Filter excessive logging of artifact download from m2e in "debug mode"

[rgrunber]

See #2420, #2785 for some background on what the problem is.

When opening a Maven project whose contents have never been downloaded to the local repository (eg. a clean $HOME/.m2/repository), and with the system property jdt.ls.debug=true enabled (most commonly in java.jdt.ls.vmargs via. -D), the client and server logs will be overwhelmed with many messages of the form :

client log

[Info  - 13:39:08] Jan. 4, 2024, 1:38:35 p.m. http-outgoing-1 << "[0x0][0x0][0x8][0x0][0x0]N[0xb0][0xab]V[0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0];[0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x10][0x0][0xed]A@[0x1][0x0][0x0]META-INF/maven/org.codehaus.mojo/build-helper-maven-plugin/PK[0x1][0x2][0x14][0x3][\n]"
[Trace - 13:39:08] Received notification 'window/logMessage'.
Params: {
    "type": 3,
    "message": "Jan. 4, 2024, 1:38:35 p.m. http-outgoing-1 << \"[0x0][0x0][0x8][0x0][0x0]N[0xb0][0xab]V[0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x4][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x0][0x10][0x0][0xed]A[0x99][0x1][0x0][0x0]org/PK[0x1][0x2][0x14][0x3][\\n]\""
}

server log

!ENTRY org.eclipse.jdt.ls.logback.appender 1 0 2024-01-04 13:39:11.561
!MESSAGE http-outgoing-8 << "=~5[0xa9][0x83]OU[0xb5][0x97][0xa8][0xab][0x2]c[0xd2][0x86][0xfc]Um]V[0xd5]8W[0xf5][0xda]mY[0xd6]V[0xae]j[0xeb]a_Q[0x9b]Z[0xbd][0xa2]j[0x87][0x97]J[0xba][0xf1][0xb6][0x97]}[0x96][0x9e][0x0][0x88]FDE[0x13][0xea][0xc4]:[0xa4]E36[0x8b][0x16][0xb4][0x8b]4![0xd8][0x80]N[0xd1][0x8a]cB[0xc1][0xa0][0xd8]$[0x11]([0xf2][0xd1][0x1b][0xf2][0xaa]8[0x84][0x13][0xf2]I[0xcc]Q[0xfd][0x8f][0x91][0x16][0xf4]*V[0xf7][0x18]m![0x1f][0xb0][0x10]^?[0x88][0xb8]7z[0x87]{[0xa3][0x3][0xc1]e[0xcd]/r>[0x9e][0x2]n[0x94][0x10]^?[0xb8][0xb8][0x96][0xbd]rN[0x7][0xef],3<[0xe8]{[0xb1][0x85][0x97][0x9e][0xf0][0x9a]LxM&[0xdc]&[0xb][0xf0][0xe7][0x91][0xcb][0xf4][0xa8][0xfb][0xe8]6fC[0xfc]n[0xde]E[0xfc][0x6]j[0x1e]@[0xa1][0xde][0xd8][0xf8](Pm|[0xf9]v[0x8a]>[0x84][0xc5][0x9]$E?[0xea][0xc5][0x80]/[0xe1]F_a[0xe9][0x15][0xab][0xf7][0xd0]+[0xf4]?%[0x13][0x1e][0xfa][0x1f]PK[0x3][0x4][0x14][0x0][0x0][0x8][0x8][0x0]Hy6PE*[0x92],[0xe9][0x1][0x0][0x0][0x12][0x4][0x0][0x0]@[0x0][0x0][0x0]org/apache/commons/io/input/ObservableInputStream$Observer.class[0x9d]RMo[0xd3]@[0x14][0x9c][0xe7]8q[0x13][0x2][0xfd][0xa2][0xa5][0x94][0xcf][0xd2][0x82]B+a[0x90][0x80]C[qh[0x5][0xc2]RD[0xf][0x85]^*!m[0x9c]u[0xe2][0xc8][0xf1]Fk;[0xe2]oqB[0xe2][0x0]w~[0x14][0xe2][0xed][0xc6]-%D [0xb0]d[0xef][0x9b][0xdd][0x99]y[0xb3]O[0xfe][0xf6][0xfd][0xf3][0x17][0x0]O[0xf1][0xa0][0x81][\n]"

This breaks the import as no progress can be made. Luckily this is mostly a problem for those developing as we're more likely to be using jdt.ls.debug.

Looking at the logger's that are generating these, I found org.apache.http.wire generating a lot of messages. After disabling it I was able to get the project to import and noticed this distribution for logging messages :

      1 aQute.bnd.metatype.MetatypeAnnotations
      1 org.eclipse.buildship.core.internal.util.gradle.PublishedGradleVersions
      1 org.eclipse.m2e.core.internal.project.registry.ProjectProcessingTracker
      1 org.eclipse.m2e.core.internal.project.registry.ProjectRegistryRefreshJob
      2 org.eclipse.m2e.core.internal.markers.MavenMarkerManager
      2 org.eclipse.m2e.core.internal.project.registry.ProjectRegistryManager
      3 aQute.bnd.osgi.PluginsContainer
      3 org.eclipse.jgit.util.SystemReader
      4 org.eclipse.aether.internal.impl.synccontext.named.NamedLockFactoryAdapterFactoryImpl
      4 org.eclipse.m2e.core.internal.embedder.EclipseLogger
      4 org.eclipse.m2e.core.internal.project.registry.DefaultMavenDependencyResolver
      5 org.eclipse.m2e.jdt.internal.AbstractJavaProjectConfigurator
      6 org.eclipse.jgit.util.FS
      6 org.eclipse.m2e.core.internal.builder.MavenBuilder
      9 org.eclipse.m2e.core.internal.project.ProjectConfigurationManager
     10 org.eclipse.m2e.core.internal.lifecyclemapping.LifecycleMappingFactory
     18 aQute.bnd.osgi.Processor
     19 aQute.bnd.osgi.Analyzer
     19 aQute.bnd.osgi.Contracts
     19 org.eclipse.jgit.internal.storage.file.Pack
     20 org.eclipse.m2e.core.internal.builder.MavenBuilderImpl
     23 org.eclipse.aether.internal.impl.DefaultLocalRepositoryProvider
     27 org.eclipse.aether.internal.impl.collect.bf.BfDependencyCollector
     40 org.apache.http.impl.conn.DefaultHttpClientConnectionOperator
     54 org.eclipse.m2e.core.project.configurator.AbstractCustomizableLifecycleMapping
    114 org.eclipse.jgit.internal.storage.file.FileSnapshot
    200 org.apache.http.conn.ssl.SSLConnectionSocketFactory
    848 org.eclipse.aether.internal.impl.DefaultRepositoryConnectorProvider
    848 org.eclipse.aether.internal.impl.DefaultTransporterProvider
    924 org.eclipse.aether.internal.impl.DefaultArtifactResolver
    960 org.apache.maven.shared.filtering.DefaultMavenResourcesFiltering
   1222 org.apache.http.client.protocol.RequestAddCookies
   1682 org.eclipse.aether.internal.impl.DefaultTrackingFileManager
   1682 org.eclipse.m2e.core.internal.embedder.AbstractTransferListenerAdapter
   3656 org.apache.http.impl.conn.DefaultManagedHttpClientConnection
   4890 org.apache.http.impl.conn.PoolingHttpClientConnectionManager
   4908 org.apache.http.impl.execchain.MainClientExec
  23536 org.apache.http.headers

I think it's safe to disable org.apache.http entirely (and I've also modified an existing one).