Thanks for the patch and review @sacha_vrand and @leper!
Could have used this very well already for D145, D156, D229 and D204. In each of these occurances I had non-visually replayed an empty extended replay file to see how my trigger script would behave. The optional and disabled by default debug messages are still in there, so that these unit-like tests can now be run from command-line to see that the scripts still work and spawn intended number of enemies (rise the water at a reasonable rate):

DEBUG [4]  Spawning 4 attackers
DEBUG [4]  Siege Ratio: 21%
DEBUG [4]  Spawning:({siege:1, champions:3})
DEBUG [4]  Templates: [{template:"units/gaul_mechanical_siege_ram", count:1}, {template:"units/gaul_champion_cavalry", count:1}, {template:"units/gaul_champion_fanatic", count:2}, {template:"units/gaul_champion_infantry", count:0}]
DEBUG [7]  Spawning 6 attackers
DEBUG [7]  Siege Ratio: 40%
DEBUG [7]  Spawning:({siege:2, champions:4})
DEBUG [7]  Templates: [{template:"units/iber_mechanical_siege_ram", count:2}, {template:"units/iber_champion_cavalry", count:2}, {template:"units/iber_champion_infantry", count:2}]
DEBUG [10]  Spawning 6 attackers
DEBUG [10]  Siege Ratio: 34%
DEBUG [10]  Spawning:({siege:2, champions:4})
DEBUG [10]  Templates: [{template:"units/rome_mechanical_siege_ballista_packed", count:1}, {template:"units/rome_mechanical_siege_ram", count:1}, {template:"units/rome_mechanical_siege_scorpio_packed", count:0}, {template:"units/rome_centurio_imperial", count:1}, {template:"units/rome_champion_cavalry", count:1}, {template:"units/rome_champion_infantry", count:1}, {template:"units/rome_legionnaire_imperial", count:1}, {template:"units/rome_legionnaire_marian", count:0}]

https://trac.wildfiregames.com/wiki/PetraBot#Batchsimulatinggames (maybe others?) should be updated I guess.

In D379#21750, @sacha_vrand wrote:

As an example of what we can do with that, here is a graph showing the win percentage of the first player on the map 'skirmishes/Acropolis Bay (2)' (doesn't seem well balanced). 250 games in conquest with Petra against Petra and a seed from 0 to 249. What do you guys think ?
Graph's link

Presumably AI devs would be very happy to find out how such graphs can be produced.
I heard you created this patch as a student project? How did it go, maybe you want to share some material (slideshow or something)?

In rP19645#21453, @elexis wrote:

Fixed by r19683.
It's not crazy to remove trailing whitespace, but I agree that it's not worth raising a concern for whitespace IMO and
what pushed me to raise a concern is not the coding convention but the conern raised in rP19667

I agree fixing the whitespace is justified, my worry is rather to the fact that we put concern for that, which require more work than fixing it directly.
In addition, the main concern should rather have been on the fact that this file use blank rather than tabs for indentation (i just noticed it when modifying the file)
And rP19667 is a bit different as we have there two competing coding conventions, and so the best way to fix it could be debatable, while it was not when only removing two trailing spaces, but i'm fine with your rP19684 :)

Hi @sacha_vrand, in case you read this and have some more interest in playing with this code, gentz has reported a bug here: #5016