f04bc944-f9b4-4a42-8b26-dcacd0e3e688

Committed 11 Mar 2025 10:27PM UTC coverage: 34.017% (-58.1%) from 92.162%

Build # f04bc944-f9b4-4a42-8b26-dcacd0e3e688

Build Type

Pull #2653

circleci

Committed by

christinach

Commit Message

Add new lc_subject_facet field.
Helps with the vocabulary work https://github.com/pulibrary/orangelight/pull/3386
In this new field we index only the lc subject heading and the subdivisions
So that when the user searches using the Details section, they can query solr for
all the subject headings and their divisions.

This is needed for the Subject browse Vocabulary work.
example: "lc_subject_facet": [
             "Booksellers and bookselling—Italy—Directories",
             "Booksellers and bookselling-Italy",
             "Booksellers and bookselling"
              ]

Pull Request Pull Request #2653: Add new lc_subject_facet field.

Run Details

1 of 3 new or added lines in 1 file covered. (33.33%)

2215 existing lines in 93 files now uncovered.

1294 of 3804 relevant lines covered (34.02%)

0.99 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

24.39

/marc_to_solr/lib/solr_deleter.rb

require 'net/http'
require 'json'

# This class allow us to delete Solr records with our current
# version of Traject. Once we upgrade to Traject version 3.1
# or later we could use Traject's native delete feature
# (see https://github.com/traject/traject/commit/c731d09c6cb15fbe572e4a294a7a127b525b7277)
# but keep in mind that Traject's implementation is still in experimental mode
# and it does not (yet) support batches.
class SolrDeleter
  def initialize(solr_url, logger = nil)
    @solr_url = solr_url
    @logger = logger
  end

  def delete(ids, batch_size = 100)
    ids.each_slice(batch_size) do |batch|
      delete_batch(batch)
    end
  end

  private

    # content_type = "text/xml"
    # request payload is XML (even if the response is in JSON via the wt=json param)
    def request_deletion(uri:, body:, content_type: 'text/xml')
      @logger&.info "Deleting #{body}"
      start_time = Time.now
      response = Faraday.post(uri) do |req|
        req.headers = { 'Content-Type' => content_type }
        req.body = body
        req.options.open_timeout = 60
        req.options.read_timeout = 60
        req.options.write_timeout = 60
      end
      @logger&.info("Delete completed in #{Time.now - start_time} seconds. URL: #{uri}")
      response
    rescue Faraday::TimeoutError
      @logger&.warn("Delete timed out after #{Time.now - start_time} seconds. URL: #{uri} : #{body}")
      nil
    end

    def build_request_body(ids:)
      output = ['<delete>']
      ids.each do |id|
        output << "<id>#{id}</id>"
      end

      output << '</delete>'
      output.join
    end

    def valid_response?(response)
      return false if response.nil?

      response.status == 200
    end

    def delete_batch(batch)
      uri = "#{@solr_url}/update?waitSearcher=false&wt=json"
      body = build_request_body(ids: batch)

      response = request_deletion(uri:, body:)
      return if valid_response?(response)

      # Only retry once
      retry_response = request_deletion(uri:, body:)
      return if valid_response?(retry_response)

      unless retry_response.nil?
        Honeybadger.notify("Error deleting Solr documents. IDs: #{batch.join(', ')}. Status: #{retry_response.status}. Body: #{retry_response.body}")
      end
    end
end

1	require 'net/http'	1✔
2	require 'json'	1✔
3
4	# This class allow us to delete Solr records with our current
5	# version of Traject. Once we upgrade to Traject version 3.1
6	# or later we could use Traject's native delete feature
7	# (see https://github.com/traject/traject/commit/c731d09c6cb15fbe572e4a294a7a127b525b7277)
8	# but keep in mind that Traject's implementation is still in experimental mode
9	# and it does not (yet) support batches.
10	class SolrDeleter	1✔
11	def initialize(solr_url, logger = nil)	1✔
UNCOV 12	@solr_url = solr_url	×
UNCOV 13	@logger = logger	×
14	end
15
16	def delete(ids, batch_size = 100)	1✔
UNCOV 17	ids.each_slice(batch_size) do \|batch\|	×
UNCOV 18	delete_batch(batch)	×
19	end
20	end
21
22	private	1✔
23
24	# content_type = "text/xml"
25	# request payload is XML (even if the response is in JSON via the wt=json param)
26	def request_deletion(uri:, body:, content_type: 'text/xml')	1✔
UNCOV 27	@logger&.info "Deleting #{body}"	×
UNCOV 28	start_time = Time.now	×
UNCOV 29	response = Faraday.post(uri) do \|req\|	×
UNCOV 30	req.headers = { 'Content-Type' => content_type }	×
UNCOV 31	req.body = body	×
UNCOV 32	req.options.open_timeout = 60	×
UNCOV 33	req.options.read_timeout = 60	×
UNCOV 34	req.options.write_timeout = 60	×
35	end
UNCOV 36	@logger&.info("Delete completed in #{Time.now - start_time} seconds. URL: #{uri}")	×
UNCOV 37	response	×
38	rescue Faraday::TimeoutError
UNCOV 39	@logger&.warn("Delete timed out after #{Time.now - start_time} seconds. URL: #{uri} : #{body}")	×
UNCOV 40	nil	×
41	end
42
43	def build_request_body(ids:)	1✔
UNCOV 44	output = ['<delete>']	×
UNCOV 45	ids.each do \|id\|	×
UNCOV 46	output << "<id>#{id}</id>"	×
47	end
48
UNCOV 49	output << '</delete>'	×
UNCOV 50	output.join	×
51	end
52
53	def valid_response?(response)	1✔
UNCOV 54	return false if response.nil?	×
55
UNCOV 56	response.status == 200	×
57	end
58
59	def delete_batch(batch)	1✔
UNCOV 60	uri = "#{@solr_url}/update?waitSearcher=false&wt=json"	×
UNCOV 61	body = build_request_body(ids: batch)	×
62
UNCOV 63	response = request_deletion(uri:, body:)	×
UNCOV 64	return if valid_response?(response)	×
65
66	# Only retry once
UNCOV 67	retry_response = request_deletion(uri:, body:)	×
UNCOV 68	return if valid_response?(retry_response)	×
69
UNCOV 70	unless retry_response.nil?	×
UNCOV 71	Honeybadger.notify("Error deleting Solr documents. IDs: #{batch.join(', ')}. Status: #{retry_response.status}. Body: #{retry_response.body}")	×
72	end
73	end
74	end

pulibrary / bibdata / f04bc944-f9b4-4a42-8b26-dcacd0e3e688

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous