b3c4a532-5da0-46f8-9c20-20018605a028

Committed 21 Mar 2025 08:11PM UTC coverage: 42.257% (-49.9%) from 92.185%

Build # b3c4a532-5da0-46f8-9c20-20018605a028

Build Type

push

circleci

Committed by

sandbergja

Commit Message

Move various gems from the default Gemfile group to a more appropriate group

Run Details

1599 of 3784 relevant lines covered (42.26%)

2.56 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✔
12	@solr_url = solr_url	×
13	@logger = logger	×
14	end
15
16	def delete(ids, batch_size = 100)	1✔
17	ids.each_slice(batch_size) do \|batch\|	×
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✔
27	@logger&.info "Deleting #{body}"	×
28	start_time = Time.now	×
29	response = Faraday.post(uri) do \|req\|	×
30	req.headers = { 'Content-Type' => content_type }	×
31	req.body = body	×
32	req.options.open_timeout = 60	×
33	req.options.read_timeout = 60	×
34	req.options.write_timeout = 60	×
35	end
36	@logger&.info("Delete completed in #{Time.now - start_time} seconds. URL: #{uri}")	×
37	response	×
38	rescue Faraday::TimeoutError
39	@logger&.warn("Delete timed out after #{Time.now - start_time} seconds. URL: #{uri} : #{body}")	×
40	nil	×
41	end
42
43	def build_request_body(ids:)	1✔
44	output = ['<delete>']	×
45	ids.each do \|id\|	×
46	output << "<id>#{id}</id>"	×
47	end
48
49	output << '</delete>'	×
50	output.join	×
51	end
52
53	def valid_response?(response)	1✔
54	return false if response.nil?	×
55
56	response.status == 200	×
57	end
58
59	def delete_batch(batch)	1✔
60	uri = "#{@solr_url}/update?waitSearcher=false&wt=json"	×
61	body = build_request_body(ids: batch)	×
62
63	response = request_deletion(uri:, body:)	×
64	return if valid_response?(response)	×
65
66	# Only retry once
67	retry_response = request_deletion(uri:, body:)	×
68	return if valid_response?(retry_response)	×
69
70	unless retry_response.nil?	×
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 / b3c4a532-5da0-46f8-9c20-20018605a028

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