fixed unmarshaling van LocalDateTime error in the DDIExporterTest.java #8517
Conversation
pdurbin
added this to Review 🦁
in IQSS/dataverse (TO BE RETIRED / DELETED in favor of project 34)
via automation
There was a problem hiding this comment.
I can't reproduce the issue on Java 11. If the goal is to support Java 17, I'm wondering if upgrading Gson is cleaner (less boilerplate). @ErykKul can you please give this a try? I left a few more comments in my review. Thanks!
| @@ -1,6 +1,6 @@ | |||
| package edu.harvard.iq.dataverse.export; | |||
|
|
|||
| import com.google.gson.Gson; | |||
| import com.google.gson.*; | |||
There was a problem hiding this comment.
For a long time we've been discouraging wildcard imports but there was some recent discussion on reconsidering this position. Nothing is written at https://guides.dataverse.org/en/5.10/developers/coding-style.html either way. 😄
| private static final Gson gson = new GsonBuilder().registerTypeAdapter(LocalDate.class, (JsonDeserializer<LocalDateTime>) (JsonElement json, Type type, JsonDeserializationContext jsonDeserializationContext) -> { | ||
| Instant instant = Instant.ofEpochMilli(json.getAsJsonPrimitive().getAsLong()); | ||
| return LocalDateTime.ofInstant(instant, ZoneId.systemDefault()); | ||
| }).create(); |
There was a problem hiding this comment.
I ran this test on my laptop and it works fine. It seems like a lot of boilerplate code.
I noticed that the InaccessibleObjectException error was first reported by @poikilotherm in an issue about Java 17:
@ErykKul are you running something newer than Java 11? In that issue, @poikilotherm suggests at upgrading Gson might fix this and what is perhaps a related issued here:
I can't reproduce the problem on Java 11 and haven't tried on anything newer. We aren't actively working on that Java 17 issue above but this pull request seems like a step in the right direction. I'm wondering, however, can you please try upgrading Gson to see if that's also a suitable fix?
There was a problem hiding this comment.
I have tried upgrading gson to 2.9.0, but it did not solve the problem. Downgrading to jdk11 helped, the test passes and in stead of errors I am getting warnings (with both gson versions; 2.9.0 and 2.2.4):
There was a problem hiding this comment.
@ErykKul I appreciate you checking if upgrading Gson helps. Like I said, the tests still run fine on my laptop with your original change. They also run fine in CI. It's a little more boilerplate but it's confined to one test. If this helps us get to Java 17, great. Thanks for the pull request!
| private static final Gson gson = new GsonBuilder().registerTypeAdapter(LocalDate.class, (JsonDeserializer<LocalDateTime>) (JsonElement json, Type type, JsonDeserializationContext jsonDeserializationContext) -> { | ||
| Instant instant = Instant.ofEpochMilli(json.getAsJsonPrimitive().getAsLong()); | ||
| return LocalDateTime.ofInstant(instant, ZoneId.systemDefault()); | ||
| }).create(); |
There was a problem hiding this comment.
@ErykKul I appreciate you checking if upgrading Gson helps. Like I said, the tests still run fine on my laptop with your original change. They also run fine in CI. It's a little more boilerplate but it's confined to one test. If this helps us get to Java 17, great. Thanks for the pull request!
Which issue(s) this PR closes: