Support Elastic Search's HTML strip char filter

https://gitlab.zendesk.com/agent/tickets/137301 (Internal)

As reported in gitlab#10693 (closed), certain code cases are not considered.

The terms searched in the ZD ticket live in a pom.xml, and according to @amulvany, the search input is being interpreted incorrectly:

artifactId-spring-boot-starter-parent/artifactId is being filtered as a path
artifactId-spring-boot-starter-parent as a single string which is common
spring-boot-starter-parent as a single string which is not so common.

The suggested solution is to implement Html Strip Char Filter in order to return the decoded value.

Example:

POST _analyze
{
  "tokenizer":      "keyword", 
  "char_filter":  [ "html_strip" ],
  "text": "<p>I&apos;m so <b>happy</b>!</p>"
}

Response:

[ \nI'm so happy!\n ] (instead of [ I'm, so, happy ]`