bt generate: sort by variable name

[thorehusfeldt]

I would like this:

               VALIDATORS               |           PROBLEM STATEMENT
           1 <=    N     <= 10          |              1 <= N <= 10               en,da
           0 <=   d_A    <= 120         |            1 <= t_A <= 120              en,da
           0 <=   d_B    <= 120         |            0 <= d_A <= 120              en,da
           1 <=   t_A    <= 120         |            1 <= t_B <= 120              en,da
           1 <=   t_B    <= 120         |            0 <= d_B <= 120              en,da

to be sorted.

[mpsijm]

I was having the same thought when validating the limits for BAPC. :smile: I was still doubting whether I wanted both to be sorted by alphabetical order (as in the left column), or by "order in the input statement" (as in the right column) :thinking: The former is easier to implement, while the latter feels more natural if some variables are dependent on earlier variables.

[thorehusfeldt]

The latter makes a lot of sense to me. Very good point.

[RagnarGrootKoerkamp]

(I agree this would be nice, but not invested enough myself to implement it right now, so I'll leave this for one of you to implement.)

[thorehusfeldt]

Please try to bt constraint a few problems with commit 7ec4b43 and tell me how that worked out.

[mpsijm]

I tried it on the BAPC problems, and it looks very nice! :smile: I also notice the gaps in case there are no matching variables, that makes it really easy to spot inconsistencies :slightly_smiling_face: For example, for King of the Hill, we used a in the validator but v in the statement:

               VALIDATORS               |           PROBLEM STATEMENT            
           1 <=    n     <= 10000       |            1 <= n <= 10000              en
           1 <=    x     <= 10000       |             1 <= x,y <= n               en
                                        |          1 <= v <= 1000000000           en
           1 <=    a     <= 1000000000  |      

And for International Irregularities, we renamed p to r in the problem statement, but did not update the validator:

               VALIDATORS               |           PROBLEM STATEMENT            
           2 <=    n     <= 100000      |            2 <= n <= 100000             en
           1 <=    q     <= 100000      |            1 <= q <= 100000             en
           0 <=    m     <= 1000000000  |          0 <= m <= 1000000000           en
           0 <=    t     <= 1000000000  | 0 <= r_1 <= \dots <= r_n <= 1000000000  en
                                        |         0 <= t_i <= 1000000000          en
                                        |             1 <= x, y <= n              en
                                        |                 x != y                  en
           0 <=    p     <= 1000000000  |                                        
           1 <=    xy    <= 100000      | 

Thanks for implementing, I think this issue can be closed now :smile: