local _ENV = mkmodule('plugins.sort') local gui = require('gui') local overlay = require('plugins.overlay') local setbelief = reqscript('modtools/set-belief') local utils = require('utils') local widgets = require('gui.widgets') local GLOBAL_KEY = 'sort' local CH_UP = string.char(30) local CH_DN = string.char(31) local function get_rating(val, baseline, range, highest, high, med, low) val = val - (baseline or 0) range = range or 100 local percentile = (math.min(range, val) * 100) // range if percentile < (low or 25) then return percentile, COLOR_RED end if percentile < (med or 50) then return percentile, COLOR_LIGHTRED end if percentile < (high or 75) then return percentile, COLOR_YELLOW end if percentile < (highest or 90) then return percentile, COLOR_GREEN end return percentile, COLOR_LIGHTGREEN end local function sort_noop(a, b) -- this function is used as a marker and never actually gets called error('sort_noop should not be called') end local function get_name(unit) return unit and dfhack.toSearchNormalized(dfhack.TranslateName(dfhack.units.getVisibleName(unit))) end local function sort_by_name_desc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local name1 = get_name(unit1) local name2 = get_name(unit2) return utils.compare_name(name1, name2) end local function sort_by_name_asc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local name1 = get_name(unit1) local name2 = get_name(unit2) return utils.compare_name(name2, name1) end local active_units = df.global.world.units.active local active_idx_cache = {} local function get_active_idx_cache() local num_active_units = #active_units if num_active_units == 0 or active_idx_cache[active_units[num_active_units-1].id] ~= num_active_units-1 then active_idx_cache = {} for i,active_unit in ipairs(active_units) do active_idx_cache[active_unit.id] = i end end return active_idx_cache end local function is_original_dwarf(unit) return df.global.plotinfo.fortress_age == unit.curse.time_on_site // 10 end local WAVE_END_GAP = 10000 local function get_most_recent_wave_oldest_active_idx(cache) local oldest_unit for idx=#active_units-1,0,-1 do local unit = active_units[idx] if not dfhack.units.isCitizen(unit) then goto continue end if oldest_unit and unit.curse.time_on_site - oldest_unit.curse.time_on_site > WAVE_END_GAP then return cache[oldest_unit.id] else oldest_unit = unit end ::continue:: end end -- return green for most recent wave, red for the first wave, yellow for all others -- rating is a three digit number that indicates the (potentially approximate) order local function get_arrival_rating(unit) local cache = get_active_idx_cache() local unit_active_idx = cache[unit.id] if not unit_active_idx then return end local most_recent_wave_oldest_active_idx = get_most_recent_wave_oldest_active_idx(cache) if not most_recent_wave_oldest_active_idx then return end local num_active_units = #active_units local rating = num_active_units < 1000 and unit_active_idx or ((unit_active_idx * 1000) // #active_units) if most_recent_wave_oldest_active_idx < unit_active_idx then return rating, COLOR_LIGHTGREEN end if is_original_dwarf(unit) then return rating, COLOR_RED end return rating, COLOR_YELLOW end local function sort_by_arrival_desc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local cache = get_active_idx_cache() if not cache[unit_id_1] then return -1 end if not cache[unit_id_2] then return 1 end return utils.compare(cache[unit_id_2], cache[unit_id_1]) end local function sort_by_arrival_asc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local cache = get_active_idx_cache() if not cache[unit_id_1] then return -1 end if not cache[unit_id_2] then return 1 end return utils.compare(cache[unit_id_1], cache[unit_id_2]) end local function get_stress(unit) return unit and unit.status.current_soul and unit.status.current_soul.personality.stress end local function get_stress_rating(unit) return get_rating(dfhack.units.getStressCategory(unit), 0, 100, 4, 3, 2, 1) end local function sort_by_stress_desc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local happiness1 = get_stress(unit1) local happiness2 = get_stress(unit2) if happiness1 == happiness2 then return sort_by_name_desc(unit_id_1, unit_id_2) end if not happiness2 then return -1 end if not happiness1 then return 1 end return utils.compare(happiness2, happiness1) end local function sort_by_stress_asc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local happiness1 = get_stress(unit1) local happiness2 = get_stress(unit2) if happiness1 == happiness2 then return sort_by_name_desc(unit_id_1, unit_id_2) end if not happiness2 then return 1 end if not happiness1 then return -1 end return utils.compare(happiness1, happiness2) end local function get_skill(skill, unit) return unit and unit.status.current_soul and (utils.binsearch(unit.status.current_soul.skills, skill, 'id')) end local function get_skill_rating(skill, unit) local uskill = get_skill(skill, unit) if not uskill then return nil end return get_rating(uskill.rating, 0, 100, 10, 5, 1, 0) end local MELEE_WEAPON_SKILLS = { df.job_skill.AXE, df.job_skill.SWORD, df.job_skill.MACE, df.job_skill.HAMMER, df.job_skill.SPEAR, } local function melee_skill_effectiveness(unit) -- Physical attributes local strength = dfhack.units.getPhysicalAttrValue(unit, df.physical_attribute_type.STRENGTH) local agility = dfhack.units.getPhysicalAttrValue(unit, df.physical_attribute_type.AGILITY) local toughness = dfhack.units.getPhysicalAttrValue(unit, df.physical_attribute_type.TOUGHNESS) local endurance = dfhack.units.getPhysicalAttrValue(unit, df.physical_attribute_type.ENDURANCE) local body_size_base = unit.body.size_info.size_base -- Mental attributes local willpower = dfhack.units.getMentalAttrValue(unit, df.mental_attribute_type.WILLPOWER) local spatial_sense = dfhack.units.getMentalAttrValue(unit, df.mental_attribute_type.SPATIAL_SENSE) local kinesthetic_sense = dfhack.units.getMentalAttrValue(unit, df.mental_attribute_type.KINESTHETIC_SENSE) -- Skills -- Finding the highest skill local skill_rating = 0 for _, skill in ipairs(MELEE_WEAPON_SKILLS) do local melee_skill = dfhack.units.getNominalSkill(unit, skill, true) skill_rating = math.max(skill_rating, melee_skill) end local melee_combat_rating = dfhack.units.getNominalSkill(unit, df.job_skill.MELEE_COMBAT, true) local rating = skill_rating * 27000 + melee_combat_rating * 9000 + strength * 180 + body_size_base * 100 + kinesthetic_sense * 50 + endurance * 50 + agility * 30 + toughness * 20 + willpower * 20 + spatial_sense * 20 return rating end local function get_melee_skill_effectiveness_rating(unit) return get_rating(melee_skill_effectiveness(unit), 350000, 2750000, 64, 52, 40, 28) end local function make_sort_by_melee_skill_effectiveness_desc() return function(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = melee_skill_effectiveness(unit1) local rating2 = melee_skill_effectiveness(unit2) if rating1 == rating2 then return sort_by_name_desc(unit_id_1, unit_id_2) end return utils.compare(rating2, rating1) end end local function make_sort_by_melee_skill_effectiveness_asc() return function(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = melee_skill_effectiveness(unit1) local rating2 = melee_skill_effectiveness(unit2) if rating1 == rating2 then return sort_by_name_desc(unit_id_1, unit_id_2) end return utils.compare(rating1, rating2) end end local RANGED_WEAPON_SKILLS = { df.job_skill.CROSSBOW, } -- Function could easily be adapted to different weapon types. local function ranged_skill_effectiveness(unit) -- Physical attributes local agility = dfhack.units.getPhysicalAttrValue(unit, df.physical_attribute_type.AGILITY) -- Mental attributes local spatial_sense = dfhack.units.getMentalAttrValue(unit, df.mental_attribute_type.SPATIAL_SENSE) local kinesthetic_sense = dfhack.units.getMentalAttrValue(unit, df.mental_attribute_type.KINESTHETIC_SENSE) local focus = dfhack.units.getMentalAttrValue(unit, df.mental_attribute_type.FOCUS) -- Skills -- Finding the highest skill local skill_rating = 0 for _, skill in ipairs(RANGED_WEAPON_SKILLS) do local ranged_skill = dfhack.units.getNominalSkill(unit, skill, true) skill_rating = math.max(skill_rating, ranged_skill) end local ranged_combat = dfhack.units.getNominalSkill(unit, df.job_skill.RANGED_COMBAT, true) local rating = skill_rating * 24000 + ranged_combat * 8000 + agility * 15 + spatial_sense * 15 + kinesthetic_sense * 6 + focus * 6 return rating end local function get_ranged_skill_effectiveness_rating(unit) return get_rating(ranged_skill_effectiveness(unit), 0, 800000, 72, 52, 31, 11) end local function make_sort_by_ranged_skill_effectiveness_desc(list) return function(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = ranged_skill_effectiveness(unit1, list) local rating2 = ranged_skill_effectiveness(unit2, list) if rating1 == rating2 then return sort_by_name_desc(unit_id_1, unit_id_2) end return utils.compare(rating2, rating1) end end local function make_sort_by_ranged_skill_effectiveness_asc(list) return function(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = ranged_skill_effectiveness(unit1, list) local rating2 = ranged_skill_effectiveness(unit2, list) if rating1 == rating2 then return sort_by_name_desc(unit_id_1, unit_id_2) end return utils.compare(rating1, rating2) end end local function make_sort_by_skill_desc(sort_skill) return function(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end if unit_id_1 == -1 then return -1 end if unit_id_2 == -1 then return 1 end local s1 = get_skill(sort_skill, df.unit.find(unit_id_1)) local s2 = get_skill(sort_skill, df.unit.find(unit_id_2)) if s1 == s2 then return sort_by_name_desc(unit_id_1, unit_id_2) end if not s2 then return -1 end if not s1 then return 1 end if s1.rating ~= s2.rating then return utils.compare(s2.rating, s1.rating) end if s1.experience ~= s2.experience then return utils.compare(s2.experience, s1.experience) end return sort_by_name_desc(unit_id_1, unit_id_2) end end local function make_sort_by_skill_asc(sort_skill) return function(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end if unit_id_1 == -1 then return -1 end if unit_id_2 == -1 then return 1 end local s1 = get_skill(sort_skill, df.unit.find(unit_id_1)) local s2 = get_skill(sort_skill, df.unit.find(unit_id_2)) if s1 == s2 then return sort_by_name_desc(unit_id_1, unit_id_2) end if not s2 then return 1 end if not s1 then return -1 end if s1.rating ~= s2.rating then return utils.compare(s1.rating, s2.rating) end if s1.experience ~= s2.experience then return utils.compare(s1.experience, s2.experience) end return sort_by_name_desc(unit_id_1, unit_id_2) end end -- Statistical rating that is higher for dwarves that are mentally stable local function get_mental_stability(unit) local altruism = unit.status.current_soul.personality.traits.ALTRUISM local anxiety_propensity = unit.status.current_soul.personality.traits.ANXIETY_PROPENSITY local bravery = unit.status.current_soul.personality.traits.BRAVERY local cheer_propensity = unit.status.current_soul.personality.traits.CHEER_PROPENSITY local curious = unit.status.current_soul.personality.traits.CURIOUS local discord = unit.status.current_soul.personality.traits.DISCORD local dutifulness = unit.status.current_soul.personality.traits.DUTIFULNESS local emotionally_obsessive = unit.status.current_soul.personality.traits.EMOTIONALLY_OBSESSIVE local humor = unit.status.current_soul.personality.traits.HUMOR local love_propensity = unit.status.current_soul.personality.traits.LOVE_PROPENSITY local perseverence = unit.status.current_soul.personality.traits.PERSEVERENCE local politeness = unit.status.current_soul.personality.traits.POLITENESS local privacy = unit.status.current_soul.personality.traits.PRIVACY local stress_vulnerability = unit.status.current_soul.personality.traits.STRESS_VULNERABILITY local tolerant = unit.status.current_soul.personality.traits.TOLERANT local craftsmanship = setbelief.getUnitBelief(unit, df.value_type['CRAFTSMANSHIP']) local family = setbelief.getUnitBelief(unit, df.value_type['FAMILY']) local harmony = setbelief.getUnitBelief(unit, df.value_type['HARMONY']) local independence = setbelief.getUnitBelief(unit, df.value_type['INDEPENDENCE']) local knowledge = setbelief.getUnitBelief(unit, df.value_type['KNOWLEDGE']) local leisure_time = setbelief.getUnitBelief(unit, df.value_type['LEISURE_TIME']) local nature = setbelief.getUnitBelief(unit, df.value_type['NATURE']) local skill = setbelief.getUnitBelief(unit, df.value_type['SKILL']) -- calculate the rating using the defined variables local rating = (craftsmanship * -0.01) + (family * -0.09) + (harmony * 0.05) + (independence * 0.06) + (knowledge * -0.30) + (leisure_time * 0.24) + (nature * 0.27) + (skill * -0.21) + (altruism * 0.13) + (anxiety_propensity * -0.06) + (bravery * 0.06) + (cheer_propensity * 0.41) + (curious * -0.06) + (discord * 0.14) + (dutifulness * -0.03) + (emotionally_obsessive * -0.13) + (humor * -0.05) + (love_propensity * 0.15) + (perseverence * -0.07) + (politeness * -0.14) + (privacy * 0.03) + (stress_vulnerability * -0.20) + (tolerant * -0.11) return rating end local function sort_by_mental_stability_desc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = get_mental_stability(unit1) local rating2 = get_mental_stability(unit2) if rating1 == rating2 then -- sorting by stress is opposite -- more mental stable dwarves should have less stress return sort_by_stress_asc(unit_id_1, unit_id_2) end return utils.compare(rating2, rating1) end local function sort_by_mental_stability_asc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = get_mental_stability(unit1) local rating2 = get_mental_stability(unit2) if rating1 == rating2 then return sort_by_stress_desc(unit_id_1, unit_id_2) end return utils.compare(rating1, rating2) end -- Statistical rating that is higher for more potent dwarves in long run melee military training -- Rating considers fighting melee opponents -- Wounds are not considered! local function get_melee_combat_potential(unit) -- Physical attributes local strength = unit.body.physical_attrs.STRENGTH.max_value local agility = unit.body.physical_attrs.AGILITY.max_value local toughness = unit.body.physical_attrs.TOUGHNESS.max_value local endurance = unit.body.physical_attrs.ENDURANCE.max_value local body_size_base = unit.body.size_info.size_base -- Mental attributes local willpower = unit.status.current_soul.mental_attrs.WILLPOWER.max_value local spatial_sense = unit.status.current_soul.mental_attrs.SPATIAL_SENSE.max_value local kinesthetic_sense = unit.status.current_soul.mental_attrs.KINESTHETIC_SENSE.max_value -- melee combat potential rating local rating = strength * 264 + endurance * 84 + body_size_base * 77 + kinesthetic_sense * 74 + agility * 33 + willpower * 31 + spatial_sense * 27 + toughness * 25 return rating end local function get_melee_combat_potential_rating(unit) return get_rating(get_melee_combat_potential(unit), 300000, 2600000, 81, 64, 46, 29) end local function sort_by_melee_combat_potential_desc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = get_melee_combat_potential(unit1) local rating2 = get_melee_combat_potential(unit2) if rating1 == rating2 then return sort_by_mental_stability_desc(unit_id_1, unit_id_2) end return utils.compare(rating2, rating1) end local function sort_by_melee_combat_potential_asc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = get_melee_combat_potential(unit1) local rating2 = get_melee_combat_potential(unit2) if rating1 == rating2 then return sort_by_mental_stability_asc(unit_id_1, unit_id_2) end return utils.compare(rating1, rating2) end -- Statistical rating that is higher for more potent dwarves in long run ranged military training -- Wounds are not considered! local function get_ranged_combat_potential(unit) -- Physical attributes local agility = unit.body.physical_attrs.AGILITY.max_value -- Mental attributes local focus = unit.status.current_soul.mental_attrs.FOCUS.max_value local spatial_sense = unit.status.current_soul.mental_attrs.SPATIAL_SENSE.max_value local kinesthetic_sense = unit.status.current_soul.mental_attrs.KINESTHETIC_SENSE.max_value -- ranged combat potential formula local rating = agility * 5 + kinesthetic_sense * 2 + spatial_sense * 5 + focus * 2 return rating end local function get_ranged_combat_potential_rating(unit) return get_rating(get_ranged_combat_potential(unit), 0, 70000, 73, 57, 41, 25) end local function sort_by_ranged_combat_potential_desc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = get_ranged_combat_potential(unit1) local rating2 = get_ranged_combat_potential(unit2) if rating1 == rating2 then return sort_by_mental_stability_desc(unit_id_1, unit_id_2) end return utils.compare(rating2, rating1) end local function sort_by_ranged_combat_potential_asc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = get_ranged_combat_potential(unit1) local rating2 = get_ranged_combat_potential(unit2) if rating1 == rating2 then return sort_by_mental_stability_asc(unit_id_1, unit_id_2) end return utils.compare(rating1, rating2) end local function get_need(unit) if not unit or not unit.status.current_soul then return end for _, need in ipairs(unit.status.current_soul.personality.needs) do if need.id == df.need_type.MartialTraining and need.focus_level < 0 then return -need.focus_level end end end local function get_need_rating(unit) local focus_level = get_need(unit) if not focus_level then return end -- convert to stress ratings so we can use stress faces as labels if focus_level > 100000 then return 0 end if focus_level > 10000 then return 1 end if focus_level > 1000 then return 2 end if focus_level > 100 then return 3 end return 6 end local function sort_by_need_desc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = get_need(unit1) local rating2 = get_need(unit2) if rating1 == rating2 then return sort_by_stress_desc(unit_id_1, unit_id_2) end if not rating2 then return -1 end if not rating1 then return 1 end return utils.compare(rating2, rating1) end local function sort_by_need_asc(unit_id_1, unit_id_2) if unit_id_1 == unit_id_2 then return 0 end local unit1 = df.unit.find(unit_id_1) local unit2 = df.unit.find(unit_id_2) if not unit1 then return -1 end if not unit2 then return 1 end local rating1 = get_need(unit1) local rating2 = get_need(unit2) if rating1 == rating2 then return sort_by_stress_asc(unit_id_1, unit_id_2) end if not rating2 then return 1 end if not rating1 then return -1 end return utils.compare(rating1, rating2) end local sort_by_any_melee_desc=make_sort_by_melee_skill_effectiveness_desc() local sort_by_any_melee_asc=make_sort_by_melee_skill_effectiveness_asc() local sort_by_any_ranged_desc=make_sort_by_ranged_skill_effectiveness_desc() local sort_by_any_ranged_asc=make_sort_by_ranged_skill_effectiveness_asc() local sort_by_teacher_desc=make_sort_by_skill_desc(df.job_skill.TEACHING) local sort_by_teacher_asc=make_sort_by_skill_asc(df.job_skill.TEACHING) local sort_by_tactics_desc=make_sort_by_skill_desc(df.job_skill.MILITARY_TACTICS) local sort_by_tactics_asc=make_sort_by_skill_asc(df.job_skill.MILITARY_TACTICS) local sort_by_axe_desc=make_sort_by_skill_desc(df.job_skill.AXE) local sort_by_axe_asc=make_sort_by_skill_asc(df.job_skill.AXE) local sort_by_sword_desc=make_sort_by_skill_desc(df.job_skill.SWORD) local sort_by_sword_asc=make_sort_by_skill_asc(df.job_skill.SWORD) local sort_by_mace_desc=make_sort_by_skill_desc(df.job_skill.MACE) local sort_by_mace_asc=make_sort_by_skill_asc(df.job_skill.MACE) local sort_by_hammer_desc=make_sort_by_skill_desc(df.job_skill.HAMMER) local sort_by_hammer_asc=make_sort_by_skill_asc(df.job_skill.HAMMER) local sort_by_spear_desc=make_sort_by_skill_desc(df.job_skill.SPEAR) local sort_by_spear_asc=make_sort_by_skill_asc(df.job_skill.SPEAR) local sort_by_crossbow_desc=make_sort_by_skill_desc(df.job_skill.CROSSBOW) local sort_by_crossbow_asc=make_sort_by_skill_asc(df.job_skill.CROSSBOW) local SORT_LIBRARY = { {label='melee effectiveness', desc_fn=sort_by_any_melee_desc, asc_fn=sort_by_any_melee_asc, rating_fn=get_melee_skill_effectiveness_rating}, {label='ranged effectiveness', desc_fn=sort_by_any_ranged_desc, asc_fn=sort_by_any_ranged_asc, rating_fn=get_ranged_skill_effectiveness_rating}, {label='name', desc_fn=sort_by_name_desc, asc_fn=sort_by_name_asc}, {label='teacher skill', desc_fn=sort_by_teacher_desc, asc_fn=sort_by_teacher_asc, rating_fn=curry(get_skill_rating, df.job_skill.TEACHING)}, {label='tactics skill', desc_fn=sort_by_tactics_desc, asc_fn=sort_by_tactics_asc, rating_fn=curry(get_skill_rating, df.job_skill.MILITARY_TACTICS)}, {label='arrival order', desc_fn=sort_by_arrival_desc, asc_fn=sort_by_arrival_asc, rating_fn=get_arrival_rating}, {label='stress level', desc_fn=sort_by_stress_desc, asc_fn=sort_by_stress_asc, rating_fn=get_stress_rating, use_stress_faces=true}, {label='need for training', desc_fn=sort_by_need_desc, asc_fn=sort_by_need_asc, rating_fn=get_need_rating, use_stress_faces=true}, {label='axe skill', desc_fn=sort_by_axe_desc, asc_fn=sort_by_axe_asc, rating_fn=curry(get_skill_rating, df.job_skill.AXE)}, {label='sword skill', desc_fn=sort_by_sword_desc, asc_fn=sort_by_sword_asc, rating_fn=curry(get_skill_rating, df.job_skill.SWORD)}, {label='mace skill', desc_fn=sort_by_mace_desc, asc_fn=sort_by_mace_asc, rating_fn=curry(get_skill_rating, df.job_skill.MACE)}, {label='hammer skill', desc_fn=sort_by_hammer_desc, asc_fn=sort_by_hammer_asc, rating_fn=curry(get_skill_rating, df.job_skill.HAMMER)}, {label='spear skill', desc_fn=sort_by_spear_desc, asc_fn=sort_by_spear_asc, rating_fn=curry(get_skill_rating, df.job_skill.SPEAR)}, {label='crossbow skill', desc_fn=sort_by_crossbow_desc, asc_fn=sort_by_crossbow_asc, rating_fn=curry(get_skill_rating, df.job_skill.CROSSBOW)}, {label='melee potential', desc_fn=sort_by_melee_combat_potential_desc, asc_fn=sort_by_melee_combat_potential_asc, rating_fn=get_melee_combat_potential_rating}, {label='ranged potential', desc_fn=sort_by_ranged_combat_potential_desc, asc_fn=sort_by_ranged_combat_potential_asc, rating_fn=get_ranged_combat_potential_rating}, } local RATING_FNS = {} local STRESS_FACE_FNS = {} for _, opt in ipairs(SORT_LIBRARY) do RATING_FNS[opt.desc_fn] = opt.rating_fn RATING_FNS[opt.asc_fn] = opt.rating_fn if opt.use_stress_faces then STRESS_FACE_FNS[opt.desc_fn] = true STRESS_FACE_FNS[opt.asc_fn] = true end end -- ---------------------- -- SquadAssignmentOverlay -- SquadAssignmentOverlay = defclass(SquadAssignmentOverlay, overlay.OverlayWidget) SquadAssignmentOverlay.ATTRS{ default_pos={x=18, y=5}, default_enabled=true, viewscreens='dwarfmode/UnitSelector/SQUAD_FILL_POSITION', version='2', frame={w=38, h=31}, frame_style=gui.FRAME_PANEL, frame_background=gui.CLEAR_PEN, autoarrange_subviews=true, autoarrange_gap=1, } -- allow initial spacebar or two successive spacebars to fall through and -- pause/unpause the game local function search_on_char(ch, text) if ch == ' ' then return text:match('%S$') end return ch:match('[%l _-]') end function SquadAssignmentOverlay:init() self.dirty = true local sort_options = {} for _, opt in ipairs(SORT_LIBRARY) do table.insert(sort_options, { label=opt.label..CH_DN, value=opt.desc_fn, pen=COLOR_GREEN, }) table.insert(sort_options, { label=opt.label..CH_UP, value=opt.asc_fn, pen=COLOR_YELLOW, }) end self:addviews{ widgets.EditField{ view_id='search', frame={l=0}, label_text='Search: ', on_char=search_on_char, on_change=function() self:refresh_list() end, }, widgets.Panel{ frame={l=0, r=0, h=15}, frame_style=gui.FRAME_INTERIOR, subviews={ widgets.CycleHotkeyLabel{ view_id='sort', frame={t=0, l=0}, label='Sort by:', key='CUSTOM_SHIFT_S', options=sort_options, initial_option=sort_by_any_melee_desc, on_change=self:callback('refresh_list', 'sort'), }, widgets.CycleHotkeyLabel{ view_id='sort_any_melee', frame={t=2, l=0, w=11}, options={ {label='melee eff.', value=sort_noop}, {label='melee eff.'..CH_DN, value=sort_by_any_melee_desc, pen=COLOR_GREEN}, {label='melee eff.'..CH_UP, value=sort_by_any_melee_asc, pen=COLOR_YELLOW}, }, initial_option=sort_by_any_melee_desc, option_gap=0, on_change=self:callback('refresh_list', 'sort_any_melee'), }, widgets.CycleHotkeyLabel{ view_id='sort_any_ranged', frame={t=2, r=8, w=12}, options={ {label='ranged eff.', value=sort_noop}, {label='ranged eff.'..CH_DN, value=sort_by_any_ranged_desc, pen=COLOR_GREEN}, {label='ranged eff.'..CH_UP, value=sort_by_any_ranged_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_any_ranged'), }, widgets.CycleHotkeyLabel{ view_id='sort_name', frame={t=2, r=0, w=5}, options={ {label='name', value=sort_noop}, {label='name'..CH_DN, value=sort_by_name_desc, pen=COLOR_GREEN}, {label='name'..CH_UP, value=sort_by_name_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_name'), }, widgets.CycleHotkeyLabel{ view_id='sort_teacher', frame={t=4, l=0, w=8}, options={ {label='teacher', value=sort_noop}, {label='teacher'..CH_DN, value=sort_by_teacher_desc, pen=COLOR_GREEN}, {label='teacher'..CH_UP, value=sort_by_teacher_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_teacher'), }, widgets.CycleHotkeyLabel{ view_id='sort_tactics', frame={t=4, l=10, w=8}, options={ {label='tactics', value=sort_noop}, {label='tactics'..CH_DN, value=sort_by_tactics_desc, pen=COLOR_GREEN}, {label='tactics'..CH_UP, value=sort_by_tactics_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_tactics'), }, widgets.CycleHotkeyLabel{ view_id='sort_arrival', frame={t=4, r=0, w=14}, options={ {label='arrival order', value=sort_noop}, {label='arrival order'..CH_DN, value=sort_by_arrival_desc, pen=COLOR_GREEN}, {label='arrival order'..CH_UP, value=sort_by_arrival_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_arrival'), }, widgets.CycleHotkeyLabel{ view_id='sort_stress', frame={t=6, l=0, w=7}, options={ {label='stress', value=sort_noop}, {label='stress'..CH_DN, value=sort_by_stress_desc, pen=COLOR_GREEN}, {label='stress'..CH_UP, value=sort_by_stress_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_stress'), }, widgets.CycleHotkeyLabel{ view_id='sort_need', frame={t=6, r=0, w=18}, options={ {label='need for training', value=sort_noop}, {label='need for training'..CH_DN, value=sort_by_need_desc, pen=COLOR_GREEN}, {label='need for training'..CH_UP, value=sort_by_need_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_need'), }, widgets.CycleHotkeyLabel{ view_id='sort_axe', frame={t=8, l=0, w=4}, options={ {label='axe', value=sort_noop}, {label='axe'..CH_DN, value=sort_by_axe_desc, pen=COLOR_GREEN}, {label='axe'..CH_UP, value=sort_by_axe_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_axe'), }, widgets.CycleHotkeyLabel{ view_id='sort_sword', frame={t=8, w=6}, options={ {label='sword', value=sort_noop}, {label='sword'..CH_DN, value=sort_by_sword_desc, pen=COLOR_GREEN}, {label='sword'..CH_UP, value=sort_by_sword_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_sword'), }, widgets.CycleHotkeyLabel{ view_id='sort_mace', frame={t=8, r=0, w=5}, options={ {label='mace', value=sort_noop}, {label='mace'..CH_DN, value=sort_by_mace_desc, pen=COLOR_GREEN}, {label='mace'..CH_UP, value=sort_by_mace_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_mace'), }, widgets.CycleHotkeyLabel{ view_id='sort_hammer', frame={t=10, l=0, w=7}, options={ {label='hammer', value=sort_noop}, {label='hammer'..CH_DN, value=sort_by_hammer_desc, pen=COLOR_GREEN}, {label='hammer'..CH_UP, value=sort_by_hammer_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_hammer'), }, widgets.CycleHotkeyLabel{ view_id='sort_spear', frame={t=10, w=6}, options={ {label='spear', value=sort_noop}, {label='spear'..CH_DN, value=sort_by_spear_desc, pen=COLOR_GREEN}, {label='spear'..CH_UP, value=sort_by_spear_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_spear'), }, widgets.CycleHotkeyLabel{ view_id='sort_crossbow', frame={t=10, r=0, w=9}, options={ {label='crossbow', value=sort_noop}, {label='crossbow'..CH_DN, value=sort_by_crossbow_desc, pen=COLOR_GREEN}, {label='crossbow'..CH_UP, value=sort_by_crossbow_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_crossbow'), }, widgets.CycleHotkeyLabel{ view_id='sort_melee_combat_potential', frame={t=12, l=0, w=16}, options={ {label='melee potential', value=sort_noop}, {label='melee potential'..CH_DN, value=sort_by_melee_combat_potential_desc, pen=COLOR_GREEN}, {label='melee potential'..CH_UP, value=sort_by_melee_combat_potential_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_melee_combat_potential'), }, widgets.CycleHotkeyLabel{ view_id='sort_ranged_combat_potential', frame={t=12, r=0, w=17}, options={ {label='ranged potential', value=sort_noop}, {label='ranged potential'..CH_DN, value=sort_by_ranged_combat_potential_desc, pen=COLOR_GREEN}, {label='ranged potential'..CH_UP, value=sort_by_ranged_combat_potential_asc, pen=COLOR_YELLOW}, }, option_gap=0, on_change=self:callback('refresh_list', 'sort_ranged_combat_potential'), }, }, }, widgets.CycleHotkeyLabel{ view_id='military', frame={l=0}, key='CUSTOM_SHIFT_Q', label='Units in other squads:', options={ {label='Include', value='include', pen=COLOR_GREEN}, {label='Only', value='only', pen=COLOR_YELLOW}, {label='Exclude', value='exclude', pen=COLOR_RED}, }, initial_option='include', on_change=function() self:refresh_list() end, }, widgets.CycleHotkeyLabel{ view_id='officials', frame={l=0}, key='CUSTOM_SHIFT_O', label='Appointed officials:', options={ {label='Include', value='include', pen=COLOR_GREEN}, {label='Only', value='only', pen=COLOR_YELLOW}, {label='Exclude', value='exclude', pen=COLOR_RED}, }, initial_option='include', on_change=function() self:refresh_list() end, }, widgets.CycleHotkeyLabel{ view_id='nobles', frame={l=0, w=20}, key='CUSTOM_SHIFT_N', label='Nobility:', options={ {label='Include', value='include', pen=COLOR_GREEN}, {label='Only', value='only', pen=COLOR_YELLOW}, {label='Exclude', value='exclude', pen=COLOR_RED}, }, initial_option='include', on_change=function() self:refresh_list() end, }, widgets.CycleHotkeyLabel{ view_id='infant', frame={l=0}, key='CUSTOM_SHIFT_M', label='Mothers with infants:', options={ {label='Include', value='include', pen=COLOR_GREEN}, {label='Only', value='only', pen=COLOR_YELLOW}, {label='Exclude', value='exclude', pen=COLOR_RED}, }, initial_option='include', on_change=function() self:refresh_list() end, }, widgets.CycleHotkeyLabel{ view_id='unstable', frame={l=0}, key='CUSTOM_SHIFT_F', label='Weak mental fortitude:', options={ {label='Include', value='include', pen=COLOR_GREEN}, {label='Only', value='only', pen=COLOR_YELLOW}, {label='Exclude', value='exclude', pen=COLOR_RED}, }, initial_option='include', on_change=function() self:refresh_list() end, }, widgets.CycleHotkeyLabel{ view_id='maimed', frame={l=0}, key='CUSTOM_SHIFT_I', label='Critically injured:', options={ {label='Include', value='include', pen=COLOR_GREEN}, {label='Only', value='only', pen=COLOR_YELLOW}, {label='Exclude', value='exclude', pen=COLOR_RED}, }, initial_option='include', on_change=function() self:refresh_list() end, }, } end local function normalize_search_key(search_key) local out = '' for c in dfhack.toSearchNormalized(search_key):gmatch("[%w%s]") do out = out .. c:lower() end return out end local function is_in_military(unit) return unit.military.squad_id > -1 end local function is_elected_or_appointed_official(unit) if #unit.occupations > 0 then return true end for _, noble_pos in ipairs(dfhack.units.getNoblePositions(unit) or {}) do if noble_pos.position.flags.ELECTED or (noble_pos.position.mandate_max == 0 and noble_pos.position.demand_max == 0) then return true end end return false end local function is_nobility(unit) for _, noble_pos in ipairs(dfhack.units.getNoblePositions(unit) or {}) do if not noble_pos.position.flags.ELECTED and (noble_pos.position.mandate_max > 0 or noble_pos.position.demand_max > 0) then return true end end return false end local function has_infant(unit) for _, baby in ipairs(df.global.world.units.other.ANY_BABY2) do if baby.relationship_ids.Mother == unit.id then return true end end return false end local function is_unstable(unit) -- stddev percentiles are 61, 48, 35, 23 -- let's go with one stddev below the mean (35) as the cutoff local _, color = get_rating(get_mental_stability(unit), -40, 80, 35, 0, 0, 0) return color ~= COLOR_LIGHTGREEN end local function is_maimed(unit) return not unit.flags2.vision_good or unit.status2.limbs_grasp_count < 2 or unit.status2.limbs_stand_count == 0 end local function filter_matches(unit_id, filter) if unit_id == -1 then return true end local unit = df.unit.find(unit_id) if not unit then return false end if filter.military == 'only' and not is_in_military(unit) then return false end if filter.military == 'exclude' and is_in_military(unit) then return false end if filter.officials == 'only' and not is_elected_or_appointed_official(unit) then return false end if filter.officials == 'exclude' and is_elected_or_appointed_official(unit) then return false end if filter.nobles == 'only' and not is_nobility(unit) then return false end if filter.nobles == 'exclude' and is_nobility(unit) then return false end if filter.infant == 'only' and not has_infant(unit) then return false end if filter.infant == 'exclude' and has_infant(unit) then return false end if filter.unstable == 'only' and not is_unstable(unit) then return false end if filter.unstable == 'exclude' and is_unstable(unit) then return false end if filter.maimed == 'only' and not is_maimed(unit) then return false end if filter.maimed == 'exclude' and is_maimed(unit) then return false end if #filter.search == 0 then return true end local search_key = dfhack.TranslateName(dfhack.units.getVisibleName(unit)) return normalize_search_key(search_key):find(dfhack.toSearchNormalized(filter.search)) end local function is_noop_filter(filter) return #filter.search == 0 and filter.military == 'include' and filter.officials == 'include' and filter.nobles == 'include' and filter.infant == 'include' and filter.unstable == 'include' and filter.maimed == 'include' end local function is_filter_equal(a, b) return a.search == b.search and a.military == b.military and a.officials == b.officials and a.nobles == b.nobles and a.infant == b.infant and a.unstable == b.unstable and a.maimed == b.maimed end local unit_selector = df.global.game.main_interface.unit_selector -- this function uses the unused itemid and selected vectors to keep state, -- taking advantage of the fact that they are reset by DF when the list of units changes local function filter_vector(filter, prev_filter) local unid_is_filtered = #unit_selector.selected >= 0 and unit_selector.selected[0] ~= 0 if is_noop_filter(filter) or #unit_selector.selected == 0 then if not unid_is_filtered then -- we haven't modified the unid vector; nothing to do here return end -- restore the unid vector unit_selector.unid:assign(unit_selector.itemid) -- clear our "we meddled" flag unit_selector.selected[0] = 0 return end if unid_is_filtered and is_filter_equal(filter, prev_filter) then -- filter hasn't changed; we don't need to refilter return end if unid_is_filtered then -- restore the unid vector unit_selector.unid:assign(unit_selector.itemid) else -- save the unid vector and set our meddle flag unit_selector.itemid:assign(unit_selector.unid) unit_selector.selected[0] = 1 end -- do the actual filtering for idx=#unit_selector.unid-1,0,-1 do if not filter_matches(unit_selector.unid[idx], filter) then unit_selector.unid:erase(idx) end end -- fix up scroll position if it would be off the end of the list if unit_selector.scroll_position + 10 > #unit_selector.unid then unit_selector.scroll_position = math.max(0, #unit_selector.unid - 10) end end local use_stress_faces = false local rating_annotations = {} local function annotate_visible_units(sort_fn) use_stress_faces = STRESS_FACE_FNS[sort_fn] rating_annotations = {} rating_fn = RATING_FNS[sort_fn] local max_idx = math.min(#unit_selector.unid-1, unit_selector.scroll_position+9) for idx = unit_selector.scroll_position, max_idx do local annotation_idx = idx - unit_selector.scroll_position + 1 local unit = df.unit.find(unit_selector.unid[idx]) rating_annotations[annotation_idx] = nil if unit and rating_fn then local val, color = rating_fn(unit) if val then rating_annotations[annotation_idx] = {val=val, color=color} end end end end local SORT_WIDGET_NAMES = { 'sort', 'sort_any_melee', 'sort_any_ranged', 'sort_name', 'sort_teacher', 'sort_tactics', 'sort_arrival', 'sort_stress', 'sort_need', 'sort_axe', 'sort_sword', 'sort_mace', 'sort_hammer', 'sort_spear', 'sort_crossbow', 'sort_melee_combat_potential', 'sort_ranged_combat_potential', } function SquadAssignmentOverlay:refresh_list(sort_widget, sort_fn) sort_widget = sort_widget or 'sort' sort_fn = sort_fn or self.subviews.sort:getOptionValue() if sort_fn == sort_noop then self.subviews[sort_widget]:cycle() return end for _,widget_name in ipairs(SORT_WIDGET_NAMES) do self.subviews[widget_name]:setOption(sort_fn) end local filter = { search=self.subviews.search.text, military=self.subviews.military:getOptionValue(), officials=self.subviews.officials:getOptionValue(), nobles=self.subviews.nobles:getOptionValue(), infant=self.subviews.infant:getOptionValue(), unstable=self.subviews.unstable:getOptionValue(), maimed=self.subviews.maimed:getOptionValue(), } filter_vector(filter, self.prev_filter or {}) self.prev_filter = filter utils.sort_vector(unit_selector.unid, nil, sort_fn) annotate_visible_units(sort_fn) self.saved_scroll_position = unit_selector.scroll_position end function SquadAssignmentOverlay:onInput(keys) if keys._MOUSE_R_DOWN or keys._MOUSE_L_DOWN and not self:getMouseFramePos() then -- if any click is made outside of our window, we may need to refresh our list self.dirty = true end return SquadAssignmentOverlay.super.onInput(self, keys) end function SquadAssignmentOverlay:onRenderFrame(dc, frame_rect) SquadAssignmentOverlay.super.onRenderFrame(self, dc, frame_rect) if self.dirty then self:refresh_list() self.dirty = false elseif self.saved_scroll_position ~= unit_selector.scroll_position then annotate_visible_units(self.subviews.sort:getOptionValue()) self.saved_scroll_position = unit_selector.scroll_position end end -- ---------------------- -- SquadAnnotationOverlay -- SquadAnnotationOverlay = defclass(SquadAnnotationOverlay, overlay.OverlayWidget) SquadAnnotationOverlay.ATTRS{ default_pos={x=56, y=5}, default_enabled=true, viewscreens='dwarfmode/UnitSelector/SQUAD_FILL_POSITION', frame={w=5, h=35}, frame_style=gui.FRAME_INTERIOR_MEDIUM, frame_background=gui.CLEAR_PEN, } function get_annotation_text(idx) local elem = rating_annotations[idx] if not elem or not tonumber(elem.val) then return ' - ' end return tostring(math.tointeger(elem.val)) end function get_annotation_color(idx) local elem = rating_annotations[idx] return elem and elem.color or nil end local to_pen = dfhack.pen.parse local DASH_PEN = to_pen{ch='-', fg=COLOR_WHITE, keep_lower=true} local FACE_TILES = {} local function init_face_tiles() for idx=0,6 do FACE_TILES[idx] = {} local face_off = (6 - idx) * 2 for y=0,1 do for x=0,1 do local tile = dfhack.screen.findGraphicsTile('INTERFACE_BITS', 32 + face_off + x, 6 + y) ensure_key(FACE_TILES[idx], y)[x] = tile end end end for idx,color in ipairs{COLOR_RED, COLOR_LIGHTRED, COLOR_YELLOW, COLOR_WHITE, COLOR_GREEN, COLOR_LIGHTGREEN, COLOR_LIGHTCYAN} do local face = {} ensure_key(face, 0)[0] = to_pen{tile=FACE_TILES[idx-1][0][0], ch=1, fg=color} ensure_key(face, 0)[1] = to_pen{tile=FACE_TILES[idx-1][0][1], ch='\\', fg=color} ensure_key(face, 1)[0] = to_pen{tile=FACE_TILES[idx-1][1][0], ch='\\', fg=color} ensure_key(face, 1)[1] = to_pen{tile=FACE_TILES[idx-1][1][1], ch='/', fg=color} FACE_TILES[idx-1] = face end end function get_stress_face_tile(idx, x, y) local elem = rating_annotations[idx] if not elem or not elem.val or elem.val < 0 then return x == 0 and y == 1 and DASH_PEN or gui.CLEAR_PEN end local val = math.min(6, elem.val) return FACE_TILES[val][y][x] end function SquadAnnotationOverlay:init() for idx = 1, 10 do self:addviews{ widgets.Label{ frame={t=idx*3+1, h=1, w=3}, text={ { text=curry(get_annotation_text, idx), pen=curry(get_annotation_color, idx), width=3, rjustify=true, }, }, visible=function() return not use_stress_faces end, }, widgets.Label{ frame={t=idx*3, r=0, h=2, w=2}, auto_height=false, text={ {width=1, tile=curry(get_stress_face_tile, idx, 0, 0)}, {width=1, tile=curry(get_stress_face_tile, idx, 1, 0)}, NEWLINE, {width=1, tile=curry(get_stress_face_tile, idx, 0, 1)}, {width=1, tile=curry(get_stress_face_tile, idx, 1, 1)}, }, visible=function() return use_stress_faces end, }, } end end OVERLAY_WIDGETS = { squad_assignment=SquadAssignmentOverlay, squad_annotation=SquadAnnotationOverlay, } dfhack.onStateChange[GLOBAL_KEY] = function(sc) if sc ~= SC_MAP_LOADED or df.global.gamemode ~= df.game_mode.DWARF then return end init_face_tiles() end --[[ local utils = require('utils') local units = require('plugins.sort.units') local items = require('plugins.sort.items') orders = orders or {} orders.units = units.orders orders.items = items.orders function parse_ordering_spec(type,...) local group = orders[type] if group == nil then dfhack.printerr('Invalid ordering class: '..tostring(type)) return nil end local specs = table.pack(...) local rv = { } for _,spec in ipairs(specs) do local nil_first = false if string.sub(spec,1,1) == '<' then nil_first = true spec = string.sub(spec,2) end local reverse = false if string.sub(spec,1,1) == '>' then reverse = true spec = string.sub(spec,2) end local cm = group[spec] if cm == nil then dfhack.printerr('Unknown order for '..type..': '..tostring(spec)) return nil end if nil_first or reverse then cm = copyall(cm) cm.nil_first = nil_first cm.reverse = reverse end rv[#rv+1] = cm end return rv end make_sort_order = utils.make_sort_order ]] return _ENV