gf2121 opened a new pull request, #14707:
URL: https://github.com/apache/lucene/pull/14707

   This tries to take advantage of the new method `Scorer#nextDocsAndScores` to 
speed up exhaustive execution of TermQuery.
   
   ```
                               TaskQPS baseline      StdDevQPS 
my_modified_version      StdDev                Pct diff p-value
                       AndStopWords       13.31      (1.1%)       13.01      
(5.1%)   -2.2% (  -8% -    4%) 0.123
                   IntervalsOrdered       31.48      (6.5%)       30.87      
(5.7%)   -1.9% ( -13% -   10%) 0.418
                             IntNRQ       12.38      (6.0%)       12.20      
(5.2%)   -1.5% ( -11% -   10%) 0.510
                        AndHighHigh       18.08      (0.9%)       17.85      
(4.4%)   -1.3% (  -6% -    4%) 0.293
                       SloppyPhrase        1.54      (6.8%)        1.53      
(8.1%)   -1.2% ( -15% -   14%) 0.686
                           SpanNear       34.44      (3.1%)       34.03      
(3.9%)   -1.2% (  -7% -    6%) 0.396
                  FilteredOrHighMed       14.78      (1.9%)       14.61      
(2.0%)   -1.1% (  -4% -    2%) 0.131
                   AndMedOrHighHigh       59.44      (2.6%)       58.82      
(2.8%)   -1.0% (  -6% -    4%) 0.324
               FilteredAndStopWords       18.63      (1.0%)       18.44      
(2.0%)   -1.0% (  -3% -    1%) 0.105
                       FilteredTerm       61.05      (1.6%)       60.51      
(2.4%)   -0.9% (  -4% -    3%) 0.274
                CombinedAndHighHigh        3.89      (1.3%)        3.86      
(3.4%)   -0.8% (  -5% -    3%) 0.407
                   FilteredOr3Terms       21.40      (3.2%)       21.23      
(2.9%)   -0.8% (  -6% -    5%) 0.502
                CountFilteredPhrase       87.15      (2.3%)       86.46      
(2.2%)   -0.8% (  -5% -    3%) 0.370
                    FilteredPrefix3       36.35      (1.2%)       36.07      
(1.9%)   -0.8% (  -3% -    2%) 0.204
                 FilteredOrHighHigh       13.22      (2.5%)       13.12      
(2.6%)   -0.8% (  -5% -    4%) 0.457
                    AndHighOrMedMed       43.97      (1.0%)       43.64      
(1.9%)   -0.7% (  -3% -    2%) 0.216
                FilteredAndHighHigh       30.83      (1.8%)       30.61      
(2.2%)   -0.7% (  -4% -    3%) 0.374
                        CountPhrase        4.02      (2.3%)        4.00      
(3.3%)   -0.6% (  -6% -    5%) 0.604
                 CombinedAndHighMed       67.11      (2.7%)       66.76      
(2.4%)   -0.5% (  -5% -    4%) 0.603
                          And3Terms      195.77      (3.7%)      194.90      
(3.2%)   -0.4% (  -7% -    6%) 0.743
                FilteredOrStopWords        9.39      (2.2%)        9.35      
(2.6%)   -0.4% (  -5% -    4%) 0.669
         FilteredOr2Terms2StopWords        9.81      (1.5%)        9.77      
(2.3%)   -0.4% (  -4% -    3%) 0.613
                             OrMany        0.97      (6.0%)        0.97     
(14.6%)   -0.3% ( -19% -   21%) 0.939
                And2Terms2StopWords      294.44      (1.2%)      293.47      
(1.6%)   -0.3% (  -3% -    2%) 0.554
                             IntSet      578.19      (2.5%)      576.33      
(2.8%)   -0.3% (  -5% -    5%) 0.760
                     FilteredOrMany        2.77      (3.6%)        2.76      
(3.1%)   -0.3% (  -6% -    6%) 0.845
                             Phrase       16.37      (1.4%)       16.33      
(1.6%)   -0.3% (  -3% -    2%) 0.667
                     FilteredPhrase       12.02      (1.5%)       11.99      
(1.3%)   -0.2% (  -3% -    2%) 0.656
        FilteredAnd2Terms2StopWords      354.15      (2.2%)      353.30      
(2.3%)   -0.2% (  -4% -    4%) 0.785
                 FilteredAndHighMed      104.82      (2.5%)      104.68      
(1.7%)   -0.1% (  -4% -    4%) 0.869
                            Respell       71.77      (2.1%)       71.68      
(2.2%)   -0.1% (  -4% -    4%) 0.878
             CountFilteredOrHighMed       47.85      (1.0%)       47.82      
(0.9%)   -0.1% (  -1% -    1%) 0.871
                CountFilteredIntNRQ       19.76      (0.6%)       19.75      
(0.5%)   -0.1% (  -1% -    1%) 0.815
                  TermDayOfYearSort      326.81      (1.2%)      326.81      
(1.0%)    0.0% (  -2% -    2%) 1.000
                             Fuzzy2       81.44      (2.8%)       81.45      
(2.7%)    0.0% (  -5% -    5%) 0.986
                         AndHighMed      114.99      (3.7%)      115.01      
(3.4%)    0.0% (  -6% -    7%) 0.985
                         TermDTSort      224.25      (1.5%)      224.32      
(1.1%)    0.0% (  -2% -    2%) 0.955
                            Prefix3       21.58      (1.9%)       21.60      
(2.9%)    0.1% (  -4% -    4%) 0.933
                       CombinedTerm       28.09      (1.1%)       28.12      
(1.4%)    0.1% (  -2% -    2%) 0.839
                           PKLookup      377.96      (2.4%)      378.76      
(1.9%)    0.2% (  -4% -    4%) 0.805
            CountFilteredOrHighHigh       41.49      (1.6%)       41.59      
(1.1%)    0.2% (  -2% -    3%) 0.650
                     FilteredIntNRQ       18.56      (1.0%)       18.61      
(0.9%)    0.3% (  -1% -    2%) 0.488
                           Wildcard       18.78      (1.9%)       18.84      
(2.5%)    0.3% (  -4% -    4%) 0.756
                  FilteredAnd3Terms      606.79      (4.8%)      609.36      
(2.1%)    0.4% (  -6% -    7%) 0.772
                             Fuzzy1       84.06      (2.6%)       84.47      
(3.1%)    0.5% (  -5% -    6%) 0.667
                    CountAndHighMed      136.86      (4.9%)      137.58      
(3.2%)    0.5% (  -7% -    9%) 0.746
                      TermTitleSort      334.08      (2.5%)      335.99      
(1.5%)    0.6% (  -3% -    4%) 0.480
                   DismaxOrHighHigh        2.70      (2.3%)        2.71      
(2.8%)    0.6% (  -4% -    5%) 0.541
                          CountTerm     9906.18      (6.0%)     9973.27      
(8.0%)    0.7% ( -12% -   15%) 0.808
                           Or3Terms       20.95      (4.1%)       21.10     
(14.4%)    0.7% ( -17% -   19%) 0.867
                    CountOrHighHigh       77.01      (2.2%)       77.73      
(1.7%)    0.9% (  -2% -    4%) 0.228
                    DismaxOrHighMed        7.20      (2.3%)        7.27      
(3.6%)    1.0% (  -4% -    7%) 0.421
                CountFilteredOrMany        9.03      (3.7%)        9.12      
(2.3%)    1.0% (  -4% -    7%) 0.402
                   CountAndHighHigh       92.10      (3.8%)       93.13      
(2.1%)    1.1% (  -4% -    7%) 0.361
                 Or2Terms2StopWords        2.50      (6.4%)        2.53     
(14.2%)    1.5% ( -17% -   23%) 0.720
                     CountOrHighMed      101.42      (4.2%)      103.09      
(3.3%)    1.7% (  -5% -    9%) 0.263
                          OrHighMed       30.02      (4.6%)       30.53     
(13.9%)    1.7% ( -16% -   21%) 0.674
                         OrHighRare       46.31      (2.3%)       47.10      
(2.0%)    1.7% (  -2% -    6%) 0.042
                        OrStopWords        3.34      (6.1%)        3.41     
(14.1%)    2.0% ( -17% -   23%) 0.635
                         OrHighHigh        8.30      (5.6%)        8.50     
(13.8%)    2.4% ( -16% -   23%) 0.568
                      TermMonthSort     3162.23      (7.1%)     3251.96      
(4.3%)    2.8% (  -7% -   15%) 0.215
                        CountOrMany       11.53      (6.7%)       11.87      
(4.2%)    3.0% (  -7% -   14%) 0.178
                 CombinedOrHighHigh        1.47      (8.9%)        1.51      
(8.6%)    3.0% ( -13% -   22%) 0.377
                  CombinedOrHighMed        3.97      (8.6%)        4.10      
(8.6%)    3.2% ( -12% -   22%) 0.337
                         DismaxTerm       51.75      (3.6%)       67.36     
(14.3%)   30.2% (  11% -   49%) 0.000
                               Term       27.45      (5.0%)       40.57     
(22.3%)   47.8% (  19% -   79%) 0.000
   ```


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org

Reply via email to