pulibrary / bibdata / 2361e00e-a596-45a2-bf05-b1c8fe4e069e

90%

main: 90%

Pull #3105

circleci

Re-implement some of AugmentTheSubject in rust

This triples the speed of AugmentTheSubject in the worst case
(i.e. no subjects match), which is also the most common case:

```
require 'benchmark/ips'
record = MARC::XMLReader.new('spec/fixtures/marc_to_solr/9918309193506421.mrx').first
augment_the_subject = AugmentTheSubject.new
headings = process_hierarchy(record, '600|*0|abcdfklmnopqrtvxyz:610|*0|abfklmnoprstvxyz:611|*0|abcdefgklnpqstvxyz:630|*0|adfgklmnoprstvxyz:650|*0|abcvxyz:651|*0|avxyz')
Benchmark.ips do |x|
  x.report { augment_the_subject.add_indigenous_studies(headings) }
end
```

```
Before
ruby 3.4.4 (2025-05-14 revision a38531fd3f) +YJIT +PRISM [arm64-darwin24]
Warming up --------------------------------------
                        21.499k i/100ms
Calculating -------------------------------------
                        214.290k (± 1.1%) i/s    (4.67 μs/i) -      1.075M in   5.017005s

After
ruby 3.4.4 (2025-05-14 revision a38531fd3f) +YJIT +PRISM [arm64-darwin24]
Warming up --------------------------------------
                        61.096k i/100ms
Calculating -------------------------------------
                        616.234k (± 0.9%) i/s    (1.62 μs/i) -      3.116M in   5.056770s
```

Helps with #2871

31 of 54 new or added lines in 3 files covered. (57.41%)

9665 of 10753 relevant lines covered (89.88%)

281.41 hits per line