IssueTracker

Hi,

Thank you for taking the time to improve libphonenumber!

We are looking into this and will respond as soon as possible.

Hi,

Thanks much for raising this issue. This is one of our known issue, which is not documented. Having phone numbers of this long are very rare. libphonenumber cannot do anything better here.

Reasons:

• JavaScript Number type has limitation of 2^53 (9007199254740992) as the max literal it can hold. When provided more than that like "90018408059449376"(17 digit), JS starts rounding the value. Unfortunately this is within the range of possible phone number length i.e 17 digit.
• Both the alternatives leads to backward-incompatible changes as the type of the field '''national_number''' has to be changed; thus leading to client-side compile issue . Eg: setNationalNumber(1234) / conversions to payloads like JSON.

1. Marking the field as [jstype = JS_STRING]
2. Or using new JS data type BigInt.

We will update FAQ at GitHub stating some of these points. Maybe you can compare with Number.MAX_VALUE and avoid such cases. Or post request to a Java/CPP/PY libphonenumber clients, as these langs have much more matured data type systems.

Added to client FAQ as known issue. https://github.com/google/libphonenumber/pull/2674

Comment #25, note that the team member who works on this is OOO for a few days and may not respond for a little while. According to the comments above, yes, AGP 3.6.3 should include the fix.

Do you have repro steps? Just to help ensure that this is, in fact, the same issue reported here? There might be multiple paths to produce the same error, etc. etc. If you're certain it's the same issue as here, then repro steps will help us determine if we should reopen this. Otherwise, it might be best to file a new bug since this one seems to have quite a bit of history on it.

Meanwhile, there are 20 +1's on this bug. Can anyone else who was experiencing this report that it was actually fixed for them starting in 3.6.0?

Thanks! David

Hey David,

I think this may have regressed. I created a fork of the original sample project with AGP 4.1.2.

https://github.com/saadfarooq/Robolectric-Databinding-Lib

The Robolectric test fails with the same error. If kotlin-kapt plugin is removed, the test passes. But then the buildFeatures databinding config says If you plan to use data binding in a Kotlin project, you should apply the kotlin-kapt plugin.

So possibly something to do with kapt?

Ooof :) Hung, are you swamped? Would you be able to take a look?

Comment #27, in the meanwhile, it may be worth trying things out with a more recent version of AGP (e.g. using beta or canary) to see if this remains broken after 4.1.2, or if this was a temporary regression.

Hung: I meant to say, I reopned this bug, but if it ends up being a different cause, or just because this bug already has so much history, feel free to reclose it and create a new one.

I tried 4.2.0-beta05 and 7.0.0-alpha8 in Arctic Fox Canary 08 and the behavior is the same. Remove kapt, test passes. Added it back, get the NoClassDefFound error.

Thank you for the sample project! This issue was mentioned earlier at comment #20, 21, 23.

This is caused by https://youtrack.jetbrains.com/issue/KT-27404, originally reported at Issue 117145132.

Basically, when using kapt, users will need to add:

• kaptTest some-random-dependency for kapt to run annotation processing on unit test source code, which in this case will include the missing data binding classes.
• kaptAndroidTest some-random-dependency for kapt to run annotation processing on androidTest source code.

Please follow the Kapt bug above for progress on this issue.

@David: I'm closing this issue as this should ideally be fixed by the Kotlin Gradle plugin.

I can not sing google account