDATIONS set
,some_char="&x bottles of beer on the wall"
,some_dropdown='Option 1'
,some_num=ranuni(0)
,some_date=round(ranuni(0)*1000,1)
/* First of the month, cycling through two years: a randomly spread date
gave every row its own day, so a date filter never returned many rows. */
,some_date=intnx('month','01JAN1960'd,mod(&x,24),'S')
,some_datetime=round(ranuni(0)*50000,1)
,some_time=round(ranuni(0)*100,1)
,some_shortnum=6+round(ranuni(0)*94,1)
+104 -4
View File
@@ -6,16 +6,19 @@
table - the engine reads only active rows, so the dependency can never
resolve at run time
2. a row is staged with active=0 while a surviving active row depends on it
3. a staged row breaks a field rule of the shared scan (core's
mp_derivedvarcheck), or carries an ACTIVE value the EDIT screen's picker
cannot produce
They should NOT abort when:
3. a new active row is added to a table that also holds an inactive row -
4. a new active row is added to a table that also holds an inactive row -
the inactive row is not part of the dependency graph, so it cannot
create a cycle or a stale reference
4. a delete-marked RLS rule naming a derived variable is submitted - the
5. a delete-marked RLS rule naming a derived variable is submitted - the
rule is being removed, so its variables are not validated
5. a delete and re-add of the same variable is submitted together - the
6. a delete and re-add of the same variable is submitted together - the
re-added active row resolves the dependency
6. an inactive row is set active again while another active row depends on
7. an inactive row is set active again while another active row depends on
it - the re-activation is what makes the dependency resolvable
The staged loads are aborted by the post edit hook. The assertion fetches the
@@ -553,6 +556,103 @@ run;
)
/**
* Test 8 - a field rule of the shared scan (mp_derivedvarcheck, core) is
* applied at submit time: an expression that could submit code of its own is
* refused by name, with the reason the scan gives. This is the same scan the
* engine runs at read time, so the case pins that the hook calls it rather
* than carrying a rule list of its own.
*/
data work.sascontroltable;
action='LOAD';
message='A refused function in VAR_LOGIC should abort';
libds="&dc_libref..MPE_DERIVED";
output;
stop;
run;
data work.jsdata;
length LIBREF $8 DSN $32 VAR_NM $32 VAR_DEPS $4000 VAR_TYPE $1
VAR_LOGIC $4000 VAR_FMT $32
_____DELETE__THIS__RECORD_____ $3;
LIBREF='DCTEST';
DSN='DC_DERIV';
VAR_NM='CALLER';
VAR_DEPS='';
VAR_TYPE='N';
VAR_LEN=8;
VAR_FMT='';
/* no semicolon in the payload: the single-expression rule is tested before
the function-name rule, so a quoted semicolon would abort on that first */
VAR_LOGIC='dosubl("data _null_")';
ACTIVE=1;
_____DELETE__THIS__RECORD_____='No';
run;
%mpe_execute(&_program,
viyacontext=&defaultcontext,
inputdatasets=work.sascontroltable work.jsdata,
outlib=web8,
outref=wb8,
mdebug=&sasjs_mdebug
)
/* the assertion compares the abort's MSG exactly, so the expected value is
assembled on one line: a literal wrapped across two source lines carries the
newline and the continuation indent with it */
%let exp8=MPE_DERIVED: CALLER: VAR_LOGIC may not call a function that;
%let exp8=&exp8 submits code or reaches the file system,;
%let exp8=&exp8 the environment or the macro symbol table;
%mpe_assert_abort(
desc=Checking a refused function in VAR_LOGIC is rejected,
outref=wb8,
msg=&exp8
)
/**
* Test 9 - the fields the hook owns rather than the scan: ACTIVE is a picker
* in the EDIT screen (HARDSELECT over the MPE_SELECTBOX pair), so a submission
* that bypasses the grid cannot land a value the screen would never offer.
*/
data work.sascontroltable;
action='LOAD';
message='An ACTIVE value the screen cannot offer should abort';
libds="&dc_libref..MPE_DERIVED";
output;
stop;
run;
data work.jsdata;
length LIBREF $8 DSN $32 VAR_NM $32 VAR_DEPS $4000 VAR_TYPE $1
VAR_LOGIC $4000 VAR_FMT $32
_____DELETE__THIS__RECORD_____ $3;
LIBREF='DCTEST';
DSN='DC_DERIV';
VAR_NM='BADACT';
VAR_DEPS='';
VAR_TYPE='N';
VAR_LEN=8;
VAR_FMT='';
VAR_LOGIC='1';
ACTIVE=2;
_____DELETE__THIS__RECORD_____='No';
run;
%mpe_execute(&_program,
viyacontext=&defaultcontext,
inputdatasets=work.sascontroltable work.jsdata,
outlib=web9,
outref=wb9,
mdebug=&sasjs_mdebug
)
%mpe_assert_abort(
desc=Checking an ACTIVE value outside the picker pair is rejected,
outref=wb9,
msg=MPE_DERIVED: BADACT: ACTIVE must be 1 or 0
)
/* tidy up so repeated runs start clean */
proc sql noprint;
delete from &dc_libref..mpe_derived
@@ -34,7 +34,7 @@
<h4> SAS Macros </h4>
@li mf_nobs.sas
@li mp_abort.sas
@li mpe_derived_check.sas
@li mp_derivedvarcheck.sas
<h4> Related Macros </h4>
@li mpe_derived.sas
@@ -100,14 +100,43 @@ run;
)
/**
* Field level checks - the same rules %mpe_derived applies at read time
* Field level checks. The rules for the derived variable itself - the name,
* the type, length, format, label and the expression - are core's, in
* mp_derivedvarcheck, so this hook and the read time engine (%mpe_derived ->
* mp_derivedvars) scan a row with the same rules and report the same reasons.
* What is added here are the two fields that belong to Data Controller rather
* than to the expression: the control table key, which the loader normalises
* but does not require, and ACTIVE, which the EDIT screen offers as a picker
* over the MPE_SELECTBOX pair - so a submission that bypasses the grid cannot
* land a value the screen would never offer.
*/
%mpe_derived_check(inds=work.staging_ds, outds=work.derived_check)
data work.derived_dcfield(keep=name reason);
set work.staging_ds;
if upcase(_____DELETE__THIS__RECORD_____)='YES' then delete;
length name $32 reason $200;
call missing(reason);
name=upcase(strip(var_nm));
if missing(libref) or missing(dsn) then reason='LIBREF and DSN are required';
else if missing(active) or active not in (0,1)
then reason='ACTIVE must be 1 or 0';
if not missing(reason) then output;
run;
%mp_derivedvarcheck(inds=work.staging_ds
,outds=work.derived_field
,reserved=%str(_____DELETE__THIS__RECORD_____)
)
/* the control table fields are reported before the expression's own rules, so
which of the two passes found a row does not change the message it gets */
data work.derived_check;
set work.derived_dcfield work.derived_field;
run;
%if %mf_nobs(work.derived_check)>0 %then %do;
data _null_;
set work.derived_check(obs=1);
call symputx('der_errmsg',cats('MPE_DERIVED: ',var_nm,': ',reason));
call symputx('der_errmsg',cats('MPE_DERIVED: ',name,': ',reason));
call symputx('der_errflag',1);
run;
%end;