Identity Map V3 #83

aulme · 2025-06-03T06:12:53Z

No description provided.

… 1 day buffer to remove flakiness

src/test/java/com/uid2/client/IdentityMapV3IntegrationTests.java

src/main/java/com/uid2/client/IdentityMapV3Response.java

jon8787 · 2025-06-06T05:14:54Z

src/test/java/com/uid2/client/IdentityMapV3IntegrationTests.java

+
+        Response response = new Response(identityMapInput);
+
+        response.assertUnmapped("INVALID", "this is not a hashed email");


curious, why have the errors changed to CAPS and a single word?

That's the code we return from the APIs. I don't know why we've changed the codes from the previous version. Perhaps @gmsdelmundo has some context one he's back next week. I didn't realize it was not always like that.
I guess this makes it look more like an error code and less like an end-user message.

So this means error handling is no longer a drop-in replacement? Especially concerned about optout - if we change this to caps we will break someone.

Good point, but a discussion on the API side, not in SDK - here we just return whatever APIs return.
We'll consider changing this in API.

@jon8787 after some discussion, we've decided to keep the APIs as is and change the SDKs to return an enum. Any unknown unmaped statuses will become "UNKNOWN".
This will still be a breaking change, but it will be obviously breaking instead of silently breaking with changing strings.

Returning an enum sounds good, although will there still be a way to get the unknown status?

Not very convinced about changing lowercase "optout" to CAPS. There may be consumers using the API directly who now have to change their optout handling when migrating to v3. What's the benefit?

I'm planning to expose the rawReason String together with reason. We'll use the Enum in all examples but if people want to examine the String, they can.
I've made parsing to enum work case-insensitive so we can decide on the casing outside of this PR.

src/main/java/com/uid2/client/IdentityMapV3Response.java

src/main/java/com/uid2/client/IdentityMapV3Helper.java

mcollins-ttd · 2025-06-10T03:51:04Z

src/main/java/com/uid2/client/IdentityMapV3Input.java

+        return new IdentityMapV3Input().withHashedPhones(hashedPhones);
+    }
+
+    private transient final Map<String, List<String>> diiMappings = new HashMap<>();


Can you add a comment about the use of transient here?

mcollins-ttd · 2025-06-10T03:52:00Z

src/main/java/com/uid2/client/IdentityMapV3Input.java

+    @SerializedName("phone_hash")
+    private final List<Identity> hashedPhones = new ArrayList<>();
+
+    public IdentityMapV3Input() {}


Do we want/need a public ctor?

from memory, might be to help consumers of the SDK to write their own tests

Also for cases where clients have a mix of different DII types, which we now support.
You could do IdentityMapV3Input.fromPhones(...).withEmails(...) if you have those cleanly separated. If you have a mix of records you can do something like:

input = new IdentityMapV3Input() for dii in diis: if dii is email: input.addEmail(dii) if dii is phone: input.addPhone(dii)

mcollins-ttd · 2025-06-10T03:55:52Z

src/main/java/com/uid2/client/IdentityMapV3Response.java

+        }
+    }
+
+    private static IdentityMapV3Input getIdentityMapInput(IdentityMapV3Input identityMapInput) {


This just returns its input and isn't used anywhere

Thanks for spotting, refactoring gone wrong at some point

mcollins-ttd · 2025-06-10T03:57:42Z

src/main/java/com/uid2/client/IdentityMapV3Input.java

+        return new IdentityMapV3Input().withHashedPhones(hashedPhones);
+    }
+
+    private transient final Map<String, List<String>> diiMappings = new HashMap<>();


Can you add a comment about what the keys/values are, or maybe change the name? Maybe something like hashedDiiToRawDii to line up with the param names of addToDiiMappings?

mcollins-ttd · 2025-06-10T05:22:27Z

src/main/java/com/uid2/client/IdentityMapV3Response.java

+        }
+    }
+
+    static public class UnmappedIdentity {


Here and elsewhere, can we use public static instead of static public?

We had it in the original version, will change though, @gmsdelmundo also noticed.

mcollins-ttd · 2025-06-10T05:55:16Z

src/main/java/com/uid2/client/IdentityMapV3Helper.java

+        return new IdentityMapV3Response(decryptedResponseString, identityMapInput);
+    }
+
+    Uid2Helper uid2Helper;


This can be private final

mcollins-ttd · 2025-06-10T06:03:38Z

src/main/java/com/uid2/client/IdentityMapV3Input.java

+        diiMappings.computeIfAbsent(hashedDii, k -> new ArrayList<>()).add(rawDii);
+    }
+
+    private String getEncodedDii(String identityType, int i) {


What does "encoded" refer to here?

"Hashed" would be the correct word, will change.

mcollins-ttd · 2025-06-10T06:23:30Z

src/test/java/com/uid2/client/IdentityMapV3IntegrationTests.java

I'd like to have some unit tests, so we have some test coverage in the CI/CD pipeline. It also means I don't have to run an operator just to test changes.

Added some unit tests.

mcollins-ttd · 2025-06-10T06:28:32Z

src/main/java/com/uid2/client/IdentityMapV3Response.java

+    }
+
+    public HashMap<String, MappedIdentity> getMappedIdentities() {
+        return mappedIdentities;


It would be good to return a copy here and in getUnmappedIdentities, or make mappedIdentities/unmappedIdentities immutable after they're populated.

Added copying. The Mapped/Unmapped Identity is already immutable so not a problem there.

aulme added 5 commits May 29, 2025 16:39

Added V3 as copy of V2

9ab7827

Serializing dii in the new format

7eade1a

Adapted V3 Identity Map clases to the new endpoint

ec27779

Ignoring .env

cf6168e

Checking refreshFrom in IdentityMapV3 tests, should be in the future,…

3df6e35

… 1 day buffer to remove flakiness

aulme requested review from mcollins-ttd and vishalegbert-ttd June 4, 2025 00:53

aulme added 2 commits June 4, 2025 14:48

Empty constructor pulbic and with methods for one dii are available

c1fb9d4

Javadocs

c3ef65d

aulme requested a review from jon8787 June 5, 2025 06:17

jon8787 reviewed Jun 6, 2025

View reviewed changes

Addressing feedback

18cfc01

vishalegbert-ttd reviewed Jun 6, 2025

View reviewed changes

src/main/java/com/uid2/client/IdentityMapV3Response.java Outdated Show resolved Hide resolved

aulme and others added 2 commits June 6, 2025 17:35

More feedback

ee0782a

[CI Pipeline] Released Snapshot version: 4.6.1-alpha-17-SNAPSHOT

7b3ae88

mcollins-ttd reviewed Jun 10, 2025

View reviewed changes

aulme force-pushed the aul-UID2-5485-identity-map-v3 branch from af5a5a1 to c70243d Compare June 11, 2025 08:05

Using enum for unmapped operations instead of strings

b8163ce

aulme force-pushed the aul-UID2-5485-identity-map-v3 branch from c70243d to b8163ce Compare June 11, 2025 08:29

aulme added 2 commits June 11, 2025 16:52

Address Matt's feedback

9a974ff

Improving unit test coverage for V3 Identity Map

df2c22a

mcollins-ttd approved these changes Jun 11, 2025

View reviewed changes

Release Workflow and others added 3 commits June 12, 2025 01:28

[CI Pipeline] Released Snapshot version: 4.6.2-alpha-19-SNAPSHOT

162e905

Adapted SDK to changes in V3 API

ec0b6fd

[CI Pipeline] Released Snapshot version: 4.6.3-alpha-20-SNAPSHOT

73ee0ab


		Response response = new Response(identityMapInput);

		response.assertUnmapped("INVALID", "this is not a hashed email");

Identity Map V3 #83

Are you sure you want to change the base?

Identity Map V3 #83

Uh oh!

Conversation

aulme commented Jun 3, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

aulme Jun 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

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!

jon8787 Jun 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

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!

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!

Choose a reason for hiding this comment

Uh oh!

mcollins-ttd Jun 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

aulme Jun 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

aulme Jun 6, 2025 •

edited

Loading

jon8787 Jun 10, 2025 •

edited

Loading

mcollins-ttd Jun 10, 2025 •

edited

Loading

aulme Jun 11, 2025 •

edited

Loading