Skip to content

Commit 5e35145

Browse files
committed
Merge remote-tracking branch 'origin/main' into claude/project-thread-r4sjiv
2 parents 4bc2d00 + a20824f commit 5e35145

118 files changed

Lines changed: 9092 additions & 2239 deletions

File tree

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.
Lines changed: 10 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,10 @@
1+
---
2+
'@object-ui/i18n': patch
3+
---
4+
5+
feat(i18n): the rest of the Studio inspectors and the package, form and editor strings move into the `studio` namespace
6+
7+
Clause-②: no
8+
Metadata-protocol: none
9+
10+
- **Third slice of the move ruled on objectstack#25922** (6431032267, Q2 = A; kept by 6431234244). The remaining `engine.inspector.*` keys and every `engine.packages.*`, `engine.form.*` and `engine.edit.*` key (551 in `en`, 560 in `zh`) now live in `@object-ui/i18n/namespaces/studio` instead of `@object-ui/app-shell`'s private table, text unchanged. The nine `engine.packages.form.help.*` rows stay `zh`-only: their English is the producer's own help text, read through `tOptional`. Studio reads the same strings; a Chinese viewer sees these labels in English for the moment the `zh` catalogue takes to load, as with the earlier slices. No export changes.
Lines changed: 13 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,13 @@
1+
---
2+
'@objectstack/spec': patch
3+
'@objectstack/core': patch
4+
---
5+
6+
fix(spec): `TranslationItemSchema.name` no longer tells authors to name a translation item after its locale tag
7+
8+
Clause-②: no
9+
Metadata-protocol: none
10+
11+
The `name` describe said a translation item was "conventionally the locale code (`zh-CN`)", but the metadata save door's item-name grammar refuses `zh-CN` (uppercase and a hyphen), so an author or agent following the describe got a 400 `INVALID_REQUEST` on the first save. The describe now says a translation item is named in snake_case like every metadata item (e.g. `zh_cn`), with the BCP-47 tag in `locale`. The save door and the runtime sync's read-side name fallback for rows stored before `locale` became required are unchanged; the sync's skip warning now points authors at `locale` only.
12+
13+
`@objectstack/core`: the authored-translation sync's no-locale skip warning now tells authors to set `locale` to the item's BCP-47 tag (e.g. `zh-CN`), instead of suggesting they name the item after it.
Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
---
2+
'@object-ui/core': patch
3+
---
4+
5+
`SPEC_ACTION_KEYS` lists `disabledReason`, the key `@objectstack/spec` added to `ActionSchema` beside `disabled` (objectstack#26786). An action that carries it is no longer counted as having an unknown key, and the inventory pin is green again.
6+
7+
Clause-②: no
8+
Metadata-protocol: none
Lines changed: 19 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,19 @@
1+
---
2+
'@objectstack/metadata-protocol': patch
3+
---
4+
5+
fix(metadata-protocol)!: the Data API refuses an empty projection and a count flag that is neither true nor false, instead of reading them as absent (#26644)
6+
7+
Clause-②: no (narrowing) — `GET /data/:object`, `POST /data/:object/query` (both transports: `RestServer` and the runtime dispatcher) and `GET /data/:object/:id` refuse a projection that names no field, and the list doors refuse a `count` / `$count` other than true or false, where they answered 200 before. Nothing is widened and no export changes.
8+
Metadata-protocol: none
9+
10+
<!-- adr-0087: not-required (no-migration-prescription) No key of any schema is added, removed, renamed or re-shaped, and no export or type changes: two query-parameter values the list normalizer read as the absent default are now refused 400 VALIDATION_ERROR naming the parameter. A caller drops the parameter or sends a well-formed value; there is nothing for a migration to rewrite and no tombstone. -->
11+
12+
**BREAKING** accept-set narrowing at the shared list normalizer (`findData`) and the by-id read (`getData`), served by every Data API door that uses them.
13+
14+
- **Before.** A blank or comma-only `$select` / `select` / `fields` (`?$select=`, `?$select=,`), and an empty array in a JSON body (`"fields": []`), became an empty projection, which every driver serves as no projection: all columns, under a 200. A `$count` / `count` other than `true` or `false` (`?$count=maybe`, `?$count=1`, a blank `?count=`) was left unlowered, so the `count=false` opt-out never matched and the count ran: a `total` was reported as if `$count=true` had been sent.
15+
- **Now.** Both are refused `400 VALIDATION_ERROR` before any read, with a message naming the parameter the caller wrote and the value (`Query parameter '$select' names no field; got ""`, `Query parameter '$count' must be true or false; got "maybe"`), as a blank `$top` already is. The accepted count values are the spec's `QueryTransportParamsSchema.count`: the boolean, or the strings `true` and `false`. `GET /data/:object/:id?select=` answers with the same code; that route's catch already reports every projection refusal (an unknown field included) as `404`, and the empty projection joins that answer.
16+
- **Unchanged.** An absent parameter keeps its default: every column, and the count runs. A named projection (`$select=a,b`, trailing commas and spaces trimmed as before), `$count=true`, `$count=false` and a JSON `null` answer exactly as before. `POST /data/:object/query` already refused a non-boolean `count` at its body schema (`400 VALIDATION_FAILED`), and still does.
17+
- **Who sees it.** A caller that builds `$select` from a list that can be empty, or sends a mistyped `$count`. It now learns that from a 400 instead of an answer that looked normal.
18+
19+
**The remedy.** Leave the parameter out to read every column or to keep the count, or send a projection naming at least one field and a count flag of `true` or `false`.
Lines changed: 20 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,20 @@
1+
---
2+
'@objectstack/formula': minor
3+
---
4+
5+
feat(formula)!: a template hole takes one formatter, and the `currency` formatter's argument is a literal ISO 4217 code — both refused at compile, never rendered as formatter syntax or in US dollars
6+
7+
Clause-②: no (narrowing) — the in-app template dialect's compile (and so `os validate` and the publish gate for every template slot) refuses a hole with a second formatter (`{{ x | default:0 | number:2 }}`) and a `{{ x | currency:ARG }}` hole whose `ARG` is not three capital letters, both of which it accepted before; `formatValue('currency', value, ARG)` now throws a `RangeError` for such an `ARG`: a path or another argument `Intl` refuses, and an empty one, used to fall back to US dollars, and a lowercase code (`eur`), which `Intl` formatted in that currency, is now refused too. No accept set is widened; the additions are exports only: `judgeHoleFormatter`, `mailHolePattern`, `parseMailHole`, `judgeMailHoleGrammar`, and the types `TemplateHoleGrammarRefusal`, `TemplateHoleGrammarRule`, `ParsedMailHole` and `MailHoleGrammarFinding`.
8+
Metadata-protocol: narrowed
9+
10+
<!-- adr-0087: not-required (no-migration-prescription) No key of any schema is added, removed, renamed or re-shaped and no export is removed: a chained formatter and a currency argument that never rendered as written are now refused where the template is compiled. The refusal names the hole and the rule it breaks, so there is nothing for a migration to rewrite. -->
11+
12+
**BREAKING** accept-set narrowing in the template dialect, and a new shared mail hole grammar.
13+
14+
- **Before, a chained hole.** `{{ record.n | default:0 | number:2 }}` compiled as `default` with the argument `0 | number:2`, and rendered that text.
15+
- **Before, a currency path.** `{{ record.amount | currency:currency_code }}` compiled; `Intl` refused the argument and the formatter caught that and used US dollars, so the amount printed in a currency nobody wrote.
16+
- **Now.** `templateEngine.compile` / `evaluate` answer a `parse` error naming the hole and the rule. `formatValue('currency', …)` throws a `RangeError` for an argument that is not a code, before it reads the value. A lowercase code (`currency:eur`), which formatted in euros before, is refused the same way. No argument still formats in US dollars.
17+
- **Shared grammar.** `judgeHoleFormatter` is the judgement both template dialects use. The mail template grammar is exported once, `mailHolePattern`, `parseMailHole` and `judgeMailHoleGrammar`, and is what `@objectstack/plugin-email`'s renderer, `judgeMailHtmlHoles` and `@objectstack/lint` read; `judgeMailHtmlHoles` no longer keeps a copy of it.
18+
- **Who sees it.** A template slot that chained formatters, or wrote a path or a lowercase code as the currency argument. No template shipped in this repository does.
19+
20+
**The remedy.** Keep one formatter per hole and decide anything more where the data is assembled. Write the currency code itself (`currency:EUR`), or format the amount where the data is assembled and pass the formatted text.
Lines changed: 10 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,10 @@
1+
---
2+
'@objectstack/service-messaging': patch
3+
---
4+
5+
fix(service-messaging): the email and inbox channels grade a `TEMPLATE_INVALID` template failure permanent, like the other template failures
6+
7+
Clause-②: no
8+
Metadata-protocol: none
9+
10+
`@objectstack/plugin-email`'s `sendTemplate` and `renderTemplate` now fail with `TEMPLATE_INVALID` for a stored template whose hole the mail grammar refuses (a second formatter, or a `currency` argument that is not a literal ISO 4217 code). Retrying cannot succeed until the template is republished, so `classifyError` on both channels grades it `permanent`, as it already grades `TEMPLATE_NOT_FOUND`, `TEMPLATE_INACTIVE`, `MISSING_VARIABLES` and `TEMPLATE_UNSUPPORTED`: the delivery goes dead at once with the code on its row, instead of spending the whole retry schedule first. Every other failure keeps its grade.
Lines changed: 19 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,19 @@
1+
---
2+
'@objectstack/plugin-email': minor
3+
---
4+
5+
feat(plugin-email)!: a mail template hole that chains a second formatter, or whose `currency` argument is not a literal ISO 4217 code, fails the render with `TEMPLATE_INVALID` instead of printing formatter syntax or US dollars
6+
7+
Clause-②: no (narrowing) — `sendTemplate` and `renderTemplate` refuse a stored template whose `subject`, `body_text` or `body_html` writes a hole with a second formatter, or `{{ x | currency:ARG }}` with an `ARG` that is not three capital letters, both of which rendered before. No accept set is widened, and the package adds no export.
8+
Metadata-protocol: narrowed
9+
10+
<!-- adr-0087: not-required (no-migration-prescription) No key of any schema is added, removed, renamed or re-shaped and no export changes: the renderer refuses a hole the authoring rule refuses, for a template that reached it without being judged. There is nothing for a migration to rewrite. -->
11+
12+
**BREAKING** a render that printed a wrong currency now fails.
13+
14+
- **Before, a chained hole.** `{{ actual_amount | default:0 | number:2 }}` rendered `0 | number:2` for an empty value and `71100` otherwise, in every face.
15+
- **Before, a currency path.** `{{ planned_amount | currency:currency_code }}` with `currency_code = 'eur'` rendered `US$1,234.50` (zh-CN), `$1,234.50` (en-US) and `1 234,50 $US` (fr-FR), in every face.
16+
- **Now.** The render throws `TEMPLATE_INVALID: email template NAME (locale=LOCALE): …`, naming the hole and the rule it breaks (one formatter per hole, or the literal-only currency code); `sendTemplate` sends nothing. The authoring doors refuse the same hole first (`email-template-hole-grammar`), so this is reached only by a template stored before that rule, until its next publish.
17+
- **The grammar.** The renderer parses holes with `@objectstack/formula`'s `parseMailHole`, the grammar the authoring rule and the placement judge read; it keeps no copy of its own.
18+
19+
**The remedy.** Republish the template with one formatter per hole and a literal currency code (`currency:EUR`), or format the amount in the caller and pass the formatted text as its own variable.
Lines changed: 20 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,20 @@
1+
---
2+
'@objectstack/lint': minor
3+
---
4+
5+
feat(lint)!: a new gating authoring rule, `email-template-hole-grammar`, refuses an email template hole that chains a second formatter or whose `currency` argument is not a literal ISO 4217 code
6+
7+
Clause-②: no (narrowing) — `os validate`, `os build`, `os lint` and the metadata publish door (`422 INVALID_METADATA`) refuse an `email_template` whose `subject`, `bodyText` or `bodyHtml` writes a hole with a second formatter (`{{ x | default:0 | number:2 }}`) or `{{ x | currency:ARG }}` with an `ARG` that is not three capital letters, both of which they accepted before. No accept set is widened; the additions are exports only: `validateEmailTemplateHoleGrammar`, `EMAIL_TEMPLATE_HOLE_GRAMMAR`, the type `EmailTemplateHoleGrammarFinding`, and the rule id `email-template-hole-grammar`.
8+
Metadata-protocol: narrowed
9+
10+
<!-- adr-0087: not-required (no-migration-prescription) No key of any schema is added, removed, renamed or re-shaped: a hole that never rendered as written (the renderer printed formatter syntax, or the amount in US dollars) is now refused where it is authored. The refusal names the hole and the rule it breaks, so there is no key for a migration to rewrite and no tombstone. -->
11+
12+
**BREAKING** accept-set narrowing at every authoring door for `email_template`.
13+
14+
- **Before, a chained hole.** `{{ actual_amount | default:0 | number:2 }}` passed every gate. The second `|` was read as part of `default`'s argument, so the hole printed the text `0 | number:2` for an empty value and the bare `71100` otherwise; `{{ x | upper | lower }}` stayed on the page as written.
15+
- **Before, a currency path.** `{{ planned_amount | currency:currency_code }}` passed every gate. The argument is not read as a path, so the amount rendered in US dollars whatever the record's currency (`$1,234.50` for a record in euros).
16+
- **Now.** The rule refuses both at `error`, naming the hole, the face and the template. A hole takes one formatter; anything more is decided in the caller that supplies the data. The currency argument is a literal code in three capital letters (`currency:EUR`), never a path. A lowercase or empty argument is refused the same way. `currency` with no argument is unchanged (US dollars).
17+
- **Which holes.** The grammar is `@objectstack/formula`'s `judgeMailHoleGrammar`, the one the mail renderer parses with; all three faces are judged.
18+
- **Who sees it.** A template that chained formatters, or wrote a path or a lowercase code as the currency argument. No template shipped in this repository does.
19+
20+
**The remedy.** Keep one formatter per hole and pass any further decided value (a default, a second format) as its own variable. Write the currency code itself (`currency:EUR`). To show each record's own currency, format the amount in the caller that supplies the data and pass the formatted text as its own variable.

‎.changeset/26670-seed-file-pool.md‎

Lines changed: 10 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,10 @@
1+
---
2+
'@objectstack/metadata-protocol': patch
3+
---
4+
5+
fix(metadata-protocol): a seed that stores many `seedFile` values no longer starves sign-ins and other requests while it runs
6+
7+
Clause-②: no
8+
Metadata-protocol: none
9+
10+
A seed that stores many `seedFile(path)` values no longer starves the requests that arrive while it runs. On a synchronous driver (better-sqlite3, the wasm fallback) every step of a seed file's work — the `sys_file` row, the bytes, the claim that commits it — resolves as a microtask, so the claims the seed loader settled after each 200-row insert batch ran as one unbroken chain and a sign-in during a background boot seed waited seconds unread behind it. The connection pool was never the bottleneck; the event loop was. The loader now hands the event loop one turn after each row's files are stored and after each row's claim, and writes a dataset that carries seed files in insert batches of 20 rows instead of 200. Datasets without seed files are unchanged.
Lines changed: 10 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,10 @@
1+
---
2+
'@object-ui/core': patch
3+
---
4+
5+
fix(core): an `api` action's target is interpolated before a registered `api` handler sends it
6+
7+
Clause-②: no
8+
Metadata-protocol: none
9+
10+
An `api` action whose `target` carries `${ctx.recordId}` (or any other `${ctx.X}` / `${param.X}` token) went out on the wire with the literal placeholder (`/api/v1/data/gapdata_item/$%7Bctx.recordId%7D`) when a host had registered its own `api` handler, as the Console does, so a record-level PATCH or POST hit a record that does not exist. A `url` action with the same token already resolved it. `ActionRunner` now runs its one target substitution on an `api` target before handing the action to a registered `api` handler, with the same scopes and the same per-value `encodeURIComponent` a `url` target gets. The built-in executor already did this and is unchanged; a registered handler for any other type still receives its target as authored.

0 commit comments

Comments
 (0)