But it looks that not understanding this API is smaller problem.
I found that *android.onVariantProperties* is never called when I call it
inside of *project.afterEvaluate*
I'm doing it because I have own configuration extension inside of my plugin.
So setup looks like
apply .....
myConfigExtension {
mysetup
}
And in the time od signing configuration I need to know about values in
myConfigExtension
Without it it at least run the first println, but it never call
RemoteSignTask
android.onVariantProperties { ApplicationVariantProperties v ->
println 'xxxxx' + "remoteSign" + v.name.capitalize() +
"Apk"
TaskProvider singingTaskProvider = project.tasks.register(
"remoteSign" + v.name.capitalize() + "Apk",
RemoteSignTask)
//def transformationRequest =
v.artifacts.use(singingTaskProvider)
.wiredWithDirectories(
{ it.apkFolder }, { it.apkFolder }
)
.toTransformMany(ArtifactType.APK.INSTANCE)
}
Dne čtvrtek 25. června 2020 v 2:23:37 UTC+2 uživatel Jerome Dochez napsal:
> we do not have such a thing as an unsigned APK (even internally because we
> package and sign simultaneously).
> what you need to do is disable signing alltogether for your app by turning
> off v1 and v2 signing through the DSL, then your APK would be unsigned...
>
> On Wed, Jun 24, 2020 at 1:35 PM Tomáš Procházka <[email protected]>
> wrote:
>
>> I'm sorry, but this API doesn't make sense at all for me.
>> This WorkerEnabledTransformation example just takes the final APK and
>> copy it somewhere else.
>> But I need to replace the default sign mechanism provided by the Android
>> plugin and nobody else (a different plugin) should know about.
>> So technically I need to tell that my task needs unsigned APK on input
>> and final APK on output.
>> If I understand correctly this talk
>> https://youtu.be/OTANozHzgPc?t=1023
>> With new API I should not care about the actual task which is responsible
>> o signing. I should care just about input and output artifacts.
>> So I should tell that I have a task that can handle unsigned APK on the
>> input and can sign it. And I want to replace any other task, which would
>> like do the same.
>> The replace mechanism described in the talk at
>> https://youtu.be/OTANozHzgPc?t=1023 gives much more sense for me.
>> I would expect something like this:
>>
>> artifact.use(signTaskProducer).toCreate(ArtifactType.APK.INSTANCE).from(ArtifactType.UNSIGNED_APK.INSTANCE)
>>
>>
>> Dne středa 24. června 2020 v 21:54:32 UTC+2 uživatel Jerome Dochez napsal:
>>
>>> toCreate API is used to create an artifact from its input, so here that
>>> would mean creating the APK file from dex, merged manifests, resources,
>>> etc...
>>>
>>> what you want is transform :
>>>
>>> val transformationRequest = artifacts.use(singingTaskProvider)
>>> .wiredWithDirectories(
>>> SigningTask::inputApkFolder,
>>> SingingTas::outputAPKFolder)
>>> .toTransformMany(ArtifactType.APK)
>>>
>>>
>>> look at WorkerEnabledTransformation, it should be very similar except
>>> you copy and sign the files instead of just copying
>>>
>>> you can do AAB but I need to add test APK which right now are not
>>> available through the public artifact types.
>>>
>>> On Wed, Jun 24, 2020 at 8:32 AM Tomáš Procházka <[email protected]>
>>> wrote:
>>>
>>>> I'm still not sure how to use it exactly, can you help a little bit.
>>>> Currently, I have this code:
>>>>
>>>> android.onVariantProperties { ApplicationVariantProperties
>>>> v ->
>>>> TaskProvider signTaskProducer = tasks.register(
>>>> "remoteSign" + variant.name.capitalize() +
>>>> "Apk",
>>>> RemoteSignTask)
>>>>
>>>> v.artifacts.use(signTaskProducer)
>>>> .wiredWith( ? )
>>>> .toCreate(ArtifactType.APK.INSTANCE)
>>>> }
>>>>
>>>> But I'm not sure what is responsible for produce unsigned task.
>>>> And I need to sign APK, AAB and also test APK
>>>> And inside of RemoteSignTask task I need to get location of all files
>>>> which should be fixed.
>>>>
>>>> My original code working until 4.0.0 was looking like this:
>>>>
>>>> project.android.applicationVariants.each {
>>>> ApplicationVariant variant ->
>>>> if (!isReleaseBuildType(variant)) {
>>>> project.logger.info("Skipping signing, '$
>>>> variant.name' is not release.")
>>>> return
>>>> }
>>>>
>>>> variant.outputs.all { ApkVariantOutput output ->
>>>> // hack to keep proper final name also when
>>>> assemble task will be skipped
>>>> output.outputFileName =
>>>> RemoteSignTask.renameToSignedName(output.outputFileName)
>>>> }
>>>> variant.setOutputsAreSigned(true)
>>>>
>>>> // Regular APK signing
>>>> def apk =
>>>> variant.getFinalArtifact(InternalArtifactType.APK.INSTANCE)
>>>> InternalArtifactType.APK.INSTANCE
>>>> def signTask = project.tasks.register(
>>>> "remoteSign" + variant.name.capitalize() +
>>>> "Apk", RemoteSignTask,
>>>> new RemoteSignTask.ConfigAction(variant, apk))
>>>>
>>>> // FIXME: IT should be completely outside this task,
>>>> but it is not possible now
>>>> def folderManifestTaskForApk = project.tasks.register(
>>>> "folderManifest" + variant.name.capitalize() +
>>>> "Apk", FolderManifestTask,
>>>> new FolderManifestTask.ConfigAction(variant,
>>>> apk))
>>>>
>>>> variant.assembleProvider.configure {
>>>> it.dependsOn(signTask)
>>>> it.dependsOn(folderManifestTaskForApk)
>>>> }
>>>> signTask.configure {
>>>> folderManifestTaskForApk.get().dependsOn(it)
>>>> }
>>>>
>>>> // App bundle signing
>>>> def bundle =
>>>> variant.getFinalArtifact(InternalArtifactType.BUNDLE.INSTANCE)
>>>> def bundleSignTask = project.tasks.register(
>>>> "remoteSign" + variant.name.capitalize() +
>>>> "Bundle", RemoteSignTask,
>>>> new RemoteSignTask.ConfigAction(variant,
>>>> bundle))
>>>>
>>>> // FIXME: IT should be completely outside this tak, but
>>>> it is not possible now
>>>> def folderManifestTaskForAAB = project.tasks.register(
>>>> "folderManifest" + variant.name.capitalize() +
>>>> "Bundle", FolderManifestTask,
>>>> new FolderManifestTask.ConfigAction(variant,
>>>> bundle))
>>>>
>>>>
>>>> variant.getVariantData().getTaskContainer().bundleTask.configure {
>>>> it.dependsOn(bundleSignTask)
>>>> it.dependsOn(folderManifestTaskForAAB)
>>>> }
>>>> bundleSignTask.configure {
>>>> folderManifestTaskForAAB.get().dependsOn(it)
>>>> }
>>>>
>>>> if (variant.testVariant != null ) {
>>>> // Test APK signing
>>>> def testApk =
>>>> variant.testVariant.getFinalArtifact(InternalArtifactType.APK.INSTANCE)
>>>> def testSignTask = project.tasks.register(
>>>> "remoteSign" + variant.name.capitalize() +
>>>> "AndroidTest", RemoteSignTask,
>>>> new
>>>> RemoteSignTask.ConfigAction(variant.testVariant, testApk))
>>>>
>>>> variant.testVariant.assembleProvider.configure {
>>>> dependsOn(testSignTask)
>>>> }
>>>>
>>>>
>>>> variant.testVariant.connectedInstrumentTestProvider.configure {
>>>> dependsOn(testSignTask)
>>>> }
>>>> }
>>>>
>>>> }
>>>>
>>>> Dne středa 24. června 2020 v 7:00:11 UTC+2 uživatel Jerome Dochez
>>>> napsal:
>>>>
>>>>> yes, just replace the artifact type with APK.
>>>>>
>>>>> On Tue, Jun 23, 2020 at 3:12 PM Tomáš Procházka <[email protected]>
>>>>> wrote:
>>>>>
>>>>>> But it looks that replace Sign mechanism should be done in the same
>>>>>> way as replace manifest merger which is shown here
>>>>>>
>>>>>> https://github.com/android/gradle-recipes/blob/master/Groovy/manifestReplacementTest/app/build.gradle
>>>>>>
>>>>>> right?
>>>>>>
>>>>>> Dne čtvrtek 11. června 2020 v 20:33:23 UTC+2 uživatel Tomáš Procházka
>>>>>> napsal:
>>>>>>
>>>>>>> This is super useful https://github.com/android/gradle-recipes
>>>>>>> Would be possible to add there how to
>>>>>>>
>>>>>>> - rename output apk/aab to a custom name
>>>>>>> - write custom apk/aab signer
>>>>>>>
>>>>>>> Please.
>>>>>>>
>>>>>>>
>>>>>>> Dne čtvrtek 21. května 2020 8:56:13 UTC+2 Tomáš Procházka napsal(a):
>>>>>>>
>>>>>>>> Hi. Thanks. An example would be really great. So in 3.6 is the only
>>>>>>>> possible way to found last one *task *which produces APK and do my
>>>>>>>> stuff in the doLast {} closure, right?
>>>>>>>>
>>>>>>>>
>>>>>>>> Dne pátek 6. března 2020 19:02:36 UTC+1 Jerome Dochez napsal(a):
>>>>>>>>>
>>>>>>>>> it's not possible to do this in 3.6. We are hoping to deliver most
>>>>>>>>> of this during the 4.1 and 4.2 releases.
>>>>>>>>>
>>>>>>>>> for your second question, you probably will just need to get the
>>>>>>>>> PublicArtifactType.APK artifacts and have a task that depend on them.
>>>>>>>>> this should actually already work in 4.0 with relatively (!)
>>>>>>>>> stable APIs, I can slap an example next week if you are interested
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Fri, Mar 6, 2020 at 8:50 AM Tomáš Procházka <
>>>>>>>>> [email protected]> wrote:
>>>>>>>>>
>>>>>>>>>> Hi.
>>>>>>>>>>
>>>>>>>>>> If. Understand it correctly.
>>>>>>>>>> New variant API is mentioned here
>>>>>>>>>> https://youtu.be/OTANozHzgPc?t=995
>>>>>>>>>> It is currently possible with 3.6.0 ?
>>>>>>>>>> Is there some more advanced article or video about this API or
>>>>>>>>>> documentation somewhere?
>>>>>>>>>>
>>>>>>>>>> For example, I'm not sure if I should use *replace *or
>>>>>>>>>> *transform* to replace the default sign mechanism?
>>>>>>>>>>
>>>>>>>>>> And I also need one new thing. I would like to count hash of all
>>>>>>>>>> APK and AAB produced during the build.
>>>>>>>>>> So I maybe can use *register, *but I don't want to force
>>>>>>>>>> production of APK or AAB. I want just wait when it happens so when
>>>>>>>>>> user
>>>>>>>>>> call bundle or assemble I need to know about every apk or bundle
>>>>>>>>>> that will
>>>>>>>>>> be produced.
>>>>>>>>>> It is possible?
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Dne pondělí 21. října 2019 18:15:13 UTC+2 Jerome Dochez napsal(a):
>>>>>>>>>>
>>>>>>>>>>> yes there will be :
>>>>>>>>>>> 1. disable all signing in the DSL (v1 and v2)
>>>>>>>>>>> 2. obtain the built APK with new variant API
>>>>>>>>>>> 3. sign them.
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> On Sun, Oct 20, 2019 at 12:16 PM Tomáš Procházka <
>>>>>>>>>>> [email protected]> wrote:
>>>>>>>>>>>
>>>>>>>>>>>> Hi Jerome. Thanks.
>>>>>>>>>>>>
>>>>>>>>>>>> Btw. Best would be if there will be a direct way how to replace
>>>>>>>>>>>> default sign mechanism, just by implementing some interface ;-)
>>>>>>>>>>>> Without modifying tasks and dependencies between them.
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> Dne pátek 18. října 2019 19:43:29 UTC+2 Jerome Dochez napsal(a):
>>>>>>>>>>>>>
>>>>>>>>>>>>> Hi Tomas
>>>>>>>>>>>>>
>>>>>>>>>>>>> we are getting closer to provide a new API that will be stable
>>>>>>>>>>>>> so you want have to handle such changes.
>>>>>>>>>>>>>
>>>>>>>>>>>>> On Tue, Oct 15, 2019 at 7:26 AM Tomáš Procházka <
>>>>>>>>>>>>> [email protected]> wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>>> I will reply myself. The correct form of getFinalArtifact
>>>>>>>>>>>>>> parameter
>>>>>>>>>>>>>> is InternalArtifactType.APK.INSTANCE
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> Dne neděle 13. října 2019 23:56:14 UTC+2 Tomáš Procházka
>>>>>>>>>>>>>> napsal(a):
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Hi.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Please, I need help again.
>>>>>>>>>>>>>>> My custom sign mechanism is again broken in plugin 4.6.0.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> InstallableVariantImpl.getFinalArtifact now return a
>>>>>>>>>>>>>>> different value, insead of BuildableArtifact it is there
>>>>>>>>>>>>>>> now Provider<FileCollection>.
>>>>>>>>>>>>>>> It is also necessary to call it in this way from Groovy now
>>>>>>>>>>>>>>> variant.getFinalArtifact(new
>>>>>>>>>>>>>>> InternalArtifactType.APK())
>>>>>>>>>>>>>>> I get Provider instance correctly, but the collection is
>>>>>>>>>>>>>>> always empty.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> @TaskAction
>>>>>>>>>>>>>>> void sign() {
>>>>>>>>>>>>>>> println '>>>>>>>>>>>>>> A2 sign task: ' +
>>>>>>>>>>>>>>> inputFiles.get().files.size()
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> This is always 0.
>>>>>>>>>>>>>>> I'm calling get() in my task, which is registered in this way
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> variant.assembleProvider.configure {
>>>>>>>>>>>>>>> dependsOn(signTask)
>>>>>>>>>>>>>>> }
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Dne čtvrtek 17. ledna 2019 12:40:06 UTC+1 Tomáš Procházka
>>>>>>>>>>>>>>> napsal(a):
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> So, here is my final solution:
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> https://gist.github.com/tprochazka/457c7eebd044c0210dcc8ba49301cda9
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> It's quite complicated. I'm using not public API, but it
>>>>>>>>>>>>>>>> looks that it works currently.
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> I created a new feature request to make it possible in some
>>>>>>>>>>>>>>>> easier way https://issuetracker.google.com/issues/122883577
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> --
>>>>>>>>>>>>>> You received this message because you are subscribed to the
>>>>>>>>>>>>>> Google Groups "adt-dev" group.
>>>>>>>>>>>>>> To unsubscribe from this group and stop receiving emails from
>>>>>>>>>>>>>> it, send an email to [email protected].
>>>>>>>>>>>>>> To view this discussion on the web visit
>>>>>>>>>>>>>> https://groups.google.com/d/msgid/adt-dev/55e89d8f-5216-4ff5-b2e0-765f80025438%40googlegroups.com
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> <https://groups.google.com/d/msgid/adt-dev/55e89d8f-5216-4ff5-b2e0-765f80025438%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>>>>>>>> .
>>>>>>>>>>>>>>
>>>>>>>>>>>>> --
>>>>>>>>>>>> You received this message because you are subscribed to the
>>>>>>>>>>>> Google Groups "adt-dev" group.
>>>>>>>>>>>> To unsubscribe from this group and stop receiving emails from
>>>>>>>>>>>> it, send an email to [email protected].
>>>>>>>>>>>> To view this discussion on the web visit
>>>>>>>>>>>> https://groups.google.com/d/msgid/adt-dev/cae178f3-d127-4803-9062-8bacbeebd250%40googlegroups.com
>>>>>>>>>>>>
>>>>>>>>>>>> <https://groups.google.com/d/msgid/adt-dev/cae178f3-d127-4803-9062-8bacbeebd250%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>>>>>> .
>>>>>>>>>>>>
>>>>>>>>>>> --
>>>>>>>>>> You received this message because you are subscribed to the
>>>>>>>>>> Google Groups "adt-dev" group.
>>>>>>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>>>>>>> send an email to [email protected].
>>>>>>>>>> To view this discussion on the web visit
>>>>>>>>>> https://groups.google.com/d/msgid/adt-dev/76a13511-fa1b-4737-9c20-8f5425867438%40googlegroups.com
>>>>>>>>>>
>>>>>>>>>> <https://groups.google.com/d/msgid/adt-dev/76a13511-fa1b-4737-9c20-8f5425867438%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>>>> .
>>>>>>>>>>
>>>>>>>>> --
>>>>>> You received this message because you are subscribed to the Google
>>>>>> Groups "adt-dev" group.
>>>>>>
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>>> an email to [email protected].
>>>>>> To view this discussion on the web visit
>>>>>> https://groups.google.com/d/msgid/adt-dev/0d56caca-91f4-40d7-8069-055051881948n%40googlegroups.com
>>>>>>
>>>>>> <https://groups.google.com/d/msgid/adt-dev/0d56caca-91f4-40d7-8069-055051881948n%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>> .
>>>>>>
>>>>> --
>>>> You received this message because you are subscribed to the Google
>>>> Groups "adt-dev" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>> an email to [email protected].
>>>>
>>> To view this discussion on the web visit
>>>> https://groups.google.com/d/msgid/adt-dev/2bb989c1-97d1-4d7c-9603-53eaf0ce66dcn%40googlegroups.com
>>>>
>>>> <https://groups.google.com/d/msgid/adt-dev/2bb989c1-97d1-4d7c-9603-53eaf0ce66dcn%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>> .
>>>>
>>> --
>> You received this message because you are subscribed to the Google Groups
>> "adt-dev" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected].
>>
> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/adt-dev/1d22642d-3d74-456e-af8b-8e17eda0d055n%40googlegroups.com
>>
>> <https://groups.google.com/d/msgid/adt-dev/1d22642d-3d74-456e-af8b-8e17eda0d055n%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>>
>
--
You received this message because you are subscribed to the Google Groups
"adt-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/adt-dev/4113877e-5987-4d9c-bc5e-d008fdc14e6fn%40googlegroups.com.
