fix: generate Specta bindings via build script

[Scott McMaster (scottmcmaster)]

Summary

Somebody moved the location of the specta-generated .ts files and changed the name of one and didn't update the regeneration hashing script, as evidenced by the increased dev build startup time and the diagnostic messages that appear.

Test Plan

Run the dev build, watch the logs.

Docs

• Docs updated
• No docs update needed

[github-actions]

	Warnings
⚠️	Please assign this PR to someone (usually yourself).

📋 PR Overview

Lines changed	8 (+4 / -4)
Files	0 added, 1 modified, 0 deleted
Draft / WIP	no
Has Test Plan	yes
New UI components	no
New Storybook stories	no
New Rust modules	no
New TS source files	no
New tests	no
package.json touched	no
Cargo.toml touched	no
Infra / CI touched	no

🔬 Coverage

Report	Lines	Statements	Functions	Branches
apps/native/coverage/coverage-summary.json	17.9%	17.9%	30.9%	54.1%

Generated by 🚫 dangerJS against da512f3

[Copilot]

Pull request overview

This PR fixes the Tauri dev startup regeneration check so it looks for the current Specta-generated TypeScript files under src/ipc, avoiding unnecessary regeneration on every dev launch.

Changes:

• Updates the SQLite generated file check from src/types/sqlite.ts to src/ipc/sqlite.ts.
• Updates the shared generated types check from src/types/shared.ts to src/ipc/types.ts.
• Aligns diagnostic “missing file” messages with the corrected paths.

---

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

[Cas Linden (CasLinden)]

The hashed files now just contain re-exports, so re-enabling the optimisation would break generation when needed.

There's also agents manually editing the output files. Breakage, merge conflicts and other issues would be less likely if we just regenerate.

I see two options:

1. hashing all files
2. doing away with the optimisation

I'd say 1 just expands the maintenance required, requiring more frequent additions to this script in the future. Is it worth the ~1-2s every run?

[Cas Linden (CasLinden)]

I suppose technically you could hash all files present in the designated /shared dir, but with the agent thing, and maintenance like this PR, which would have caused an issue down the line, I think the case for removal is strong.

[Scott McMaster (scottmcmaster)]

The hashed files now just contain re-exports, so re-enabling the optimisation would break generation when needed.

There's also agents manually editing the output files. Breakage, merge conflicts and other issues would be less likely if we just regenerate.

I see two options:

1. hashing all files
2. doing away with the optimisation

I'd say 1 just expands the maintenance required, requiring more frequent additions to this script in the future. Is it worth the ~1-2s every run?

Sorry, I don't understand this point. The script is just for dev builds to not waste multiple seconds on every startup. I'm not seeing the issue, but maybe if you clarify I can think of a way to fix it, because technically there always should be a way to make this work in any sane build system.

[Cas Linden (CasLinden)]

The hashed files now just contain re-exports

Look at the contents of shared_types.rs

this is just a typo

[Cas Linden (CasLinden)]

Unqueued this.

[Cas Linden (CasLinden)]

I have an explanation ready, let me just prove it / check I am not just wrong.

[Cas Linden (CasLinden)]

Scott McMaster (@scottmcmaster)

This commit changed shared_types.rs from a file that contains the types, to one that re-exports them.

So since then (weeks before the change this PR points to), changing types in nested files in the /shared leads to false negatives as the type files themselves are not what the script hashes and therefore checks for changes.

Here's two scenarios I ran on the branch with screenshots:

Scenario 1: New type is added, requires addition in specta_gen_ts.rs, which is also hashed. So it works.
Scenario 2: An existing type is changed, no hashed files change, script emits false negative.

[Cas Linden (CasLinden)]

Scott McMaster (@scottmcmaster)

please cherry pick: fdff307

Or close this and merge this