Probectors by 7h3kk1d · Pull Request #1998 · hazelgrove/hazel

7h3kk1d · 2025-11-03T14:29:17Z

Adds a rich "projector-style" display on probed values. Also allows for the underlying syntax to be rewritten by the probed interface. Currently there's a calculator for integers, table editor, and card view.

Calculator

Screencast.From.2025-11-21.11-12-44.webm

Tables

The rich table probe allows for some column actions to rewrite the underlying syntax via function pipelining. If placed on a pattern the table is read-only
Screencast From 2025-11-21 11-16-20.webm

- Does not work with ephemeral refractor models

…refractor

Move suit, rank, card, and hand type definitions, along with utility functions like card_to_exp, string_to_suit, and conversion helpers, from CardProj.re and CardRenderer.re into new CardTypes and CardSyntax modules. This reduces code duplication and improves maintainability without altering functionality.

Implement on_mousedown event handlers for Singleton and Hand card views to toggle between Choose, Flipped, and Show modes. Double-click or shift-click adjusts mode accordingly, enhancing user interaction by allowing direct mode changes via mouse events.

Add CalculatorRenderer.rei defining types for calculator operations, state, and actions, including RichProbe integration. Update ProbeProj.re to reference CalculatorRenderer directly instead of CalculatorRenderer.M for correct projector registration.

- Enhanced renderer selection to only apply a renderer if it can handle the current expression, preventing potential rendering failures when expressions change. - Updated switch case from specific `Some(renderer)` to guarded `Some(renderer) when renderer.can_handle(exp)`, and changed fallback from `None` to `_` for clarity.

…r CalculatorRenderer and CardRenderer - Add `value` types (`int` for Calculator, `state` for Card) - Implement `parse` functions to extract values from expressions - Refactor `init` functions to accept parsed values instead of expressions - Update `render` functions to use pre-parsed values; remove redundant parsing - Reorder type definitions for clarity in CalculatorRenderer

…refractor

src/haz3lcore/projectors/implementations/ProbeProj.re

…low for term creation

codecov · 2026-02-04T19:17:39Z

Codecov Report

❌ Patch coverage is 8.16754% with 877 lines in your changes missing coverage. Please review.
✅ Project coverage is 47.57%. Comparing base (55c94b4) to head (b703642).

Files with missing lines	Patch %	Lines
...3lcore/projectors/implementations/TableRenderer.re	3.05%	413 Missing ⚠️
.../haz3lcore/projectors/implementations/ProbeProj.re	4.16%	115 Missing ⚠️
...e/projectors/implementations/CalculatorRenderer.re	3.63%	106 Missing ⚠️
...z3lcore/projectors/implementations/CardRenderer.re	9.82%	101 Missing ⚠️
.../haz3lcore/projectors/implementations/CardTypes.re	27.27%	32 Missing ⚠️
src/haz3lcore/projectors/ProjectorPerform.re	0.00%	31 Missing ⚠️
...haz3lcore/projectors/implementations/CardSyntax.re	38.46%	16 Missing ⚠️
...c/haz3lcore/projectors/implementations/CardUtil.re	0.00%	13 Missing ⚠️
src/language/builtins/BuiltinsBase.re	57.14%	12 Missing ⚠️
...c/haz3lcore/projectors/implementations/CardProj.re	0.00%	9 Missing ⚠️
... and 10 more

Additional details and impacted files

@@                   Coverage Diff                   @@
##           dataframe-projector    #1998      +/-   ##
=======================================================
- Coverage                50.30%   47.57%   -2.73%     
=======================================================
  Files                      231      237       +6     
  Lines                    25450    26282     +832     
=======================================================
- Hits                     12802    12504     -298     
- Misses                   12648    13778    +1130

Files with missing lines	Coverage Δ
...z3lcore/projectors/implementations/CSVProjector.re	`0.00% <ø> (ø)`
...z3lcore/projectors/implementations/CheckboxProj.re	`16.66% <ø> (ø)`
...az3lcore/projectors/implementations/LivelitProj.re	`4.76% <ø> (ø)`
...az3lcore/projectors/implementations/SliderFProj.re	`16.66% <ø> (ø)`
...haz3lcore/projectors/implementations/SliderProj.re	`16.66% <ø> (ø)`
.../haz3lcore/projectors/implementations/TableProj.re	`19.71% <ø> (ø)`
...z3lcore/projectors/implementations/TextAreaProj.re	`7.57% <ø> (ø)`
...c/haz3lcore/projectors/implementations/TypeProj.re	`5.76% <ø> (ø)`
src/haz3lcore/zipper/Refractors.re	`50.00% <80.00%> (+2.17%)`	⬆️
src/haz3lcore/zipper/action/Action.re	`0.00% <0.00%> (ø)`
... and 18 more

... and 39 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

7h3kk1d · 2026-02-04T19:29:07Z

src/web/app/common/RefractorView.re

+        Refractors.to_projector(
+          Option.value(
+            TermData.segment(id, term_data)
+            |> Option.map(Segment.unparenthesize)
+            |> Option.map(Segment.trim_secondary(Left))
+            |> Option.map(Segment.trim_secondary(Right))
+            |> Option.map(Segment.parenthesize),
+            ~default=
+              Base.Secondary({
+                id: Id.invalid,
+                content: Whitespace(""),
+              }),
+          ),
+          id,
+          entry,
+        );


@disconcision I need the syntax on refractors so I can use lift_syntax to rewrite the underlying syntax. This is how I have it working after all the recent changes.

The old way I was doing it is here:

hazel/src/web/app/common/ProjectorView.re

Lines 390 to 399 in e9ebe54

let id = projector.p.id;

let seg =

TermData.segment(Id.recover_original(id), syntax.term_data);

/* Add syntax to info */

let info_with_syntax: ProjectorBase.info = {

...projector.info,

syntax: Option.value(seg, ~default=projector.info.syntax),

};

I was similarly pulling the segment out of the termdata map. The old ProjectorCore took a segment and the new on takes a piece so I'm parenthesizing it. I'm also having to trim secondary to avoid issues with parsing it to a term in lift_syntax.

I'm not sure if you have thoughts here on a better way to do this.

…ed clarity

…utput

Copilot

Pull request overview

Copilot reviewed 58 out of 58 changed files in this pull request and generated 10 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-04T19:41:00Z

src/haz3lcore/projectors/ProjectorPerform.re

+        | Some(z) => Ok(z)
+        | None => Error(Cant_project)
+        };
+      | (None, None) => assert(false)


The assert(false) on line 241 indicates a case that is expected to never happen (when a refractor is being processed but neither manual nor ephemeral refractor exists). However, using assert(false) causes the program to crash at runtime if this assumption is violated. Consider using a more graceful error handling approach, such as returning Error(Cant_project) or adding a descriptive error message explaining what invariant was violated.

Suggested change

| (None, None) => assert(false)

| (None, None) =>

// Invariant violation: refractor-kind projector without manual or ephemeral refractor.

// Fail gracefully instead of crashing.

Error(Cant_project)

Copilot · 2026-02-04T19:41:01Z

src/haz3lcore/projectors/implementations/TableRenderer.re

+        | _ =>
+          failwith("TableRenderer: apply_transformation: not an expression"),
+        info.syntax,
+      )
+    ) {
+    | Some(s) => s
+    | None => failwith("TableRenderer: apply_transformation: lift failed")


The failwith calls on lines 163 and 168 will crash the application if the syntax is not an expression or if lift_syntax fails. This could happen with malformed data or unexpected edge cases. Consider returning option types or Result types and handling errors gracefully at the call site, especially since this is user-facing functionality that modifies syntax interactively.

Suggested change

| _ =>

failwith("TableRenderer: apply_transformation: not an expression"),

info.syntax,

)

) {

| Some(s) => s

| None => failwith("TableRenderer: apply_transformation: lift failed")

| other =>

/* If the syntax node is not an expression, leave it unchanged. */

other,

info.syntax,

)

) {

| Some(s) => s

| None =>

/* If lifting fails, fall back to the original syntax instead of crashing. */

info.syntax

Copilot · 2026-02-04T19:41:01Z

src/haz3lcore/projectors/implementations/CalculatorRenderer.re

+          "CalculatorRenderer: apply_arithmetic_operation: not an expression",
+        ),
+      info.syntax,
+    )
+  ) {
+  | Some(s) => s
+  | None =>
+    failwith("CalculatorRenderer: apply_arithmetic_operation: lift failed")


Similar to TableRenderer, the failwith calls on lines 100 and 107 will crash the application. These should handle errors more gracefully, especially since the calculator is an interactive user-facing feature. Return option or Result types and handle errors at the UI level.

Copilot · 2026-02-04T19:41:01Z

src/haz3lcore/projectors/implementations/TableProj.re

+    | None => failwith("TextArea: get: Not a table")
+    }
+  | None => failwith("TextArea: get: Not a table")


The error messages on lines 58 and 60 say "TextArea: get: Not a table" but this is in TableProj.re, not TextAreaProj.re. The error message should be corrected to "TableProj: get: Not a table" to accurately identify where the error originated.

Suggested change

| None => failwith("TextArea: get: Not a table")

}

| None => failwith("TextArea: get: Not a table")

| None => failwith("TableProj: get: Not a table")

}

| None => failwith("TableProj: get: Not a table")

Copilot · 2026-02-04T19:41:02Z

src/haz3lcore/pretty/ExpToSegment.re

+let project_table_if = (should_project, pieces) =>
+  if (should_project) {
+    switch (MakeTerm.for_projection([pieces])) {
+    | None => failwith("ExpToSegment.fold_if")


The error message on line 1054 says "ExpToSegment.fold_if" but it's actually in the project_table_if function. The error message should be updated to "ExpToSegment.project_table_if" to accurately identify the source of the error.

Suggested change

| None => failwith("ExpToSegment.fold_if")

| None => failwith("ExpToSegment.project_table_if")

Copilot · 2026-02-04T19:41:02Z

src/haz3lcore/projectors/implementations/ProbeProj.re

+  switch (probe_model_of_sexp(sexp)) {
+  | model => model
+  | exception _ => {active_renderer: None}


The probe_model_of_sexp function on lines 106-110 appears to call itself recursively (probe_model_of_sexp(sexp) calls probe_model_of_sexp). This creates a custom deserialization function that falls back to a default value on exception. However, the pattern match on line 108 (| model => model) is redundant since it just returns the model unchanged. The function could be simplified to just the exception case, or if the intent is to override the generated deserialization, the naming should be clearer to avoid confusion.

Suggested change

switch (probe_model_of_sexp(sexp)) {

| model => model

| exception _ => {active_renderer: None}

/* Safely deserialize probe_model from sexp, falling back to default on error. */

switch (sexp) {

| Sexplib.Sexp.List(

[Sexplib.Sexp.List([Sexplib.Sexp.Atom("active_renderer"), ar_sexp])]

) =>

/* Decode active_renderer field; if it raises, fall back to None. */

switch (oactive_renderer_of_sexp(ar_sexp)) {

| active_renderer => {active_renderer}

| exception _ => {active_renderer: None}

}

/* If the structure doesn't match the expected record layout, use default. */

| _ => {active_renderer: None}

Copilot · 2026-02-04T19:41:02Z

src/web/www/style/variables.css

  --modal-z: 101;
-}
+
+  /* TABLE TABLES */


The PR title "Probectors" appears to be a portmanteau of "Probe" and "Projector" but this may not be immediately clear to all developers. Consider using a more descriptive title like "Rich Projector Displays for Probed Values" or documenting the term "Probectors" in the codebase if it's intended to be part of the project's terminology.

Copilot · 2026-02-04T19:41:02Z

src/haz3lcore/projectors/implementations/TableRenderer.re

+             action: () =>
+               Effect.Many([
+                 local(CloseMenu),
+                 parent(
+                   SetSyntax(
+                     OptUtil.get_or_fail(
+                       (left ? "move left" : "move right") ++ " failed",
+                       move_column(info, dyn_type, h, left),
+                     ),
+                   ),
+                 ),
+               ]),


The OptUtil.get_or_fail call on line 645 will crash if move_column returns None. Since this is user-facing interactive functionality, the failure should be handled more gracefully - perhaps by showing an error message to the user or disabling the move buttons when the operation is not valid. The current implementation checks can_move_left/can_move_right but then still uses get_or_fail which suggests a logic inconsistency.

Suggested change

action: () =>

Effect.Many([

local(CloseMenu),

parent(

SetSyntax(

OptUtil.get_or_fail(

(left ? "move left" : "move right") ++ " failed",

move_column(info, dyn_type, h, left),

),

),

),

]),

action: () => {

let direction_label = left ? "move left" : "move right";

switch (move_column(info, dyn_type, h, left)) {

| Some(new_syntax) =>

Effect.Many([

local(CloseMenu),

parent(SetSyntax(new_syntax)),

])

| None =>

JsUtil.alert(direction_label ++ " failed");

local(CloseMenu)

};

},

Copilot · 2026-02-04T19:41:03Z

src/web/www/style/variables.css

  --modal-z: 101;
-}
+
+  /* TABLE TABLES */


The CSS comment "TABLE TABLES" on line 230 appears to be a typo or placeholder. It should probably be "TABLE STYLES" or "TABLE VARIABLES" to match the pattern of other section comments in this file (e.g., "HACKS").

Copilot · 2026-02-04T19:41:03Z

test/haz3ltest.re

+    // @ Test_Editing.tests
    @ Test_AutoProbe.tests
-    @ Test_Indentation.tests
+    // @ Test_Indentation.tests


Tests for Test_Editing and Test_Indentation have been commented out without explanation. If these tests are failing or being temporarily disabled during development, this should be noted in the PR description or the tests should be fixed before merging.

Add max-height constraint (350px) with auto overflow-y to modals containing tables. This prevents modals with large tables from extending beyond viewport height and ensures content remains accessible through scrolling.

…projector-refractor

…ize ID membership checks

…lumn menu

…tionality

…projector-refractor

This reverts part of commit 0d6b44e.

…projector-refractor

7h3kk1d and others added 29 commits October 2, 2025 15:06

Merge branch 'probemoar' into dataframe-projector

5ad0a15

Speculative refactor extracting table renderer

704e9f4

Attempt to render table for probes.

adfd919

- Does not work with ephemeral refractor models

Merge remote-tracking branch 'origin/probemoar' into table-projector-…

571591b

…refractor

tables at least rendering

ed5b9af

Merge remote-tracking branch 'origin/dev' into table-projector-refractor

b0cf6c9

reformat

c88cde2

I have broken projectors but made refractors work

05bb739

Fix submenu action to track current column in table modal

6b96eb3

Refactor value_view to use span for table badge

429bed6

Removing table code from probe code

26ea407

comment out debug

15eeaee

Adding RichProbe signature

35c3f01

cleanup

391a300

Temp commit

43c8f93

calculatin

a101ce4

Enhance calculator modal close button styles and positioning

47db3e5

First class modules

b8997ab

halfbaked cards

f3aa2ee

Merge remote-tracking branch 'origin/probemoar' into table-projector-…

bff8aae

…refractor

disable failing tests from probemoar

cf29da8

fix warnings

ae6da67

refactor(CardProj): remove unused imports for CardSyntax and CardUtil

125a225

refactor: remove unused imports across multiple modules

52c5863

7h3kk1d commented Nov 5, 2025

View reviewed changes

src/haz3lcore/projectors/implementations/ProbeProj.re Show resolved Hide resolved

Trim secondary segments when constructing syntax for refractors to al…

3bc417c

…low for term creation

7h3kk1d commented Feb 4, 2026

View reviewed changes

7h3kk1d added 3 commits February 4, 2026 14:29

Refactor build_column_menu to use explicit parameter types for improv…

6c93e95

…ed clarity

Remove debug print statements from lift_syntax function for cleaner o…

6fce622

…utput

Simplify focusable pointer function by removing debug print statement

93462d8

7h3kk1d requested a review from Copilot February 4, 2026 19:34

Copilot started reviewing on behalf of 7h3kk1d February 4, 2026 19:34 View session

Copilot AI reviewed Feb 4, 2026

View reviewed changes

7h3kk1d added 2 commits February 4, 2026 14:43

style(projectors): add scroll constraint to table modals

8d0567c

Add max-height constraint (350px) with auto overflow-y to modals containing tables. This prevents modals with large tables from extending beyond viewport height and ensures content remains accessible through scrolling.

feat(TableRenderer): add "Group By" action to column menu

69e9be5

7h3kk1d changed the base branch from dev to dataframe-projector February 4, 2026 20:01

7h3kk1d and others added 18 commits February 4, 2026 15:29

Merge remote-tracking branch 'origin/dataframe-projector' into table-…

92bddf1

…projector-refractor

Fix styling

2e52488

Merge remote-tracking branch 'origin/dataframe-projector' into table-…

ee2f5bf

…projector-refractor

Make auto probe models persistent

52147a3

Refactor add_ids_from_auto_term to preserve existing models and optim…

8b43d5a

…ize ID membership checks

Merge branch 'dataframe-projector' into table-projector-refractor

b2acebd

Add option type handling: drop Nones and provide default values in co…

ba7870b

…lumn menu

Add clear and noop column transformations to enhance column menu func…

9f9075a

…tionality

Remove comments

a0438b8

Merge remote-tracking branch 'origin/dataframe-projector' into table-…

db04bff

…projector-refractor

Update table rich probe to match projector

aff6787

Readd comparison functions

144cb72

This reverts part of commit 0d6b44e.

Fix error message in get function for table retrieval

6383832

Merge branch 'dataframe-projector' into table-projector-refractor

8f32034

Merge branch 'dataframe-projector' into table-projector-refractor

62955e2

Merge remote-tracking branch 'origin/dataframe-projector' into table-…

fe67545

…projector-refractor

Merge remote-tracking branch 'origin/dataframe-projector' into table-…

47f1e7a

…projector-refractor

Merge remote-tracking branch 'origin/dataframe-projector' into table-…

b703642

…projector-refractor

	let id = projector.p.id;

	let seg =
	TermData.segment(Id.recover_original(id), syntax.term_data);

	/* Add syntax to info */
	let info_with_syntax: ProjectorBase.info = {
	...projector.info,
	syntax: Option.value(seg, ~default=projector.info.syntax),
	};

	\| None => failwith("ExpToSegment.fold_if")
	\| None => failwith("ExpToSegment.project_table_if")

-  switch (probe_model_of_sexp(sexp)) {
-  | model => model
-  | exception _ => {active_renderer: None}
+  /* Safely deserialize probe_model from sexp, falling back to default on error. */
+  switch (sexp) {
+  | Sexplib.Sexp.List(
+      [Sexplib.Sexp.List([Sexplib.Sexp.Atom("active_renderer"), ar_sexp])]
+    ) =>
+    /* Decode active_renderer field; if it raises, fall back to None. */
+    switch (oactive_renderer_of_sexp(ar_sexp)) {
+    | active_renderer => {active_renderer}
+    | exception _ => {active_renderer: None}
+    }
+  /* If the structure doesn't match the expected record layout, use default. */
+  | _ => {active_renderer: None}

Comments

Conversation

7h3kk1d commented Nov 3, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Calculator

Tables

Uh oh!

Uh oh!

codecov bot commented Feb 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

7h3kk1d Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 4, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

7h3kk1d commented Nov 3, 2025 •

edited

Loading

codecov bot commented Feb 4, 2026 •

edited

Loading