Represent every values as text

[laurenthuberdeau]

Results:

Small improvement. Considering that the change is small and doesn't impact the ease of development, I suggest we merge it. We'll just have to be careful to not silently pack/unpack the data by using the pattern synonyms in Semantics.

Before:

benchmarking clique/5
time                 92.27 μs   (91.50 μs .. 93.54 μs)
                     0.999 R²   (0.998 R² .. 1.000 R²)
mean                 93.09 μs   (92.42 μs .. 93.81 μs)
std dev              2.302 μs   (1.893 μs .. 2.923 μs)
variance introduced by outliers: 21% (moderately inflated)

benchmarking clique/10
time                 333.2 μs   (332.1 μs .. 334.6 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 333.3 μs   (332.5 μs .. 334.5 μs)
std dev              2.966 μs   (2.033 μs .. 4.108 μs)

benchmarking clique/25
time                 2.258 ms   (2.250 ms .. 2.269 ms)
                     1.000 R²   (0.999 R² .. 1.000 R²)
mean                 2.268 ms   (2.260 ms .. 2.282 ms)
std dev              32.68 μs   (23.77 μs .. 50.49 μs)

benchmarking clique/40
time                 5.671 ms   (5.649 ms .. 5.696 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 5.657 ms   (5.641 ms .. 5.668 ms)
std dev              41.02 μs   (33.25 μs .. 52.26 μs)

benchmarking line/5
time                 89.98 μs   (89.39 μs .. 90.80 μs)
                     0.999 R²   (0.999 R² .. 1.000 R²)
mean                 89.22 μs   (88.71 μs .. 89.87 μs)
std dev              1.948 μs   (1.427 μs .. 2.500 μs)
variance introduced by outliers: 18% (moderately inflated)

benchmarking line/10
time                 371.5 μs   (370.4 μs .. 372.9 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 371.6 μs   (370.9 μs .. 372.5 μs)
std dev              2.809 μs   (2.343 μs .. 3.419 μs)

benchmarking line/25
time                 3.178 ms   (3.156 ms .. 3.220 ms)
                     0.999 R²   (0.998 R² .. 1.000 R²)
mean                 3.182 ms   (3.165 ms .. 3.207 ms)
std dev              68.43 μs   (51.06 μs .. 89.99 μs)

benchmarking line/40
time                 10.65 ms   (10.57 ms .. 10.78 ms)
                     0.999 R²   (0.998 R² .. 1.000 R²)
mean                 10.69 ms   (10.64 ms .. 10.76 ms)
std dev              158.2 μs   (117.6 μs .. 202.3 μs)

benchmarking loop/50
time                 494.2 μs   (490.0 μs .. 499.0 μs)
                     0.999 R²   (0.998 R² .. 0.999 R²)
mean                 492.9 μs   (488.3 μs .. 499.6 μs)
std dev              18.57 μs   (14.20 μs .. 24.43 μs)
variance introduced by outliers: 30% (moderately inflated)

benchmarking loop/100
time                 1.401 ms   (1.393 ms .. 1.413 ms)
                     0.999 R²   (0.999 R² .. 1.000 R²)
mean                 1.411 ms   (1.406 ms .. 1.419 ms)
std dev              21.32 μs   (15.94 μs .. 33.33 μs)

benchmarking loop/500
time                 21.63 ms   (21.38 ms .. 21.91 ms)
                     0.999 R²   (0.998 R² .. 1.000 R²)
mean                 21.21 ms   (21.10 ms .. 21.38 ms)
std dev              321.7 μs   (214.5 μs .. 500.1 μs)

benchmarking loop/1000
time                 83.70 ms   (83.27 ms .. 84.30 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 83.69 ms   (83.42 ms .. 84.04 ms)
std dev              491.0 μs   (300.5 μs .. 784.5 μs)

benchmarking tight/50
time                 202.9 μs   (200.8 μs .. 205.2 μs)
                     0.999 R²   (0.999 R² .. 0.999 R²)
mean                 207.1 μs   (204.8 μs .. 210.4 μs)
std dev              9.345 μs   (6.641 μs .. 14.57 μs)
variance introduced by outliers: 44% (moderately inflated)

benchmarking tight/100
time                 719.7 μs   (718.6 μs .. 720.9 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 723.3 μs   (721.7 μs .. 725.7 μs)
std dev              6.186 μs   (4.581 μs .. 8.501 μs)

benchmarking tight/500
time                 16.53 ms   (16.49 ms .. 16.57 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 16.53 ms   (16.51 ms .. 16.55 ms)
std dev              54.79 μs   (40.68 μs .. 81.67 μs)

benchmarking tight/1000
time                 71.07 ms   (70.91 ms .. 71.27 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 71.07 ms   (70.98 ms .. 71.15 ms)
std dev              159.1 μs   (116.5 μs .. 234.5 μs)

benchmarking parse/50
time                 6.128 ms   (6.102 ms .. 6.153 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 6.122 ms   (6.105 ms .. 6.137 ms)
std dev              46.25 μs   (35.18 μs .. 67.21 μs)

benchmarking parse/100
time                 12.29 ms   (12.19 ms .. 12.47 ms)
                     0.998 R²   (0.995 R² .. 1.000 R²)
mean                 12.20 ms   (12.15 ms .. 12.38 ms)
std dev              237.8 μs   (81.32 μs .. 459.2 μs)

benchmarking parse/500
time                 68.82 ms   (67.25 ms .. 70.19 ms)
                     0.999 R²   (0.999 R² .. 1.000 R²)
mean                 67.61 ms   (67.15 ms .. 68.14 ms)
std dev              903.0 μs   (593.0 μs .. 1.430 ms)

benchmarking parse/1000
time                 136.0 ms   (135.5 ms .. 136.3 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 135.8 ms   (135.1 ms .. 136.1 ms)
std dev              693.6 μs   (261.1 μs .. 1.136 ms)
variance introduced by outliers: 11% (moderately inflated)

Now:

benchmarking clique/5  
time                 86.66 μs   (86.54 μs .. 86.79 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 86.97 μs   (86.84 μs .. 87.15 μs)
std dev              502.2 ns   (395.5 ns .. 642.6 ns)
                       
benchmarking clique/10 
time                 318.3 μs   (317.7 μs .. 318.8 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 318.3 μs   (317.9 μs .. 318.7 μs)
std dev              1.275 μs   (1.091 μs .. 1.568 μs)
                       
benchmarking clique/25 
time                 2.183 ms   (2.155 ms .. 2.215 ms)
                     0.998 R²   (0.996 R² .. 1.000 R²)
mean                 2.174 ms   (2.158 ms .. 2.215 ms)
std dev              85.12 μs   (29.05 μs .. 156.3 μs)
variance introduced by outliers: 24% (moderately inflated)
                       
benchmarking clique/40 
time                 5.467 ms   (5.407 ms .. 5.535 ms)
                     0.999 R²   (0.999 R² .. 1.000 R²)
mean                 5.399 ms   (5.377 ms .. 5.428 ms)
std dev              80.40 μs   (60.65 μs .. 101.5 μs)
                       
benchmarking line/5    
time                 85.57 μs   (85.47 μs .. 85.68 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 85.62 μs   (85.54 μs .. 85.72 μs)
std dev              305.2 ns   (247.2 ns .. 377.1 ns)
                       
benchmarking line/10   
time                 354.1 μs   (353.7 μs .. 354.6 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 354.1 μs   (353.7 μs .. 354.5 μs)
std dev              1.286 μs   (1.060 μs .. 1.688 μs)
                       
benchmarking line/25   
time                 2.987 ms   (2.974 ms .. 3.005 ms)
                     1.000 R²   (0.999 R² .. 1.000 R²)
mean                 2.976 ms   (2.972 ms .. 2.986 ms)
std dev              22.72 μs   (14.04 μs .. 41.08 μs)
                       
benchmarking line/40   
time                 10.14 ms   (9.999 ms .. 10.23 ms)
                     0.999 R²   (0.999 R² .. 1.000 R²)
mean                 9.950 ms   (9.917 ms .. 10.01 ms)
std dev              118.9 μs   (83.44 μs .. 150.1 μs)
                       
benchmarking loop/50   
time                 470.3 μs   (462.8 μs .. 476.5 μs)
                     0.999 R²   (0.997 R² .. 1.000 R²)
mean                 460.4 μs   (458.1 μs .. 464.3 μs)
std dev              10.18 μs   (6.753 μs .. 17.01 μs)
variance introduced by outliers: 13% (moderately inflated)
                       
benchmarking loop/100  
time                 1.334 ms   (1.327 ms .. 1.340 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 1.339 ms   (1.337 ms .. 1.342 ms)
std dev              9.650 μs   (7.881 μs .. 12.18 μs)
                       
benchmarking loop/500  
time                 20.16 ms   (20.02 ms .. 20.33 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 20.10 ms   (20.03 ms .. 20.18 ms)
std dev              172.6 μs   (126.2 μs .. 238.3 μs)
                       
benchmarking loop/1000 
time                 79.66 ms   (78.37 ms .. 83.24 ms)
                     0.997 R²   (0.990 R² .. 1.000 R²)
mean                 79.51 ms   (78.79 ms .. 81.74 ms)
std dev              1.859 ms   (487.7 μs .. 3.167 ms)
                       
benchmarking tight/50  
time                 186.1 μs   (185.7 μs .. 186.8 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 187.0 μs   (186.4 μs .. 188.0 μs)
std dev              2.400 μs   (1.341 μs .. 4.237 μs)
                       
benchmarking tight/100 
time                 675.2 μs   (673.7 μs .. 676.7 μs)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 677.1 μs   (675.8 μs .. 678.7 μs)
std dev              5.148 μs   (4.071 μs .. 7.105 μs)
                       
benchmarking tight/500 
time                 15.51 ms   (15.37 ms .. 15.63 ms)
                     1.000 R²   (0.999 R² .. 1.000 R²)
mean                 15.73 ms   (15.60 ms .. 16.13 ms)
std dev              508.2 μs   (152.4 μs .. 992.6 μs)
variance introduced by outliers: 12% (moderately inflated)
                       
benchmarking tight/1000
time                 67.31 ms   (66.96 ms .. 67.74 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 67.53 ms   (67.23 ms .. 68.53 ms)
std dev              839.8 μs   (277.8 μs .. 1.430 ms)
                       
benchmarking parse/50  
time                 6.063 ms   (6.038 ms .. 6.086 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 6.055 ms   (6.036 ms .. 6.089 ms)
std dev              72.71 μs   (51.55 μs .. 114.3 μs)
                       
benchmarking parse/100 
time                 12.05 ms   (11.97 ms .. 12.11 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 12.01 ms   (11.97 ms .. 12.04 ms)
std dev              91.99 μs   (77.59 μs .. 118.5 μs)
                       
benchmarking parse/500 
time                 66.37 ms   (66.01 ms .. 66.71 ms)
                     1.000 R²   (1.000 R² .. 1.000 R²)
mean                 66.79 ms   (66.57 ms .. 67.39 ms)
std dev              609.9 μs   (265.2 μs .. 1.038 ms)
                       
benchmarking parse/1000
time                 137.4 ms   (134.2 ms .. 140.6 ms)
                     0.999 R²   (0.998 R² .. 1.000 R²)
mean                 135.1 ms   (134.0 ms .. 136.5 ms)
std dev              1.907 ms   (1.253 ms .. 3.004 ms)
variance introduced by outliers: 11% (moderately inflated)

[laurenthuberdeau]

Same trick that I used when I changed the representation of IP addresses in portal-suite:
https://github.com/Simspace/portal-suite/blob/dev/common-utilities/haskell/src/SimSpace/Utils/Cidr.hs#L36-L43

[laurenthuberdeau]

Also, this pattern is complete in a strange way because the cases are not disjoint. For example, the following code always returns "true" because T matches everything.

case I 5 of
  T _ -> "text"
  I _  -> "int"
  B _ -> "bool"