Allow import of all projects in Eclipse on Windows

Gradle has a "shortcut" which omits the target and source compatibility
we set when it thinks it is not necessary (eg gradle is running on the
same version as the target compat). However, the way we compile against
java 9 is to set javac to use java 9, while gradle still runs on java 8.
This change makes -source and -target explicit for now, until these
"optimizations" can be removed from gradle.

closes elastic#18039

Build: Explicitly set target and source compatibility for javac

fix exists method for list settings when using numbered setting format

This change fixes the generation of plugin properties files when the
version changes. Before, it would not regenerate, and running integTest
would fail with an incompatibile version error.

Build: Fix plugin properties generation when version changes

…in set processor

add ability to disable ability to override values of existing fields in set processor

This commit addresses a checkstyle line-length violation in
SetProcessor.java.

…y dsl.

Fix a limitation that prevent from hierarchical inner hits be defined in query dsl.

Removed the nested_path, parent_child_type and query options from inner hits dsl. These options are only set by ES
upon parsing the has_child, has_parent and nested queries are using their respective query builders.

These options are still used internally, when these options are set a new private copy is created based on the
provided InnerHitBuilder and configuring either nested_path or parent_child_type and the inner query of the query builder
being used.

Closes elastic#11118

Today we use a sliced lock strategy for acquiring locks to prevent
concurrent updates to the same document. The number of sliced locks is
computed as a linear function of the number of logical
processors. Unfortunately, the probability of a collision against a
sliced lock is prone to the birthday problem and grows faster than
expected. In fact, the mathematics works out such that for a fixed
target probability of collision, the number of lock slices should grow
like the square of the number of logical processors. This is
less-than-ideal, and we can do better anyway. This commit introduces a
strategy for avoiding lock contention within the internal
engine. Ideally, we would only have lock contention if there were
concurrent updates to the same document. We can get close to this ideal
world by associating a lock with the ID of each document. This
association can be held in a concurrent hash map. Now, the JDK
ConcurrentHashMap also uses a sliced lock internally, but it has several
strategies for avoiding taking the locks and these locks are only held
for a very short period of time. This implementation associates a
reference count with the lock that is associated with a document ID and
automatically removes the document ID from the concurrent hash map when
the reference count reaches zero.

Relates elastic#18060

Add new IPv6 types to docs where it's supported

…oMatchDocsQuery is returned instead now.

…18004)

This commit modifes the EsThreadPoolTestCase#info helper method to
return null when info for the thread pool can not be found. This really
should only happen for the "same" thread pool, and so we also assert
that we only get to a place where there is no info if the thread pool
that info was requested for is in fact the "same" thread pool. Not
returning null here and instead throwing an exception would fail tests
that tried to lookup info on the "same" thread pool.

Fixes elastic#17683.

Closes elastic#17814

name and uuid because the cluster state may contain an active index of
the same name (but different uuid).

Closes elastic#18058
Closes elastic#18054

The source code example for the initial example was missing the correct JSON object formatting and syntax.  That has been fixed with my change.

… terms

When the termslookup (mocked in this case) doesn't return any terms, the
query used to rewrite to an empty boolean query. Now it rewrites to a
MatchNoDocsQuery. This changes the test expectation accordingly.

Closes elastic#18071

This commit removes a racy but unnecessary assertion in scaling thread
pool idle test. Namely, the main test thread can reach the removed
assertion before the last few threads in the thread pool have completed
their tasks and caused the completed tasks count on the underlying
executor to be updated. But this assertion is unnecessary. The main test
thread already waits on a latch that is only decremented immediately
before a task completes. This ensures that it was in fact the case that
every submitted task was executed.

Closes elastic#18072

…r_sort

Updated misspelling in documentation

`ignore_above` doesn't apply to analyzed `text` fields

Closes elastic#18077

Today when running in production mode the bootstrap checks are
completely unforgiving. But there are cases where an end-user might not
have the ability to modify some of the system-level settings that cause
the bootstrap checks to trip (e.g., guest settings that are inherited
from a host and can not be modified). This commit adds a setting that
allows system-level bootstrap checks to be ignored for these
end-users. We classify certain bootstrap checks into system-level checks
and only those bootstrap checks will be ignored if this flag is
enabled. All other bootstrap checks are still subject to being enforced
if the user is in production mode. We will still log warnings for these
bootstrap checks because the end-user does still need to be made aware
that they are running in a configuration that is less-than-ideal from a
resiliency perspective.

Relates elastic#18088

…ed. elastic#18082

Closes elastic#18062

This commit marks the method MaxMapCountCheck#isSystemCheck as final as
this method should not be modifiable by inheriting classes (used in
tests).

This commit removes an unnecessary if statement in Bootstrap#check. The
removed if statement was duplicating the conditionals in the nested if
statements and was merely an artifact of an earlier refactoring.

Don't try to compute completion stats on a reader after we already closed it

Conflicts:
	core/src/main/java/org/elasticsearch/index/shard/IndexShard.java

Closes elastic#18096

Add tests/doc for boolean fields with expressions

With this commit we compress HTTP responses provided the client
supports it (as indicated by the HTTP header 'Accept-Encoding').

We're also able to process compressed HTTP requests if needed.

The default compression level is lowered from 6 to 3 as benchmarks
have indicated that this reduces query latency with a negligible
increase in network traffic.

Closes elastic#7309

Fixes an issue where updating index metadata on a index from a version that it does not have in its static list of known versions fails.

This commit adds a note to the contributing docs on how to configure
assertions inside Eclipse and IntelliJ.

Relates elastic#18107

Today, the constructor for IngestDocument#FieldPath does a string
concatentation and two object allocation on every field path. This
commit removes these unnecessary operations.

Relates elastic#18108

Relates to elastic#17689

This reverts commit 3ca02d6.

The query shard reset() method resets some internal state in the
query shard context, like clearing query names, the filter flag
or named queries. The problem with this method being public is
that it currently (miss?) used for modifying an existing context
for recursive invocatiob, but the contexts that have been reseted
that way cannot be properly set back to their previous state.

This PR is a step towards removing reset() entirely by first making
it only be used internally in QueryShardContext. In places where
reset() was used we can either create new QueryShardContexts or
modify the existing context because it is discarded afterwards anyway.

as of elastic@42526ac
5.0.0 alpha2 have no settingsBuilder() method.

Update java client.asciidoc - Settings.settingsBuilder() removal.

…r_sort

Folds the helper class for random object generation into the
abstract sort test class. Removes a few references to ESTestCase
that were not needed due to inheriting from it along the way.

…to_geovalidationmethod

…alue_other_than_for_sort

Get (mostly) rid of random test data generator for sort tests

…to_geovalidationmethod

…stancesortbuilder_to_geovalidationmethod

Introduces GeoValidationMethod to GeoDistanceSortBuilder

This commit removes the method Strings#splitStringToArray and replaces
the call sites with invocations to String#split. There are only two
explanations for the existence of this method. The first is that
String#split is slightly tricky in that it accepts a regular expression
rather than a character to split on. This means that if s is a string,
s.split(".")  does not split on the character '.', but rather splits on
the regular expression '.' which splits on every character (of course,
this is easily fixed by invoking s.split("\\.") instead). The second
possible explanation is that (again) String#split accepts a regular
expression. This means that there could be a performance concern
compared to just splitting on a single character. However, it turns out
that String#split has a fast path for the case of splitting on a single
character and microbenchmarks show that String#split has 1.5x--2x the
throughput of Strings#splitStringToArray. There is a slight behavior
difference between Strings#splitStringToArray and String#split: namely,
the former would return an empty array in cases when the input string
was null or empty but String#split will just NPE at the call site on
null and return a one-element array containing the empty string when the
input string is empty. There was only one place relying on this behavior
and the call site has been modified accordingly.

PathTrie has a constructor that allows for an arbitrary separtor and
wildcard, but this constructor is unused and internally we always use
'/' as the separator and '*' as the wildcard. There are no tests for the
case where the separator differs from the default separator and
wildcard. This commit removes this constructor and now all instances of
PathTrie have the default separator and wildcard.

…s (such as `top_hits`) which require access to score information.

The score is recomputed lazily for each document belonging to a top bucket.
Relates to elastic#9825

Remove Strings#splitStringToArray

Remove arbitrary separator/wildcard from PathTrie

…_score

Add the ability to use the breadth_first mode with nested aggregations (such as `top_hits`) which require access to score information.

…pt (2)

Currently we have a lot of methods left in QueryShardContext that
take parsers or BytesReference arguments to do some xContent
parsing on the shard. While this still seems necessary in some cases
(e.g. percolation, phrase suggester), the shard context should only
be concerned with generating lucene queries from QueryBuilders.

This change removes all of the parseX() methods in favour of two
public methods toQuery(QueryBuilder) and toFilter(QueryBuilder) that
either call the query builders toFilter() or toQuery() method and
move all code required for parsing out to the respective callers.

* camel_case all the things
* `?pretty` all the things (we tell people to do this in bug reports)

Closes elastic#18136

Adding random shuffling of xContent to InnterHitBuilderTests shows
that the scriptFields are stored in order as a list internally although
they are an unordered json objects in the query dsl.

This changes the internal representation to a set and updates
serialization accordingly.

* Reorganize scripting documentation

* Further changes to tidy up scripting docs

Closes elastic#18116

* Add note about .lat/lon potentially returning null

* Added .value to expressions example

* Fixed two bad ASCIIDOC links

This commit adds support for ES_JAVA_OPTS to the elasticsearch-plugin
script.

Relates elastic#18140

Expressions don't have nulls, only doubles. If the field is missing, then its
treated as 0.0. You can query .empty to see if its missing and substitute something else.

See elastic#18132 (comment)

Today we specify the client option for the JVM when executing plugin
commands. Yet, this option does nothing on a 64-bit capable JDK as such
JDKs always select the Hotspot server VM. And for 32-bit JDKs, running
plugin commands with the server VM is okay. Thus, we should just remove
this unnecessary flag and just let the default VM be selected.

Relates elastic#18142

Added documentation for the cluster.indices.tombstones.size property for maximum tombstones in the cluster state.

All other values are errors.

Add java test for throttling. We had a REST test but it only ran against
one node so it didn't catch serialization errors.

Add Simple round trip test for rethrottle request

This commit introduces a handshake when initiating a light
connection. During this handshake, node information, cluster name, and
version are received from the target node of the connection. This
information can be used to immediately validate that the target node is
a member of the same cluster, and used to set the version on the
stream. This will allow us to extend APIs that are used during initial
cluster recovery without a major version change.

Relates elastic#15971

This change makes the vagrant tasks extend LoggedExec, so that the
entire vagrant output can be dumped on failure (and completely logged
when using --info). It should help for debugging issues like elastic#18122.

Tests: improve logging for vagrant to emit entire output on failure

painless: optimize/simplify dynamic field and method access

Today we softly warn about running with the client VM. However, we
should really refuse to start in production mode if running with the
client VM as the performance of the client VM is too devastating for a
server application. This commit adds an option to jvm.options to ensure
that we are starting with the server VM (on all 32-bit non-Windows
platforms on server-class machines (2+ CPUs, 2+ GB physical RAM) this is
the default and on all 64-bit platforms this is the only option) and
adds a bootstrap check for the client VM.

Relates elastic#18155

Closes elastic#18150

Adds infrastructure so `gradle :docs:check` will extract tests from
snippets in the documentation and execute the tests. This is included
in `gradle check` so it should happen on CI and during a normal build.

By default each `// AUTOSENSE` snippet creates a unique REST test. These
tests are executed in a random order and the cluster is wiped between
each one. If multiple snippets chain together into a test you can annotate
all snippets after the first with `// TEST[continued]` to have the
generated tests for both snippets joined.

Snippets marked as `// TESTRESPONSE` are checked against the response
of the last action.

See docs/README.asciidoc for lots more.

Closes elastic#12583. That issue is about catching bugs in the docs during build.
This catches *some* bugs in the docs during build which is a good start.

The snippet in the docs creates and index and uses it with the
_analyze api. The trouble is that if the index hasn't been created
fully the _analyze API will fail. This adds a
GET _cluster/health?wait_for_status=yellow
which fixes the issue.

While this does make the docs more cluttered, it also makes the snippets
actually runnable.

Closes elastic#18165

o/e/snapshots/Snapshot and o/e/snapshots/SnapshotInfo contain the same
fields and represent the same information.  Snapshot was used to
maintain snapshot information to the snapshot repository, while
SnapshotInfo was used to represent the snapshot information as presented
through the REST layer.  This removes the Snapshot class and combines
all uses into the SnapshotInfo class.

Closes elastic#18167

closes elastic#18162

This gives better coverage and consistency with the scripting APIs, by
whitelisting the primary search scripting API classes and using them instead
of only Map and List methods.

For example, accessing fields can now be done with `.value` instead of `.0`
because `getValue()` is whitelisted. For now, access to a document's fields in
this way (loads) are fast-pathed in the code, to avoid dynamic overhead.

Access to geo fields and geo distance functions is now supported.

TODO: date support (e.g. whitelist ReadableDateTime methods as a start)
TODO: improve docs (like expressions and groovy have for document's fields)
TODO: remove fast-path hack

Closes elastic#18169

Squashed commit of the following:

commit ec9f24b
Author: Robert Muir <rmuir@apache.org>
Date:   Thu May 5 17:59:37 2016 -0400

    cutover to <Def> instead of <Object> here

commit 9edb155
Author: Robert Muir <rmuir@apache.org>
Date:   Thu May 5 17:03:02 2016 -0400

    add fast-path for docvalues field loads

commit f8e38c0
Author: Robert Muir <rmuir@apache.org>
Date:   Thu May 5 16:47:31 2016 -0400

    Painless: add fielddata accessors (.value/.values/.distance()/etc)

All in service of the snippets passing consistently.

This was broken recently as part of making the vagrant tasks extend
LoggedExec. This change fixes the progress logger to not be started
until we start seeing output from vagrant.

Packaging: Make rpm not include parent dirs

This commit fixes a typo in the error message when the now unsupported
environment variable ES_MAX_MEM is set on Windows.

Small typo fix in startup script.

Relates elastic#18168

This commit marks InnerHitsBuilderTests#testFromAndToXContent as
awaiting a fix.

In preparation for a unified release process, we need to be able to
generate the pom files independently of trying to actually publish. This
change adds back the maven-publish plugin just for that purpose. The
nexus plugin still exists for now, so that we do not break snapshots,
but that can be removed at a later time once snapshots are happenign
through the unified tools. Note I also changed the dir jars are written
into so that all our artifacts are under build/distributions.

QueryBuilder has generics, but those are never used: all call sites use
`QueryBuilder<?>`. Only `AbstractQueryBuilder` needs generics so that the base
class can contain a default implementation for setters that returns `this`.

The `ip` field uses a binary representation internally. This breaks when
rendering sort values in search responses since elasticsearch tries to write a
binary byte[] as an utf8 json string. This commit extends the `DocValueFormat`
API in order to give fields a chance to choose how to render values.

Closes elastic#6077

This was mostly untested and had some bugs.

Closes elastic#18085

Closes elastic#18164

This adds some random shuffling of xContent to some more test cases.

Relates to elastic#5831

This was lost whene releasing alpha2 since the version constant got renamed.

This would be better handled by escape analysis.

… in fieldname

We should prevent highlighting if a field is anything but a text or keyword field.
However, someone might implement a custom field type that has text and still want to
highlight on that. We cannot know in advance if the highlighter will be able to
highlight such a field and so we do the following:
If the field is only highlighted because the field matches a wildcard we assume
it was a mistake and do not process it.
If the field was explicitly given we assume that whoever issued the query knew
what they were doing and try to highlight anyway.

closes elastic#17537

Due to trying to modify a map while iterating it, a concurrent modification
in the pipeline stats could be thrown. This uses an iterator to prevent this.

Closes elastic#18126

This causes round tripping through xcontent to fail.

Closes elastic#18166

The files removed are below the 140 character line length limit.

This commit removes handshaking from the transport client. This
handshaking is not needed because of the existence of the liveness
check.

Relates elastic#18174

Relates to elastic#18160

This commit adds support for the semicolon character as a valid query
string parameter delimiter.

Relates elastic#18186

…wildcard-highlighting

Exclude all but string fields from highlighting if wildcards are used…

closes elastic#8041

This commit modifies the packaging for the RPM package so that edits to
config files will not get lost during removal and upgrade.

Relates elastic#18188

Most of the current implementations of BaseNodesResponse (plural Nodes) ignore FailedNodeExceptions.

- This adds a helper function to do the grouping to TransportNodesAction
- Requires a non-null array of FailedNodeExceptions within the BaseNodesResponse constructor
- Reads/writes the array to output
- Also adds StreamInput and StreamOutput methods for generically reading and writing arrays

Build: Add pom generation to assemble task

This commit fixes the exception messages for lifecycles when stopping in
illegal states.

Relates elastic#18189

…lastic#18190)

fixes elastic#18139

We recently added correct serialization for null values, but the helper
method used does not allow null. This fixes serialization to handle the
null.

Mappings: Fix ip mapper to correctly serialize a null null_value

This is a follow up to elastic#18173 and includes adding pom generation to the
fake build-tools project, which is really just buildSrc, but builds
during normal builds.

This commit adds a link to the field data docs from the field data
fields docs for completeness.

Closes elastic#17672

Floating point numbers are added as `float`, and Strings are added as `text` with `keyword sub-field

With this commit we eagerly evaluate content length in HttpServer
and also pass the same value to ResourceHandlingHttpChannel. With
this change it easier to reason about the content length that is
freed leaving no doubt that it must be identical to the reserved
amount.

With this commit we free all bytes reserved on the request
circuit breaker.

Closes elastic#18144

This makes the test generation support both while we move from
`// AUTOSENSE` to `// CONSOLE`.

Will bother elastic#18160

Integration tests failed:
https://elasticsearch-ci.elastic.co/job/elastic+elasticsearch+master+multijob-intake/483/console

We'll see if the rest tests were hiding some other failure.

This commit modifies the packaging tests to account for the fact that
rpm behaves differently with respect to preserving directories marked as
"CONFIG | NOREPLACE" on older versions versus newer versions. Older
versions will leave the directory as-is while newer versions will append
the suffix ".rpmsave" to the directory name.

Relates elastic#18216

Log the method and the path.

The plugin script parses command-line options looking for Java system
properties and extracts these arguments to pass to the java command when
starting the JVM. Since elasticsearch-plugin allows arbitrary user
arguments to the JVM via ES_JAVA_OPTS, this parsing is unnecessary. This
commit removes this unnecessary 

Relates elastic#18207

Tests: Delay starting progress loggers for vagrant until test is running

build-tools project

…ping

but not in the index.

Build: Add fake project to include buildSrc as normal project

Do not return fieldstats information for fields that exist in the mapping but not in the index.

This makes it much easier to apply to other projects.

Fixes to doc tests infrastructure:
* Fix comparing lists. Was totally broken.
* Fix order of actual vs expected parameters.
* Allow multiple `// TESTRESPONSE` lines with substitutions to join
into one big list of subtitutions. This makes lets the docs look
tidier.
* Exclude build from snippet scanning
* Allow subclasses of ESRestTestCase access to the admin execution context

…se (elastic#18102)

0 too!

Since `// AUTOSENSE` has been replaced by `// CONSOLE` we should use
the new name for the variable name.

It is deprecated and should be replaced with // CONSOLE.

This commit increases the timeout on the Netty handshake tests because
the previous value could cause timeout exceptions on slow machines.

Relates elastic#18227

Remove performance hack for accessing a document's fields, its not needed.
Add support for accessing is-getter methods like List.isEmpty() as .empty

Closes elastic#18201

This removes dead/duplicate code and makes the `_index` field not configurable.
(Configuration used to jus be ignored, now we would throw an exception if any
is provided.)

…18215

The current code tries to handle the case that document versions are either
missing or stored in payloads rather than doc values. Since none of the 2.x
releases allowed this, we can remove this logic.

The ability to configure index names that are different from the full name was
removed in 2.0.

It currently has twice the same method, once with a MapperService instance and
once with a DocumentMapper. This commits only keeps the former.

All implementations of SearchParseElement have been removed since they are no longer used now that parsing is done on the coordinating node. The SearchParseElement and FetchSubPhaseParseElement classes are not removed as currently they are needed for plugins that add a custom fetch sub phase. These will be removed in a follow up PR that will allow fetch sub phase plugins to register a parser in a different way.

Updated the full text query intro to add `match_phrase` and `match_phrase_prefix`

painless: Array load/store and length with invokedynamic

…elastic#17499

…n really be used like arrays

In 7d1fd17 the parsing of command-line
properties in the plugin script was removed. That commit missed
additional parsing of the properties for es.default.path.conf. This
commit removes that parsing and also replaces the use of eval with exec.

Relates elastic#18239

Improving parsing of sigma param for Extended Stats Bucket Aggregation

… null query

Relates elastic#18051

Today when join validation fails, we log a warning but do not log the
exception that led to the join validation failing. This commit modifies
this so that we do log this exception.

painless: Add synthetic length property as alias to Lists, so they can be used like arrays

This makes defensive copying of ShardRouting objects obsolete whenever we do a reroute and trashes less objects.

An additional sanity check introduced by elastic#17821 makes some tests fail. This check verifies that
only one shard with same shard id is allocated to a node. This commit fixes a bug in
ClusterStateCreationUtils which would construct a cluster state that allocated two shards with same
id to the same node.

This commit switches the command-line scripts to use bash instead of sh
so that we can take advantage of features that bash provides like
arrays.

Relates elastic#18251

In the plugin script we set the setting default.path.conf to control the
path to the configuration file. But the default directory might not
exist in which case we can run into exceptions. This commit restores a
guard against this directory not existing.

I am unable to set ec2 discovery tags because this setting was
accidentally omitted from the register settings list in
Ec2DiscoveryPlugin.java. I get this:

java.lang.IllegalArgumentException: unknown setting [discovery.ec2.tag.project]

painless: use better typing for dynamic method calls

painless: implement needsScore() correctly.

This commit adds a hard requirement to the RPM and Debian packages for
/bin/bash to be present, and adds a note regarding this to the migration
docs.

Relates elastic#18259

This commit modifies the list plugins command to produce deterministic
output by sorting the plugins by comparing paths.

Relates elastic#18260

painless: retrieve _score directly from Scorer

Painless doc access

Fix bracket shortcuts

Add TAG_SETTING to list of allowed tags for the ec2 discovery plugin.

Add 'ctx' keyword to painless.

…00ms

* Docs: First pass at improving analyzer docs

I've rewritten the intro to analyzers plus the docs
for all analyzers to provide working examples.

I've also removed:

* analyzer aliases (see elastic#18244)
* analyzer versions (see elastic#18267)
* snowball analyzer (see elastic#8690)

Next steps will be tokenizers, token filters, char filters

* Fixed two typos

This commit modifies two logging statements in the
IndexingMemoryController to log the key for the setting
indices.memory.index_buffer_size instead of the object.

Relates elastic#18191

Today we encourage users to set their minimum and maximum heap settings
equal to each other to prevent the heap from resizing. Yet, the default
heap settings do not start Elasticsearch in this fashion. This commit
addresses this discrepancy by setting the default min heap to '512m' and
the default max heap to the default min heap.

Relates elastic#16334

…lastic#18031

It will keep using the caching terms enum for keyword/text fields and falls back
to IndexSearcher.count for fields that do not use the inverted index for
searching (such as numbers and ip addresses). Note that this probably means that
significant terms aggregations on these fields will be less efficient than they
used to be. It should be ok under a sampler aggregation though.

This moves tests back to the state they were in before numbers started using
points, and also adds a new test that significant terms aggs fail if a field is
not indexed.

In the long term, we might want to follow the approach that Robert initially
proposed that consists in collecting all documents from the background filter in
order to compute frequencies using doc values. This would also mean that
significant terms aggregations do not require fields to be indexed anymore.

painless: _score as double, not float

still works correclty.

(same name and UUID) exists in the cluster state.  This resolves a
situation where if an index data folder was copied into a node's data
directory while the node is running and that index had a tombstone in
the cluster state, the index would still get imported.

Closes elastic#18250
Closes elastic#18249

…ow the

second type.

request host, as the request is not a cross origin.

Relates elastic#18256

This commit ensures that if CORS is enabled, then Origin headers are
checked regardless of whether the request came from a browser or not.
In the past, we only proceeded with CORS checks if the User-Agent was a
browser.

…sh get, etc) for painless.

but i have no clue how to test it, it seems this feature never worked via REST?

Should we drop the feature instead?

…hanks @s1monw

This commit fixes the Debian package requires clause for bash. For the
RPM it is okay to specify the binary, but for the Debian package the
package name must be specified.

_value support in painess?

painless: remove input, support params instead

Long priority over Float

painless: add debugging tool to tests

Closes elastic#18286

This commit adds bats tests that the RPM and Debian packages depend on
bash.

Relates elastic#18292

Rearranges the FingerprintAnalyzer so that AsciiFolding comes earlier in the chain (after lowercasing, before stop removal, for maximum deduping power)

Closes elastic#18266

This should help the tests pass more consistently. Should.

This also removes from `?pretty` from the docs. It isn't a thing
with `// CONSOLE`.

Relates to
https://elasticsearch-ci.elastic.co/job/elastic+elasticsearch+master+multijob-os-compatibility/os=oraclelinux/399/console

As most of our log messages are not sentences and do not end with
periods, this commit removes a period from the end of the min master
node bootstrap check log message.

Make Line Number Available in Painless

Relates elastic#18274

Pretty sure we're not making a brand new `/var/logs` directory when everything else goes into `/var/log`

Relates elastic#18305

Another unstable snippet....

https://elasticsearch-ci.elastic.co/job/elastic+elasticsearch+master+multijob-os-compatibility/os=sles/402/console

This commit modifies the init.d scripts to use bash now that bash is a
required dependency.

Relates elastic#18308