SloppyPhraseScorer sometimes misses documents that ExactPhraseScorer finds. [LUCENE-3821]

The general bug is a case where a phrase with no slop is found,
but if you add slop its not.

I committed a test today (TestSloppyPhraseQuery2) that actually triggers this case,
jenkins just hasn't had enough time to chew on it.

ant test -Dtestcase=TestSloppyPhraseQuery2 -Dtests.iter=100 is enough to make it fail on trunk or 3.x

---
Migrated from [LUCENE-3821](https://issues.apache.org/jira/browse/LUCENE-3821) by Naomi Dushay, updated Mar 13 2012
Attachments: [LUCENE-3821_test.patch](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-3821/LUCENE-3821_test.patch), [LUCENE-3821.patch](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-3821/LUCENE-3821.patch) (versions: 4), [LUCENE-3821-SloppyDecays.patch](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-3821/LUCENE-3821-SloppyDecays.patch), [schema.xml](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-3821/schema.xml), [solrconfig-test.xml](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-3821/solrconfig-test.xml)

Start with TestSloppyPhraseQuery2 and run `ant test -Dtestcase=TestSloppyPhraseQuery2 -Dtests.iter=100` on trunk or 3.x to reproduce the failure. Read SloppyPhraseScorer alongside ExactPhraseScorer and use the attached patches for context. Done means the regression test passes without sloppy phrase queries missing documents found by exact phrase queries.

Written by the indexing model from the issue text.