feat(#448): support encrypting submission payload by garethbowen · Pull Request #766 · getodk/web-forms

garethbowen · 2026-04-02T04:38:37Z

Closes #448

I have verified this PR works in these browsers (latest versions):

What else has been done to verify that this works as intended?

Minimal testing. I'd really like to add an e2e test one day that actually verifies that the submission can be decrypted, but that requires actually starting up backend. If we end up in the same repo as frontend then there already is a framework to do this so postponing it until we figure out where we're going to go!

Why is this the best possible solution? Were any other approaches considered?

How does this change affect users? Describe intentional changes to behavior and behavior that could have accidentally been affected by code changes. In other words, what are the regression risks?

Do we need any specific form for testing your changes? If so, please attach one.

No specific form but it only works in central (not demo) and only if you set up encryption as laid out in the test plan.

What's changed

changeset-bot · 2026-04-02T04:38:40Z

🦋 Changeset detected

Latest commit: 0b4a7c1

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 5 packages

Name	Type
@getodk/xforms-engine	Minor
@getodk/scenario	Minor
@getodk/common	Minor
@getodk/web-forms	Minor
@getodk/xpath	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

garethbowen · 2026-04-02T04:39:20Z

packages/xforms-engine/src/parse/model/SubmissionManifestDefinition.ts

+		// 	);
+		// 	el3.textContent = this.signature;
+		// 	manifest.appendChild(el3);
+		// }


This part is still waiting on a decision. I suspect it's going to get removed completely because it's a) optional, and b) not implemented in central.

garethbowen · 2026-04-02T04:54:25Z

@latin-panda This is ready for review. I'm still waiting on a couple of answers and it'll need a fair amount of QA testing, but as I'm AFK for the next week I thought I'd send it over for review in the meantime. Thanks!

latin-panda

Really interesting! I left some comments below

latin-panda · 2026-04-08T23:08:15Z

packages/xforms-engine/src/lib/client-reactivity/instance-state/quarantine/README.md

+
+## Implementation
+
+The symmetric encryption parts of the spec are implemnted using CryptoJS because the particularly algorithm required by the spec is not supported by Subtle Crypto, and we use CryptoJS elsewhere.


Suggested change

The symmetric encryption parts of the spec are implemnted using CryptoJS because the particularly algorithm required by the spec is not supported by Subtle Crypto, and we use CryptoJS elsewhere.

The symmetric encryption parts of the spec are implemented using CryptoJS because the particular algorithm required by the spec is not supported by Subtle Crypto, and we use CryptoJS elsewhere.

latin-panda · 2026-04-08T23:35:17Z

packages/xforms-engine/src/lib/client-reactivity/instance-state/quarantine/symmetric.ts

+	symmetricKey: Uint8Array<ArrayBuffer>,
+	seed: Seed
+): Promise<File> => {
+	const content = await getBlobData(attachment);


I'm curious about the amount of memory WF is allocating to submissions overall, including attachments and raw data, since it loads the submission into memory again. It'd be good to check how large forms perform on low-end devices, as part of the test plan.

latin-panda · 2026-04-08T23:41:54Z

packages/xforms-engine/src/lib/client-reactivity/instance-state/quarantine/symmetric.ts

+	const bytes = [];
+	for (let i = 0; i < buffer.length; i += 4) {
+		bytes.push(
+			(buffer[i]! << 24) | (buffer[i + 1]! << 16) | (buffer[i + 2]! << 8) | buffer[i + 3]!


This increases by 4, and it will go out of bounds. However, the bitwise operation treats undefined as 0. It would be better to still check that buffer[i], buffer[i + 1], etc., are defined to make it more robust.

latin-panda · 2026-04-08T23:45:17Z

packages/xforms-engine/src/lib/client-reactivity/instance-state/quarantine/symmetric.ts

+		readonly instanceId: string,
+		readonly symmetricKey: Uint8Array<ArrayBuffer>
+	) {
+		const key = CryptoJS.lib.WordArray.create(symmetricKey);


Does it make sense to use this function here too?

Suggested change

const key = CryptoJS.lib.WordArray.create(symmetricKey);

const key = arrayBufferToWordArray(symmetricKey);

latin-panda · 2026-04-08T23:50:50Z

packages/xforms-engine/src/parse/model/SubmissionManifestDefinition.ts

+		const el = document.createElementNS(ODK_SUBMISSIONS_NAMESPACE_URI, 'base64EncryptedKey');
+		el.textContent = this.base64EncryptedKey;
+		manifest.appendChild(el);
+
+		const el2 = document.createElementNS(ODK_SUBMISSIONS_NAMESPACE_URI, 'encryptedXmlFile');
+		el2.textContent = ENCRYPTED_SUBMISSION_ATTACHMENT_NAME;
+		manifest.appendChild(el2);


Suggested change

const el = document.createElementNS(ODK_SUBMISSIONS_NAMESPACE_URI, 'base64EncryptedKey');

el.textContent = this.base64EncryptedKey;

manifest.appendChild(el);

const el2 = document.createElementNS(ODK_SUBMISSIONS_NAMESPACE_URI, 'encryptedXmlFile');

el2.textContent = ENCRYPTED_SUBMISSION_ATTACHMENT_NAME;

manifest.appendChild(el2);

const keyEl = document.createElementNS(ODK_SUBMISSIONS_NAMESPACE_URI, 'base64EncryptedKey');

el.textContent = this.base64EncryptedKey;

manifest.appendChild(el);

const xmlFileEl = document.createElementNS(ODK_SUBMISSIONS_NAMESPACE_URI, 'encryptedXmlFile');

el2.textContent = ENCRYPTED_SUBMISSION_ATTACHMENT_NAME;

manifest.appendChild(el2);

latin-panda · 2026-04-08T23:59:01Z

packages/xforms-engine/src/parse/model/SubmissionManifestDefinition.ts

+import {
+	ODK_SUBMISSIONS_NAMESPACE_URI,
+	OPENROSA_XFORMS_NAMESPACE_URI,
+} from '@getodk/common/constants/xmlns.ts';


How about moving this to /instance-state? It seems like this file is handling a lot of DOM element work, and currently parse/model isn't importing from client-reactivity. What do you think?

latin-panda · 2026-04-09T00:02:52Z

packages/xforms-engine/src/lib/client-reactivity/instance-state/quarantine/symmetric.ts

+
+class Seed {
+	readonly ivSeedArray;
+	counter = 0;


I think it can be private

Suggested change

counter = 0;

private counter = 0;

garethbowen added 14 commits April 2, 2026 13:45

feat(#448): support encrypting submission payload

7b1f510

backup

721dea4

progress

4b3a81c

iv works

1c5763a

correct instanceid

d0d8bd0

successful decryption!

5aedd2f

quaratine the dangerous code, and clean things up

085e470

fix compile step

399aa84

encrypt attachments also

f48efc7

make test pass

7fd202f

refactor

9b0a118

add end of file new line

f10b284

lint

bcdc8c3

define significant bytes to remove erroneous padding for attachments

78d1796

garethbowen commented Apr 2, 2026

View reviewed changes

garethbowen added 2 commits April 2, 2026 17:44

cleanup

ad727e5

changeset and feature matrix

0b4a7c1

garethbowen marked this pull request as draft April 2, 2026 04:54

garethbowen requested a review from latin-panda April 2, 2026 04:54

latin-panda requested changes Apr 9, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(#448): support encrypting submission payload#766

feat(#448): support encrypting submission payload#766
garethbowen wants to merge 16 commits intomainfrom
448-submission-encryption

garethbowen commented Apr 2, 2026

Uh oh!

changeset-bot bot commented Apr 2, 2026 •

edited

Loading

Uh oh!

garethbowen Apr 2, 2026

Uh oh!

garethbowen commented Apr 2, 2026

Uh oh!

latin-panda left a comment

Uh oh!

latin-panda Apr 8, 2026

Uh oh!

latin-panda Apr 8, 2026

Uh oh!

latin-panda Apr 8, 2026

Uh oh!

latin-panda Apr 8, 2026

Uh oh!

latin-panda Apr 8, 2026

Uh oh!

latin-panda Apr 8, 2026

Uh oh!

latin-panda Apr 9, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants


		## Implementation

		The symmetric encryption parts of the spec are implemnted using CryptoJS because the particularly algorithm required by the spec is not supported by Subtle Crypto, and we use CryptoJS elsewhere.

	const key = CryptoJS.lib.WordArray.create(symmetricKey);
	const key = arrayBufferToWordArray(symmetricKey);

Conversation

garethbowen commented Apr 2, 2026

I have verified this PR works in these browsers (latest versions):

What else has been done to verify that this works as intended?

Why is this the best possible solution? Were any other approaches considered?

How does this change affect users? Describe intentional changes to behavior and behavior that could have accidentally been affected by code changes. In other words, what are the regression risks?

Do we need any specific form for testing your changes? If so, please attach one.

What's changed

Uh oh!

changeset-bot bot commented Apr 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🦋 Changeset detected

Uh oh!

Choose a reason for hiding this comment

Uh oh!

garethbowen commented Apr 2, 2026

Uh oh!

latin-panda left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

changeset-bot bot commented Apr 2, 2026 •

edited

Loading