During merging, write empty (headers-only) numeric vectors when vector-valued FieldInfos is present [LUCENE-9992]

When we merge segments having all documents with vectors in some field deleted, we write FieldInfo for that field, but no index files, and then when we read the index we get exceptions as shown by this ci test failure:

Build: 

1 tests failed.
FAILED:  org.apache.lucene.codecs.lucene90.TestLucene90HnswVectorFormat.testDeleteAllVectorDocs

Error Message:
org.apache.lucene.index.CorruptIndexException: Problem reading index from RawDirectoryWrapper(ByteBuffersDirectory@70f29f52 lockFactory=org.apache.lucene.store.SingleInstanceLockFactory@486fbd6) (resource=RawDirectoryWrapper(ByteBuffersDirectory@70f29f52 lockFactory=org.apache.lucene.store.SingleInstanceLockFactory@486fbd6))

Stack Trace:
org.apache.lucene.index.CorruptIndexException: Problem reading index from RawDirectoryWrapper(ByteBuffersDirectory@70f29f52 lockFactory=org.apache.lucene.store.SingleInstanceLockFactory@486fbd6) (resource=RawDirectoryWrapper(ByteBuffersDirectory@70f29f52 lockFactory=org.apache.lucene.store.SingleInstanceLockFactory@486fbd6))
        at __randomizedtesting.SeedInfo.seed([DA9C80A622E56DF7:82EB7A7E3E790C0C]:0)
        at org.apache.lucene.index.SegmentCoreReaders.<init>(SegmentCoreReaders.java:160)
        at org.apache.lucene.index.SegmentReader.<init>(SegmentReader.java:89)
        at org.apache.lucene.index.ReadersAndUpdates.getReader(ReadersAndUpdates.java:179)
        at org.apache.lucene.index.ReadersAndUpdates.getReadOnlyClone(ReadersAndUpdates.java:221)
        at org.apache.lucene.index.IndexWriter.lambda$getReader$0(IndexWriter.java:534)
        at org.apache.lucene.index.StandardDirectoryReader.open(StandardDirectoryReader.java:137)
        at org.apache.lucene.index.IndexWriter.getReader(IndexWriter.java:596)
        at org.apache.lucene.index.IndexWriter.getReader(IndexWriter.java:452)
        at org.apache.lucene.index.BaseVectorFormatTestCase.testDeleteAllVectorDocs(BaseVectorFormatTestCase.java:564)
        at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
        at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.base/java.lang.reflect.Method.invoke(Method.java:566)
        at com.carrotsearch.randomizedtesting.RandomizedRunner.invoke(RandomizedRunner.java:1754)
        at com.carrotsearch.randomizedtesting.RandomizedRunner$8.evaluate(RandomizedRunner.java:942)
        at com.carrotsearch.randomizedtesting.RandomizedRunner$9.evaluate(RandomizedRunner.java:978)
        at com.carrotsearch.randomizedtesting.RandomizedRunner$10.evaluate(RandomizedRunner.java:992)
        at org.apache.lucene.util.TestRuleSetupTeardownChained$1.evaluate(TestRuleSetupTeardownChained.java:44)
        at org.apache.lucene.util.AbstractBeforeAfterRule$1.evaluate(AbstractBeforeAfterRule.java:43)
        at org.apache.lucene.util.TestRuleThreadAndTestName$1.evaluate(TestRuleThreadAndTestName.java:45)
        at org.apache.lucene.util.TestRuleIgnoreAfterMaxFailures$1.evaluate(TestRuleIgnoreAfterMaxFailures.java:60)
        at org.apache.lucene.util.TestRuleMarkFailure$1.evaluate(TestRuleMarkFailure.java:44)
        at org.junit.rules.RunRules.evaluate(RunRules.java:20)
        at com.carrotsearch.randomizedtesting.rules.StatementAdapter.evaluate(StatementAdapter.java:36)
        at com.carrotsearch.randomizedtesting.ThreadLeakControl$StatementRunner.run(ThreadLeakControl.java:370)
        at com.carrotsearch.randomizedtesting.ThreadLeakControl.forkTimeoutingTask(ThreadLeakControl.java:819)
        at com.carrotsearch.randomizedtesting.ThreadLeakControl$3.evaluate(ThreadLeakControl.java:470)
        at com.carrotsearch.randomizedtesting.RandomizedRunner.runSingleTest(RandomizedRunner.java:951)
        at com.carrotsearch.randomizedtesting.RandomizedRunner$5.evaluate(RandomizedRunner.java:836)
        at com.carrotsearch.randomizedtesting.RandomizedRunner$6.evaluate(RandomizedRunner.java:887)
        at com.carrotsearch.randomizedtesting.RandomizedRunner$7.evaluate(RandomizedRunner.java:898)
        at org.apache.lucene.util.AbstractBeforeAfterRule$1.evaluate(AbstractBeforeAfterRule.java:43)
        at com.carrotsearch.randomizedtesting.rules.StatementAdapter.evaluate(StatementAdapter.java:36)
        at org.apache.lucene.util.TestRuleStoreClassName$1.evaluate(TestRuleStoreClassName.java:38)
        at com.carrotsearch.randomizedtesting.rules.NoShadowingOrOverridesOnMethodsRule$1.evaluate(NoShadowingOrOverridesOnMethodsRule.java:40)
        at com.carrotsearch.randomizedtesting.rules.NoShadowingOrOverridesOnMethodsRule$1.evaluate(NoShadowingOrOverridesOnMethodsRule.java:40)
        at com.carrotsearch.randomizedtesting.rules.StatementAdapter.evaluate(StatementAdapter.java:36)
        at com.carrotsearch.randomizedtesting.rules.StatementAdapter.evaluate(StatementAdapter.java:36)
        at org.apache.lucene.util.TestRuleAssertionsRequired$1.evaluate(TestRuleAssertionsRequired.java:53)
        at org.apache.lucene.util.AbstractBeforeAfterRule$1.evaluate(AbstractBeforeAfterRule.java:43)
        at org.apache.lucene.util.TestRuleMarkFailure$1.evaluate(TestRuleMarkFailure.java:44)
        at org.apache.lucene.util.TestRuleIgnoreAfterMaxFailures$1.evaluate(TestRuleIgnoreAfterMaxFailures.java:60)
        at org.apache.lucene.util.TestRuleIgnoreTestSuites$1.evaluate(TestRuleIgnoreTestSuites.java:47)
        at org.junit.rules.RunRules.evaluate(RunRules.java:20)
        at com.carrotsearch.randomizedtesting.rules.StatementAdapter.evaluate(StatementAdapter.java:36)
        at com.carrotsearch.randomizedtesting.ThreadLeakControl$StatementRunner.run(ThreadLeakControl.java:370)
        at com.carrotsearch.randomizedtesting.ThreadLeakControl.lambda$forkTimeoutingTask$0(ThreadLeakControl.java:826)
        at java.base/java.lang.Thread.run(Thread.java:834)
        Suppressed: java.lang.RuntimeException: CheckIndex failed
                at org.apache.lucene.util.TestUtil.checkIndex(TestUtil.java:331)
                at org.apache.lucene.util.TestUtil.checkIndex(TestUtil.java:306)
                at org.apache.lucene.store.BaseDirectoryWrapper.close(BaseDirectoryWrapper.java:42)
                at org.apache.lucene.index.BaseVectorFormatTestCase.testDeleteAllVectorDocs(BaseVectorFormatTestCase.java:549)
                ... 39 more
Caused by: java.io.FileNotFoundException: No sub-file with id _Lucene90HnswVectorFormat_0.vem found in compound file "_2.cfs" (fileName=_2_Lucene90HnswVectorFormat_0.vem files: [.fnm, _Lucene90_0.tip, _Lucene90_0.tmd, _Lucene90_0.doc, _Lucene90_0.tim, .fdm, .fdx, .fdt])
        at org.apache.lucene.codecs.lucene90.Lucene90CompoundReader.openInput(Lucene90CompoundReader.java:168)
        at org.apache.lucene.store.Directory.openChecksumInput(Directory.java:152)
        at org.apache.lucene.codecs.lucene90.Lucene90HnswVectorReader.readMetadata(Lucene90HnswVectorReader.java:98)
        at org.apache.lucene.codecs.lucene90.Lucene90HnswVectorReader.<init>(Lucene90HnswVectorReader.java:67)
        at org.apache.lucene.codecs.lucene90.Lucene90HnswVectorFormat.fieldsReader(Lucene90HnswVectorFormat.java:91)
        at org.apache.lucene.codecs.perfield.PerFieldVectorFormat$FieldsReader.<init>(PerFieldVectorFormat.java:209)
        at org.apache.lucene.codecs.perfield.PerFieldVectorFormat.fieldsReader(PerFieldVectorFormat.java:77)
        at org.apache.lucene.index.SegmentCoreReaders.<init>(SegmentCoreReaders.java:153)
        ... 47 more

Build Log:
[...truncated 664 lines...]
ERROR: The following test(s) have failed:
  - org.apache.lucene.codecs.lucene90.TestLucene90HnswVectorFormat.testDeleteAllVectorDocs (:lucene:core)
    Test output: /home/jenkins/jenkins-slave/workspace/Lucene/Lucene-NightlyTests-main/checkout/lucene/core/build/test-results/test/outputs/OUTPUT-org.apache.lucene.codecs.lucene90.TestLucene90HnswVectorFormat.txt
    Reproduce with: gradlew :lucene:core:test --tests "org.apache.lucene.codecs.lucene90.TestLucene90HnswVectorFormat.testDeleteAllVectorDocs" -Ptests.jvms=4 -Ptests.haltonfailure=false -Ptests.jvmargs=-XX:TieredStopAtLevel=1 -Ptests.seed=DA9C80A622E56DF7 -Ptests.multiplier=2 -Ptests.nightly=true -Ptests.badapples=false -Ptests.file.encoding=UTF-8 -Ptests.linedocsfile=/home/jenkins/jenkins-slave/workspace/Lucene/Lucene-NightlyTests-main/test-data/enwiki.random.lines.txt

---
Migrated from [LUCENE-9992](https://issues.apache.org/jira/browse/LUCENE-9992) by Michael Sokolov (@msokolov), updated Jun 09 2021
Pull requests: https://github.com/apache/lucene/pull/172

Start with BaseVectorFormatTestCase.testDeleteAllVectorDocs and reproduce it using the Gradle command and seed in the report. Trace the failure through Lucene90HnswVectorFormat, Lucene90HnswVectorReader, and PerFieldVectorFormat; done means the headers-only vector case can be read without the reported missing-sub-file exception.

Written by the indexing model from the issue text.