feat(gui): author the Form View in place with an edit mode by yangzhang75 · Pull Request #8455 · apache/texera

yangzhang75 · 2026-09-09T02:00:27Z

What changes were proposed in this PR?

Closes #8026. Part of the Form View stack (parent issue #8011), on main now that #8442 (PR14) has merged. Two commits: 871649877 is the feature; 01d0e2d35 is a test-only follow-up for the Windows CI leg (last bullet). The tip is 01d0e2d35.

Adds the authoring side of the page: with write access, an Edit toggle turns the form into in-place authoring.

toggleAuthoring: authoring enables graph modification and turns the inspect panel live (the panel acts as an editor, actsAsEditor); leaving it returns the page to the reader state (inert panel, modification disabled). Entering edit mode is refused without write access at the method itself, not only by hiding the button.
Rename an exposed input or any of its sub-fields, and hide a sub-field, through a new editable-label-wrapper formly wrapper (the label is the input; leaving it empty falls back to the schema label; the reader's label is tied to its control with for). Overrides are stored per binding and reapplied on every rebuild.
Reorder inputs by drag (CDK drag-drop, mapped by binding id so a hidden card cannot desync the order) or from the keyboard with Move up / Move down on each card, remove one, or expose a new one by clicking a step on the preview.
Write the author's instruction (markdown, with preview; the title is edited in the header, which is a row rather than one button so the input is not nested in a control) and per-input help text; "From X" attribution shows only while authoring. A binding whose operator no longer exists renders as a removable card for the author and is hidden from readers; it is never deleted silently.
A repeated (array) input keeps its title above its rows in reader mode too: the shared array widget prints its label at the bottom beside its add button, so without this the title jumped from above the rows in edit mode to below them on Done. The static label is tied to the control with for.
Carries the same rebuild-while-typing fix as fix(workflow-form): hold a rebuild that arrives while typing instead of dropping it #8499 (a tick box is not typing; a rebuild that arrives mid-typing is held and runs on focusout) so the tick boxes work here on their own; whichever merges first, the other rebases and the duplicate hunk disappears.
Open canvas saves first and hands over only once the save has completed: the switch is a full-page load, which aborts a request still in flight, so navigating right after firing the save could lose the edit it was meant to carry across. A failed save keeps the author on the form with the error shown; a reader with nothing to save goes straight through. (feat(gui): wire the Form View entry points #8456 does the same on the canvas side.)
Pick which extra results to feature: the picker offers the non-terminal view-result steps (plus already-chosen ones) and rebuilds live when an eye is toggled on the canvas. The final step's result always shows, so it is not offered.
Keyboard focus is visible on the pill, remove, tab and eye buttons (:focus-visible).
The embedded preview's right-click menu carries the structure lock. Edit mode re-enables workflow modification for the property panel, and the menu's re-shaping commands used to follow that flag alone, so the structure-locked preview would have offered cut, paste, delete and disable again. texera-context-menu now takes [structureLocked] from the editor and gates those commands on modification AND no lock (canModify); copy, the result toggles, execute-to and export do not re-shape the graph and are unchanged.
While authoring, the control keeps a programmatic name: the visible name box edits the label but is not the label, so a visually hidden <label for> that follows the current name (the author's, else the schema's) sits beside it. Readers keep the static label as before.
Graph edits go through the shared graph; formBinding edits stay local until fix(workflow): keep workflowSettings and formBinding in the shared model #8351 shares them.
01d0e2d35 (test only, workspace menu): the menu's export test module-mocked the CommonJS file-saver package with vi.mock. Under the Angular unit-test builder that call is not reliably hoisted (Vitest warns about it on every platform), and with this stack's shared-chunk graph it stopped applying on the Windows leg: the real saveAs ran and the spy saw 0 calls, on 7 of the last 8 Windows runs of feat(gui): author the Form View in place with an edit mode #8455/feat(gui): wire the Form View entry points #8456 while main and fix(workflow-form): hold a rebuild that arrives while typing instead of dropping it #8499 pass. The export now goes through the existing injectable FileSaverService (as the dashboard downloads already do) and the spec stubs that with TestBed. No behaviour change.

The diff is over 1000 lines because it is 813 lines of spec against 1410 of source (of which 444 are stylesheet); the authoring header, wrapper and reorder logic are one feature and do not split further without landing a half-usable edit mode.

Any related issues, documentation, discussions?

Closes #8026. Part of the Form View feature (parent issue #8011).

How was this PR tested?

Unit tests (vitest). Direct-construction tests cover the authoring gate (a reader cannot enter edit mode and the graph stays unmodifiable), rename/hide overrides reaching the rendered fields, drag and keyboard reorder with the id-mapped indices (including the end stops and a card the config no longer holds), broken bindings kept for explicit removal, the result picker's range and its live rebuild on an eye toggle, the instruction/help-text editing, and the switch's save-then-navigate order (navigates on complete, stays on error, straight through for a reader). The rendered spec covers the edit-mode header shape (title input outside the toggle, aria-controls), the Move buttons, and the panel turning live in edit mode. The editable-label wrapper has its own TestBed spec (real component: decorate, rename, hide toggle, label association for readers and, while authoring, the hidden label following the current name). The context menu's spec renders it under a structure lock with modification enabled and checks cut, paste, delete, disable and enable are off while copy and the result toggle stay; the editor's spec opens the right-click menu on a locked editor and checks the lock arrived. Each new guard was deletion-checked (removing it turns the corresponding test red). eslint, prettier and the production (AOT) build pass; every changed line is statement and function covered.

video

Screen.Recording.2026-09-10.at.6.05.08.PM.mov

Was this PR authored or co-authored using generative AI tooling?

Yes. Generated-by: Claude Code (Claude Fable 5.1, Anthropic). Co-authored with Claude, reviewed line by line by the author before submission.

🤖 Generated with Claude Code

https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY

github-actions · 2026-09-09T02:01:49Z

Automated Reviewer Suggestions

Based on the git blame history of the changed files, we recommend the following reviewers:

Contributors with relevant context: @PG1204, @Yicong-Huang, @mengw15
You can notify them by mentioning @PG1204, @Yicong-Huang, @mengw15 in a comment.

codecov-commenter · 2026-09-09T02:10:10Z

Codecov Report

❌ Patch coverage is 48.56115% with 143 lines in your changes missing coverage. Please review.
✅ Project coverage is 93.47%. Comparing base (1fbd346) to head (01d0e2d).
⚠️ Report is 6 commits behind head on main.

Files with missing lines	Patch %	Lines
...mponent/workflow-form/workflow-form.component.html	0.00%	132 Missing ⚠️
...abel-wrapper/editable-label-wrapper.component.html	76.00%	2 Missing and 4 partials ⚠️
...component/workflow-form/workflow-form.component.ts	95.04%	0 Missing and 5 partials ⚠️

❌ Your patch status has failed because the patch coverage (48.56%) is below the target coverage (60.00%). You can increase the patch coverage or adjust the target coverage.

Additional details and impacted files

@@             Coverage Diff              @@
##               main    #8455      +/-   ##
============================================
- Coverage     93.69%   93.47%   -0.23%     
  Complexity     4826     4826              
============================================
  Files          1209     1211       +2     
  Lines         49871    50110     +239     
  Branches       6099     6135      +36     
============================================
+ Hits          46727    46840     +113     
- Misses         1652     1768     +116     
- Partials       1492     1502      +10

Flag	Coverage Δ		*Carryforward flag
access-control-service	`80.18% <ø> (ø)`		Carriedforward from 1fbd346
agent-service	`99.32% <ø> (ø)`		Carriedforward from 1fbd346
amber	`89.91% <ø> (ø)`		Carriedforward from 1fbd346
computing-unit-managing-service	`77.14% <ø> (ø)`		Carriedforward from 1fbd346
config-service	`87.12% <ø> (ø)`		Carriedforward from 1fbd346
file-service	`83.65% <ø> (ø)`		Carriedforward from 1fbd346
frontend	`95.64% <48.56%> (-0.52%)`	⬇️
notebook-migration-service	`83.73% <ø> (ø)`		Carriedforward from 1fbd346
pyamber	`98.47% <ø> (ø)`		Carriedforward from 1fbd346
workflow-compiling-service	`77.19% <ø> (ø)`		Carriedforward from 1fbd346

*This pull request uses carry forward flags. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

Copilot

Warning

Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.

Pull request overview

Adds in-place authoring capabilities to the Workflow Form View, including edit-mode UI, result rendering/curation, and read-only step inspection via the property panel.

Changes:

Add hasNonEmptyResult to distinguish empty results from “no result”.
Extend WorkflowFormComponent with authoring mode (rename/hide/reorder/expose/remove inputs), results display + picker, and read-only property-panel inspection.
Introduce editable-label-wrapper Formly wrapper and adjust property editor components to support embedded read-only usage.

Reviewed changes

Copilot reviewed 19 out of 19 changed files in this pull request and generated 8 comments.

Show a summary per file

File	Description
frontend/src/app/workspace/service/workflow-result/workflow-result.service.ts	Adds `hasNonEmptyResult` helper for UI logic that should ignore empty results.
frontend/src/app/workspace/service/workflow-result/workflow-result.service.spec.ts	Adds unit test coverage for `hasNonEmptyResult` behavior.
frontend/src/app/workspace/service/form-binding/form-binding.service.ts	Updates docstring clarifying “featured results” behavior and terminal-step always-shown rule.
frontend/src/app/workspace/component/workflow-form/workflow-form.spec-harness.ts	Expands harness with result/view-result/highlight mocks needed by new form-view behavior.
frontend/src/app/workspace/component/workflow-form/workflow-form.rendered.spec.ts	Adds rendered-template coverage for results section and stubs heavy children.
frontend/src/app/workspace/component/workflow-form/workflow-form.component.ts	Core feature work: authoring toggle, input overrides, results rendering, step inspection, and chart fitting.
frontend/src/app/workspace/component/workflow-form/workflow-form.component.spec.ts	Adds unit tests for authoring gate, results filtering, view-result-change behavior, inspection panel, etc.
frontend/src/app/workspace/component/workflow-form/workflow-form.component.scss	Adds styles for embedded property panel, results cards, and authoring UI.
frontend/src/app/workspace/component/workflow-form/workflow-form.component.html	Adds edit-mode header, authoring instruction editor, drag-drop inputs, results section, and results picker.
frontend/src/app/workspace/component/property-editor/property-editor.component.ts	Adds `persistPlacement` and `broadcastEditing` inputs for embedded/read-only panel usage.
frontend/src/app/workspace/component/property-editor/property-editor.component.spec.ts	Adds tests for `persistPlacement=false` and new `broadcastEditing` input wiring.
frontend/src/app/workspace/component/property-editor/operator-property-edit-frame/operator-property-edit-frame.component.ts	Guards shared-model writes behind `broadcastEditing` for read-only inspection mode.
frontend/src/app/workspace/component/property-editor/operator-property-edit-frame/operator-property-edit-frame.component.spec.ts	Adds tests asserting broadcast/version-sync occur (or not) based on `broadcastEditing`.
frontend/src/app/common/type/workflow.ts	Clarifies `resultOperatorIds` semantics in the config type doc.
frontend/src/app/common/formly/formly-config.ts	Registers the new `editable-label-wrapper` Formly wrapper.
frontend/src/app/common/formly/editable-label-wrapper/editable-label-wrapper.component.ts	Adds the wrapper component plus `decorate()` helper for author rename/hide UI.
frontend/src/app/common/formly/editable-label-wrapper/editable-label-wrapper.component.spec.ts	Adds component-level tests for wrapper behavior, handlers, and template branches.
frontend/src/app/common/formly/editable-label-wrapper/editable-label-wrapper.component.scss	Styles for editable label + hide toggle, including “dimmed when hidden” authoring view.
frontend/src/app/common/formly/editable-label-wrapper/editable-label-wrapper.component.html	Template for editable label input + optional hide toggle (authoring) vs static label (reader).

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

yangzhang75 · 2026-09-11T02:35:11Z

/request-review: @mengw15

### What changes were proposed in this PR? Closes apache#8496. A regression from apache#8125: since then, every workflow's second autosave (and every one after it) fails with 500, silently on the operator canvas and as "Could not save" on the Form View. `persistWorkflow` wrote `is_public` from the request. The frontend feeds the saved row straight back as its metadata, and that row names the flag `isPublic` while the rest of the frontend calls it `isPublished`, so the very next save went out without the flag, the update wrote NULL into a NOT NULL column, and the request failed. A stale `isPublic: false` on a save could likewise un-publish a published workflow. - Backend: `saveWorkflowFields` now writes name, description and content only. Publishing stays with `/public` and `/private`, `default_view` with `/set-default-view`, and the timestamps are not rewritten, so a save can never clobber a concurrent change to any of them. - Frontend: `WorkflowPersistService.persistWorkflow` no longer sends `isPublic` (the endpoint does not read it, and the value is not reliably known after the first save), and `WorkflowUtilService.parseWorkflowInfo` carries a persist response's `isPublic` over to `isPublished`, so metadata fed back from a save keeps the publish state instead of dropping it. The Form View stack is not blocked by this: apache#8455 touches none of these files, and apache#8456 touches `workflow-persist.service.ts` only in `createWorkflow` (adding `defaultView`), a different function; a dry-run merge of the two is clean. ### Any related issues, documentation, discussions? Closes apache#8496. Found while verifying apache#8455 on a flag-on instance (parent apache#8011). ### How was this PR tested? Backend: `WorkflowResourceSpec` gains two tests, a save carrying no flag neither fails nor changes `is_public` after `/public`, and a save carrying `false` does not un-publish; the existing default-view save test was updated to send no flag, as the frontend does. `WorkflowResourceSpec` and `PublishedCopySchemaSpec` pass (87 tests), scalafmt clean. Frontend: the persist spec asserts the save payload carries no `isPublic` and that the response's `isPublic` comes back as `isPublished`; `parseWorkflowInfo` gains tests for the carry-over and for leaving a present `isPublished` alone. Full suite passes (5777), changed lines fully covered, eslint, prettier and the production (AOT) build pass. End to end, against a running stack rebuilt with this change: the exact second-save payload that returned 500 now returns 200 with `is_public` unchanged; a create-through-persist with the new payload inserts with `is_public = false`; and in a headless browser the Form View renames a workflow twice with every `/api/workflow/persist` answering 200, no `isPublic` key in any request body, and no "Could not save". ### Was this PR authored or co-authored using generative AI tooling? Yes. Generated-by: Claude Code (Claude Fable 5.1, Anthropic). Co-authored with Claude, reviewed line by line by the author before submission. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com> Co-authored-by: Meng Wang <mengw15@uci.edu>

…of dropping it (apache#8499) ### What changes were proposed in this PR? Closes apache#8497. The Form View skips rebuilding its input cards while the reader is typing, so a rebuild cannot throw away a half-entered value (from apache#8437). Two things were wrong with that skip. - It counted any focused `INPUT` as typing, tick boxes included. Ticking a property in the step panel focuses the tick box, so the rebuild that should add the card was skipped and the tick looked like it did nothing until something else rebuilt the page (the way apache#8455's edit mode surfaces it). - A skipped rebuild was dropped, not postponed. A schema refresh that landed while someone was typing (the column names an attribute dropdown offers, after a compile) never reached the cards until the next compile or a reload. Now only text-like inputs, textareas, selects and content-editables count as typing, and a rebuild that does arrive mid-typing is held and runs once the focus leaves the text control: a `focusout` host listener decides one tick later, so tabbing to the next text field keeps it held while any other focus change runs it. Both subscribers (compilation state, form-binding changes) go through the one `rebuildFormOrDefer`. apache#8455 carries the same change so it stays self-contained and testable on its own; whichever merges first, the other rebases and the duplicate hunk disappears. ### Any related issues, documentation, discussions? Closes apache#8497. Found while verifying apache#8455 on a flag-on instance (parent apache#8011). ### How was this PR tested? Unit tests (vitest): a focused tick box is not typing; a held rebuild runs once on focusout for both the compilation and the binding stream; it stays held when the focus only moves to another text field; a focusout with nothing held rebuilds nothing; a queued focusout callback does not rebuild a second time when the control that took the focus already rebuilt (the tick box case). The rendered spec dispatches a real bubbling `focusout` from a control inside the page and asserts the host listener drains the held rebuild. Each was checked to fail with its half of the fix removed, including removing the `@HostListener` decorator. Full suite passes (5780), changed lines fully covered, eslint, prettier and the production (AOT) build pass. In a headless browser against a dev server built from this branch: typing a value into an exposed input and tabbing out keeps the value through the held rebuild, with no console errors and every autosave answering 200. The tick-box path was exercised the same way on apache#8455, which carries the identical change: the card follows the tick within a second, in both directions. ### Was this PR authored or co-authored using generative AI tooling? Yes. Generated-by: Claude Code (Claude Fable 5.1, Anthropic). Co-authored with Claude, reviewed line by line by the author before submission. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>

Copilot

🟡 Changes recommended

Edit mode exposes structural context-menu mutations and rebuilds can discard pending debounced form edits.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Review details

Suppressed comments (4)

frontend/src/app/workspace/component/workflow-form/workflow-form.component.ts:1132

This immediate re-read can discard the user's latest input value. Each card writes through a debounced form.valueChanges, but readConfig() calls buildForm(), whose formsRebuilt.next() unsubscribes that pending debounce. If someone types a value and clicks Done before the debounce expires, the old graph value is rebuilt and the edit is lost. Flush pending card values before rebuilding, or defer mode/structural rebuilds until those writes have committed.

    // Edit mode is what makes operator properties (and the embedded canvas) editable here.
    this.applyEditability();
    this.readConfig();

frontend/src/app/workspace/component/workflow-form/workflow-form.component.ts:599

A disabled operator can retain its viewResult flag, but execution builds the logical plan from getSubDAG(), which excludes disabled operators, so its result is not materialized. This filter currently offers such an unchosen disabled step as a selectable result. Exclude disabled viewed operators while retaining already-chosen ones so stale selections can still be removed.

    this.resultChoices = this.operators()
      .filter(op => !terminals.has(op.operatorID) && (viewed.has(op.operatorID) || chosen.has(op.operatorID)))

frontend/src/app/workspace/component/workflow-form/workflow-form.component.ts:861

The stored override is assigned to node.props.label immediately above, so using that value as fallback makes the editor claim that clearing a custom name keeps the same custom name. Capture the schema label before applying override.displayName and pass that original value as the fallback.

              fallback: (node.props?.label as string) || path,

frontend/src/app/common/formly/editable-label-wrapper/editable-label-wrapper.component.html:55

This for is dangling for the reader-mode use introduced here: the wrapper is attached to an array field, while ArrayTypeComponent renders rows and buttons but no element with the array field's id (array.type.ts:20-55). Therefore the visible array title does not label any control or group. Give the title its own id and apply role="group" plus aria-labelledby to the container around fieldComponent.

<label
  class="lbl-static"
  *ngIf="!props.authoring && (props.authorName || props.schemaLabel)"
  [attr.for]="id">

Files reviewed: 13/13 changed files
Comments generated: 2
Review effort level: Balanced

The workspace menu's export test module-mocked the CommonJS file-saver package (vi.mock("file-saver")). Under the Angular unit-test builder that call ends up inside the built spec chunk rather than at its top level (Vitest warns about it on every platform), and once this stack shifted the shared chunk graph the mock stopped applying on the Windows leg: the real saveAs ran and the spy saw 0 calls (7 of the last 8 Windows runs on apache#8455 and apache#8456, while main and apache#8499 pass). Route the download through the existing injectable FileSaverService, as the dashboard downloads already do, and stub that with TestBed, which does not depend on mock hoisting. No behaviour change. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY

Add an Edit/Done toggle (write access only) that turns the Form View into in-place authoring without leaving the page: rename an input or its sub-fields by clicking the title, hide a sub-field (kept faded, not removed), reorder inputs by drag, expose a new one by clicking a step on the embedded preview (the property panel goes live with tick boxes), remove an input, write the instruction (write/preview), and pick which steps report their results. An author sees broken inputs (their operator was deleted) to repair or remove them; a reader never does. Leaving edit mode locks everything back to read-only. Every edit goes through the same shared graph and form-binding config the operator canvas edits. Adds a reusable editable-label formly wrapper (renders a plain label for a reader, an editable input with an optional hide eye for an author). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY

The workspace menu's export test module-mocked the CommonJS file-saver package (vi.mock("file-saver")). Under the Angular unit-test builder that call ends up inside the built spec chunk rather than at its top level (Vitest warns about it on every platform), and once this stack shifted the shared chunk graph the mock stopped applying on the Windows leg: the real saveAs ran and the spy saw 0 calls (7 of the last 8 Windows runs on apache#8455 and apache#8456, while main and apache#8499 pass). Route the download through the existing injectable FileSaverService, as the dashboard downloads already do, and stub that with TestBed, which does not depend on mock hoisting. No behaviour change. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY

yangzhang75 · 2026-09-11T18:22:30Z

Closing: at 2200 lines this was too much for one review, so it is split in two, both rebased onto current main (which now carries #8498 and #8499, so the duplicated rebuild-while-typing hunk is gone):

feat(gui): add the Form View edit mode, with instruction and result authoring #8516: the edit mode, with instruction and result authoring (closes feat(workflow-form): Form View edit mode, with instruction and result authoring #8026)
feat(gui): author the Form View's inputs in place #8517: authoring the inputs in place: rename, hide, reorder, remove (closes feat(workflow-form): author the Form View's inputs in place (rename, hide, reorder, remove) #8027, reused for this half), stacked on feat(gui): add the Form View edit mode, with instruction and result authoring #8516

Together they are byte-for-byte the tree that was here at 01d0e2d35, rebased, plus two strengthened tests; every fix from this PR's review threads is in one of the two. #8456 is now stacked on #8517.

The workspace menu's export test module-mocked the CommonJS file-saver package (vi.mock("file-saver")). Under the Angular unit-test builder that call ends up inside the built spec chunk rather than at its top level (Vitest warns about it on every platform), and once this stack shifted the shared chunk graph the mock stopped applying on the Windows leg: the real saveAs ran and the spy saw 0 calls (7 of the last 8 Windows runs on apache#8455 and apache#8456, while main and apache#8499 pass). Route the download through the existing injectable FileSaverService, as the dashboard downloads already do, and stub that with TestBed, which does not depend on mock hoisting. No behaviour change. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY

…uthoring (apache#8516) ### What changes were proposed in this PR? Closes apache#8026. Part of the Form View stack (parent issue apache#8011), on main. This is the first half of what was apache#8455 (closed, split in two at 2200 lines); the second half, authoring the inputs themselves, is apache#8517 stacked on this one. Four commits: the first is the feature, the second and third are test-only follow-ups (the Windows CI leg and the template coverage report), and the last opens the result picker to everyone with one saved list of shown results (the design changes asked for in review, mengw15's list shape included). The review commit is the branch tip. Adds the edit mode of the page and the page-level authoring it enables. - `toggleAuthoring`: with write access, Edit turns the page into in-place authoring; Done returns it to the reader state. Edit mode is the only state that enables workflow modification here (`applyEditability`), and entering it is refused without write access at the method itself, not only by hiding the button. - The inspect panel goes live in edit mode: it acts as an editor (`actsAsEditor`), its expose tick boxes are on (`exposeChoosing`), and `inert` comes off, so the author opens a step on the embedded preview and chooses which of its settings people fill in right there. It stays inert and read-only for everyone else, as before. - The embedded preview's right-click menu carries the structure lock. The menu's re-shaping commands followed the modification flag alone, so re-enabling it for the panel would have offered cut, paste, delete and disable on the structure-locked preview. `texera-context-menu` now takes `[structureLocked]` from the editor and gates those on modification AND no lock (`canModify`); copy, the result toggles, execute-to and export do not re-shape the graph and are unchanged. - Write the author's instruction in place: the heading is edited in the header (a row rather than one button, so the input is not nested in a control) and the body as markdown with Write / Preview tabs. A reader still sees it rendered, only when there is text. - Pick which results to show, for everyone: the picker lists the final steps and the intermediate steps that have view-result on the canvas, and rebuilds live when an eye is toggled or the graph's shape or names change (a step deleted, disabled, given a downstream link, or renamed, a co-editor's edit included), from the graph's own streams rather than only after the debounced compilation. What shows is one saved list, `shownResultIds` (the shape mengw15 proposed): absent, every final step shows, as on the canvas; once the author has chosen, exactly the listed steps show, and `[]` means none, a choice the earlier pair of opposite-signed lists (featured / turned off) could not store and could contradict each other on. In edit mode a pill adds the step to or removes it from that list; the first choice starts the list from the final steps at that moment, so nothing the author saw disappears. The accepted cost: a step that becomes final after the author has chosen does not appear by itself. A definition whose only content is that list, an empty one included, is still saved. A listed step whose eye has since been turned off stays offered in edit mode only, so the author can take it off the list; a reader is not offered a pill that could never turn on. Anyone else, a writer merely viewing included, gets the same picker but changes only their own view of the page; nothing is written, so a reader without write access can choose too, and the choice lasts for the page. A disabled step is neither offered nor shown, eye or no eye: the compiled plan leaves it out. - Edit mode is the only state that unlocks the graph from this page, and only while no run is in flight (the canvas rule, kept so that entering edit mode mid-run cannot undo it). The lock is a root-level flag with writers that know nothing of this page: the execute service unlocks it whenever a run ends (completed, failed, killed, reset), and the computing-unit selector embedded here unlocks it when it finds no run on the chosen unit. Rather than chase each caller, the page clamps at the stream they all report to (`getWorkflowModificationEnabledStream`): whenever the flag turns on while the page must stay locked, it is turned off again, so a writer merely viewing can never reach the preview's view-result command. In edit mode the canvas rule stands: locked while running, unlocked when the run ends. The clamp runs a microtask after the unlocking call, never inside it: `enableWorkflowModification` enables undo/redo after it emits, and the stream still has other subscribers to reach, so a nested disable would leave them on the stale "true"; run afterwards, the disable is the last word and every consumer sees one locked state. The execute service flips the lock before it emits the new state; by the time the clamp looks, the execution-state handler has re-applied the same rule with the final state, so in edit mode the unlock stands. The step panel follows the same rule (`panelLive`), not edit mode alone: the property frame does not consult the lock before its own writes (the version sync on mount, the schema defaults ajv fills in, the editing marker), so a step selected while a run is in flight stays a read-only, inert mount even in edit mode and turns live when the run ends. Done dismisses the step panel before leaving edit mode, while the frame is still an editor: that is the only state in which the property editor clears the "currently editing" marker co-editors see. - The embedded preview grows no editing buttons: no link tools on hover (remove, breakpoint), and a selected operator unfolds its state and port counts but not its delete, chat and add/remove-port buttons. None of them can act on the structure-locked preview, so they only suggested it could be edited. - Open canvas saves first and hands over only once the save has completed: the switch is a full-page load, which aborts a request still in flight. Saves go out one at a time, in order: two persists in flight at once can reach the backend out of order and the older content would win, so the switch's save waits for an autosave already on its way, and the page hands over only once the queue has drained: that save and any asked for while it was in flight (the page stays interactive until the hand-over) have completed; each request carries the workflow as it was when the save was asked for, so the last one enqueued is the latest. The drain outlives the page: the final save on the way out joins the same queue rather than racing an autosave still in flight, and the queue is closed after it. A failed save keeps the author on the form with the error shown and does not stop the queue; a reader with nothing to save goes straight through. A save's response feeds back the server-owned metadata (the timestamp, and the normalised name when nothing changed) but never undoes a rename made while it was in flight, and repaints nothing once the page is gone. (apache#8456 does the save-then-navigate part on the canvas side.) - Keyboard focus is visible on the pill and tab buttons (`:focus-visible`). The picker's pills are tracked by step (`trackBy`): a toggle re-reads the config and rebuilds the choices as new objects, and re-created buttons would have dropped the keyboard focus from the pill just pressed. The instruction body carries an `aria-label`, since its placeholder is gone as soon as there is text. - `29c54bb85` (test only, workspace menu): the menu's export test module-mocked the CommonJS `file-saver` package with `vi.mock`. Under the Angular unit-test builder that call is not reliably hoisted (Vitest warns about it on every platform), and with this stack's shared-chunk graph it stopped applying on the Windows leg (7 of the last 8 Windows runs of the unsplit PR, while main passes). The export now goes through the existing injectable `FileSaverService`, as the dashboard downloads already do, and the spec stubs that with TestBed. No behaviour change. - `695c2fb4f` (test only, Form View rendered spec): the spec swapped the property panel for a stub by overriding the page's imports, which JIT-recompiles the page; a JIT template has no mapping back to the `.component.html`, so the page's template read as 0% covered on codecov from apache#8442 on (the 10 "missing" template lines there, 132 on the unsplit PR). The real panel's template is blanked instead, with its lifecycle hooks switched off, so the page stays AOT-compiled and its template is measured again. The edit-mode markup this PR adds is then covered through the DOM: Edit / Done, the Write / Preview tabs, heading and body writing through, the picker's pills and empty hint. The diff is about 1400 added lines because 725 of them are spec against 694 of source (of which 192 are stylesheet); the two test-only commits and the review-driven redesign of the picker are a large part of it, and the feature itself is under 700 lines. Not in this PR: renaming, hiding, reordering, help text and removal of the exposed inputs, and the author's view of a broken input. Those are apache#8517. ### Any related issues, documentation, discussions? Closes apache#8026. Part of the Form View feature (parent issue apache#8011). Replaces the first half of apache#8455. ### How was this PR tested? Unit tests (vitest). Direct-construction tests cover the authoring gate (enter, leave, refused without write access, always allowed to leave), the result picker's range (final steps until the author chooses, then exactly the saved list, `[]` included; viewed and listed intermediates; disabled steps left out), its live rebuild on an eye toggle, the edit-mode toggle writing the default (the saved list started from the final steps on the first choice), a viewer's own toggle writing nothing and giving way on entering edit mode, a reader not offered a saved pick whose eye is off, the lock clamped back whenever anything else unlocks the graph outside edit mode, unlocked in edit mode once a run has ended in the execute service's real order (unlock before state), and kept locked when edit mode is entered mid-run, the switch's save queued behind an autosave in flight with navigation after both, the final save on the way out drained behind an autosave in flight after the page is gone, a failed save not stopping the queue, a failed final save still reported without throwing, an older save's response not undoing a rename made meanwhile and no metadata repaint after the page is gone, a shown step dropped from the cards and the picker the moment the graph's shape makes it unavailable, a pill renamed the moment its step's display name changes, the hand-over waiting for a save queued behind the switch's and staying put when that save fails, the step panel dismissed on Done while the frame is still an editor, the instruction writes going through the binding service, and the switch's save-then-navigate order (navigates on complete, stays on error, straight through for a reader). The rendered spec covers the edit-mode header shape (title input outside the toggle, `aria-controls`), the panel turning live in edit mode (editor on, tick boxes on, inert off), and the panel staying read-only while a run is in flight even in edit mode, turning live when the run ends. The context menu's spec renders it under a structure lock with modification enabled and checks cut, paste, delete, disable and enable are off while copy and the result toggle stay; the editor's spec opens the right-click menu on a locked editor and checks the lock arrived, and checks a locked preview grows no link tools on hover or on add, and keeps a selected operator's delete, chat and port buttons hidden while its state still unfolds, while the canvas shows all of them. The binding service's spec checks the first choice starts the list from the default handed in, later choices flip within the list ignoring the default, and turning off the only shown step leaves `[]` rather than an absent list; the action service's spec checks a definition whose only content is `shownResultIds: []` is still carried in the saved content. Each new guard was deletion-checked (removing it turns the corresponding test red). eslint, prettier and the production (AOT) build pass; every changed line, template lines included, is statement and function covered (5830 tests). #### Video https://github.com/user-attachments/assets/ca2f8e99-357a-48f7-95cb-6bf4e57c1bc4 ### Was this PR authored or co-authored using generative AI tooling? Yes. Generated-by: Claude Code (Claude Fable 5.1, Anthropic). Co-authored with Claude, reviewed line by line by the author before submission. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY --------- Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com> Co-authored-by: Meng Wang <mengw15@uci.edu> Co-authored-by: mengw15 <125719918+mengw15@users.noreply.github.com>

### What changes were proposed in this PR? Closes apache#8027. Part of the Form View stack (parent issue apache#8011), on main now that apache#8516 (the edit mode) has merged. This is the second half of what was apache#8455 (closed, split in two at 2200 lines). The review commit is the branch's single commit, on main. Lets the author shape the exposed inputs where they appear, in the edit mode apache#8516 adds. - Rename an input or any of its sub-fields, and hide a sub-field, through a new `editable-label-wrapper` formly wrapper: the label is the input, so what the author types is what the reader reads; leaving it empty falls back to the schema label (captured before the stored override is applied, so a renamed sub-field's placeholder and tooltip promise what clearing the box really yields). The control keeps a programmatic name in both modes (a visually hidden `<label for>` while authoring, the static label for readers); a repeated field has no labelable control carrying its id, so its title names the rows as a group (`role="group"` + `aria-labelledby`) instead of pointing a label at nothing. Overrides are stored per binding and reapplied on every rebuild. In a repeated section every row shares one override, so the controls (name box, eye) sit on the first row only; later rows show the same name and hidden state statically and follow the first row's edits at once, and a scalar array's rows are walked as rows, never as the input's root, so the input's own title box appears once above them. - Reorder inputs by drag (CDK drag-drop, mapped by binding id so a card the config no longer holds cannot move the wrong field) or from the keyboard with Move up / Move down on each card, disabled at the ends. - Write per-input help text (saved without rebuilding the form on every keystroke), see "From X" attribution while authoring, and remove an input. A rename is written through on every keystroke (input, not change), so the hidden label and the saved name follow the box as it is typed and nothing is lost if the page is left while the box still has the focus. A step renamed on the canvas or in the live panel (a co-editor's rename included) rebuilds the cards, held while the reader is typing like the compilation path, so the "From X" attribution never goes stale; nothing else emits for a rename. - A binding whose operator no longer exists renders as a removable card carrying the reason for the author and is hidden from readers; it is never deleted silently on a re-read. - A repeated (array) input keeps its title above its rows in reader mode too: the shared array widget prints its label at the bottom beside its add button, so without this the title jumped from above the rows in edit mode to below them on Done. - Keyboard focus is visible on the Move, Remove and eye buttons (`:focus-visible`), and it is never dropped by the edit it triggers: a rename, a hide or help text is presentation only and is shown by the control that took it, and the page does not rebuild on such a write's own announcement: every config write announces on `formBindingChanged$` and the page rebuilds on that stream, so a presentation write is marked while it is made and its announcement skipped (it still reaches the autosave), while structural writes (expose from the panel, remove, reorder) rebuild as before, once (the callers that re-read themselves mark their write too). The test harness's form-binding mock now announces like the real service, so this chain is under test rather than severed; the Move buttons at the ends are `aria-disabled` rather than disabled, so a move that reaches the top or bottom keeps the focus on the button; after Remove the focus goes to the next card's Remove, else the previous one's, else the Inputs heading. The Move and Remove buttons are named with their input (every card has the same three), and the eye is a proper toggle (constant name, state in `aria-pressed`). The diff is about 1650 added lines because 830 of them are spec against 829 of source (of which 246 are stylesheet); the wrapper and the card's author row are one feature, and the wrapper's four files plus the review-driven fixes (once-per-path controls, self-reflected writes) are the bulk of the rest. ### Any related issues, documentation, discussions? Closes apache#8027. Part of the Form View feature (parent issue apache#8011). Replaces the second half of apache#8455. ### How was this PR tested? Unit tests (vitest). Direct-construction tests cover rename and hide overrides reaching the rendered fields (root title and sub-fields, keyed by path), the cards rebuilt when a step's display name changes, the name box writing through on each input event, a repeated section's controls on its first row only with later rows following a rename and a hide, the schema label kept as the fallback of an already renamed sub-field, a scalar array's rows walked as rows (one title box), the page not rebuilding on its own presentation writes though each is announced while a structural announcement still rebuilds, and the wrapper naming a repeated field as a group and fading a hidden field in a follower row, drag and keyboard reorder with the id-mapped indices (including the end stops and a card the config no longer holds), help text saved without a rebuild, broken bindings shown to an author only and kept for explicit removal, and the repeated input's static title. The rendered spec drives the card's author row through the DOM: the provenance line and drag handle, the Move buttons in both directions with their end states, the help-text box writing through, Remove, the broken card's reason, and the drop hand-off to `onDrop`. The editable-label wrapper has its own TestBed spec (real component: decorate, the name box's change renaming, the eye hiding, the reader's label association and the hidden label while authoring). Each new guard was deletion-checked (removing it turns the corresponding test red). eslint, prettier and the production (AOT) build pass; every changed line, template lines included, is statement and function covered. #### video https://github.com/user-attachments/assets/566f8bd7-5596-4085-9317-79974cfdfad2 ### Was this PR authored or co-authored using generative AI tooling? Yes. Generated-by: Claude Code (Claude Fable 5.1, Anthropic). Co-authored with Claude, reviewed line by line by the author before submission. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01FVvP3ttj22f9LB4p9u2anY Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>

github-actions Bot assigned yangzhang75 Sep 9, 2026

github-actions Bot added the frontend Changes related to the frontend GUI label Sep 9, 2026

yangzhang75 force-pushed the formview-pr15 branch 2 times, most recently from 9913052 to e8ffb24 Compare September 9, 2026 06:49

This was referenced Sep 9, 2026

feat(gui): wire the Form View entry points #8456

Merged

Form View: a form-based view for running a workflow #8011

Open

mengw15 requested a balanced review from Copilot September 9, 2026 17:54

Copilot AI reviewed Sep 9, 2026

View reviewed changes

Copilot started reviewing on behalf of mengw15 September 9, 2026 21:09 View session

yangzhang75 force-pushed the formview-pr15 branch 5 times, most recently from 565dfda to 8f0570b Compare September 10, 2026 23:30

yangzhang75 force-pushed the formview-pr15 branch 3 times, most recently from f13abfa to fe33f0c Compare September 11, 2026 02:31

yangzhang75 marked this pull request as ready for review September 11, 2026 02:34

github-actions Bot requested a review from mengw15 September 11, 2026 02:35

mengw15 requested a balanced review from Copilot September 11, 2026 16:40

Copilot started reviewing on behalf of mengw15 September 11, 2026 16:41 View session

Copilot AI reviewed Sep 11, 2026

View reviewed changes

Comment thread frontend/src/app/workspace/component/workflow-form/workflow-form.component.ts

Comment thread frontend/src/app/common/formly/editable-label-wrapper/editable-label-wrapper.component.html

yangzhang75 force-pushed the formview-pr15 branch from fe33f0c to 818ec4f Compare September 11, 2026 17:11

yangzhang75 and others added 2 commits September 11, 2026 10:22

yangzhang75 force-pushed the formview-pr15 branch from 818ec4f to 01d0e2d Compare September 11, 2026 17:41

This was referenced Sep 11, 2026

feat(gui): add the Form View edit mode, with instruction and result authoring #8516

Merged

feat(gui): author the Form View's inputs in place #8517

Merged

yangzhang75 closed this Sep 11, 2026

Navigation Menu

Uh oh!

feat(gui): author the Form View in place with an edit mode - #8455

feat(gui): author the Form View in place with an edit mode#8455
yangzhang75 wants to merge 2 commits into
apache:mainfrom
yangzhang75:formview-pr15

yangzhang75 commented Sep 9, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 9, 2026 •

edited

Loading

Uh oh!

codecov-commenter commented Sep 9, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

yangzhang75 commented Sep 11, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

yangzhang75 commented Sep 11, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

yangzhang75 commented Sep 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What changes were proposed in this PR?

Any related issues, documentation, discussions?

How was this PR tested?

video

Was this PR authored or co-authored using generative AI tooling?

Uh oh!

github-actions Bot commented Sep 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Automated Reviewer Suggestions

Uh oh!

codecov-commenter commented Sep 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

yangzhang75 commented Sep 11, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟡 Changes recommended

Suppressed comments (4)

Uh oh!

Uh oh!

Uh oh!

yangzhang75 commented Sep 11, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

yangzhang75 commented Sep 9, 2026 •

edited

Loading

github-actions Bot commented Sep 9, 2026 •

edited

Loading

codecov-commenter commented Sep 9, 2026 •

edited

Loading