Stop using deprecated textNoStem fields - use textUnstemmed instead

[ndushay]

Prereq: integration tests in Argo for appropriate functionality of Solr fields.

The goal of this ticket is to stop using some deprecated solr field types for Argo.

In argo-xx schema.xml:

    <!-- Text tokenized without stemming -->
    <dynamicField name="*_text_unstemmed_i"   type="textUnstemmed" indexed="true"  stored="false" multiValued="false"/>
    <dynamicField name="*_text_unstemmed_im"  type="textUnstemmed" indexed="true"  stored="false" multiValued="true"/>
    <dynamicField name="*_text_unstemmed_si"  type="textUnstemmed" indexed="true"  stored="true"  multiValued="false"/>
    <dynamicField name="*_text_unstemmed_sim" type="textUnstemmed" indexed="true"  stored="true"  multiValued="true"/>
    <!-- DEPRECATED:  textNoStem is a deprecated type -->
    <dynamicField name="*_text_nostem_i"  type="textNoStem" indexed="true"  stored="false" multiValued="false"/>
    <dynamicField name="*_text_nostem_im" type="textNoStem" indexed="true"  stored="false" multiValued="true"/>
...
snip
...0
   <!-- DEPRECATED:  use textUnstemmed, as WordDelimiterFilterFactory is deprecated.  Analyzed Text, no Stemming or Synonyms -->
    <fieldtype name="textNoStem" class="solr.TextField" positionIncrementGap="100" autoGeneratePhraseQueries="true">
      <analyzer>
        <tokenizer class="solr.WhitespaceTokenizerFactory"/>
        <filter class="solr.ICUFoldingFilterFactory"/>
        <!-- NFKC, case folding, diacritics removed -->
        <filter class="solr.WordDelimiterFilterFactory" splitOnCaseChange="1" generateWordParts="1" catenateWords="1" splitOnNumerics="0" generateNumberParts="1" catenateNumbers="1" catenateAll="0" preserveOriginal="0" stemEnglishPossessive="0"/>
        <filter class="solr.RemoveDuplicatesTokenFilterFactory"/>
      </analyzer>
    </fieldtype>

Existing fields using type TextNoStem

• descriptive_text_nostem_i
• source_id_text_nostem_i
• author_text_nostem_im
• contributor_text_nostem_im

NOTE: if a single stored field will do the job, store it (e.g. (primary) author field can be used for display and search if stored)

Steps

• sul-solr-configs allow new fields
  • qa
  • stage
  • prod
• argo's solr config files
• dor-service-app solr config files
• dor_indexing gem populates new fields of type TextUnstemmed (in addition to old types)
  • PR for this
  • release new dor-indexing gem version
• dor_indexing_app uses new gem release (and populates new fields)
  • PR for this
  • deployed to qa
  • deployed to stage
  • deployed to prod
• dor-services-app app uses new gem release (and populates new fields)
  • PR for this
  • deployed to qa
  • deployed to stage
  • deployed to prod
• new field population tested
  • solr queries for argo-qa
  • solr queries for argo-stage
• argo solr prod fields have begun populated
• argo solr prod fields have finished populated
• argo PR (on HOLD) for replacing usage of old field names
  • argo changes deployed to qa
  • argo changes deployed to stage
  • argo tested and vetted on qa and/or stage
• sul-solr-configs argo_xx files PR to use new fields in search results for stage and qa
  • solr config changes deployed to qa
  • solr config changes deployed to stage
• argo solr prod new fields have begun populating (starting: )
• argo solr prod new fields have finished populating
• roll out new fields in argo prod
  • add prod to argo PR (take out of HOLD) for replacing usage of old field names
    • argo changes deployed to prod
  • argo solr config file PRs to use new fields in search results
    • sul-solr config changes deployed to prod
    • argo repo gets solr config changes
    • dor-services-app repo gets solr config changes
• remove old fields
  • argo
  • dor_indexing_app
  • indexes
    • qa
    • stage
    • prod
• sul-solr-config removes field types
• argo Solr config files updated in the Solr cluster
  • qa
  • stage
  • prod
• argo repo solr config changes
• dor-services-app repo solr config changes