"""500 fully clothed fictional identities, eleven age groups, ten expressions."""
from collections import Counter
import hashlib
import json
from pathlib import Path
import random
from vivarium_prompt_templates import CHARACTER_STYLE, CHARACTER_PROMPT_TEMPLATE

ROOT=Path(__file__).resolve().parents[1]
STATE=ROOT/'experiments/character-production'
AGES=[('baby',0,2),('preschool',3,5),('younger_schoolchild',6,8),('older_schoolchild',9,12),
      ('younger_teen',13,15),('older_teen',16,18),('young_adult',19,29),('middle_adult',30,49),
      ('older_adult',50,64),('younger_senior',65,79),('older_senior',80,95)]
EMOTIONS=[('neutral','a relaxed neutral expression with a slight friendly smile'),
 ('joyful','natural joy: an open smiling mouth, raised cheeks and bright happy eyes'),
 ('grateful','warm gratitude: a gentle closed-mouth appreciative smile and softly relaxed eyelids'),
 ('playful','gentle playfulness: a small lopsided mischievous grin and one eyebrow slightly raised'),
 ('angry','mild controlled anger: eyebrows drawn inward and lowered, lips firmly pressed together'),
 ('afraid','mild apprehension: raised inner eyebrows, slightly widened eyes and gently parted lips'),
 ('sad','quiet mild sadness: raised inner eyebrows, softly sad eyes and downturned mouth corners'),
 ('annoyed','mild annoyance: partly lowered eyelids, one brow lowered and a faint flat line of the lips'),
 ('pain','brief mild discomfort: a slight wince, gently narrowed eyes and compressed lips, with no visible injury'),
 ('disgust','mild distaste: a slight nose wrinkle, subtly raised upper lip and asymmetric downturned mouth')]
ETHNICITIES=[('white_caucasian','white European-ancestry', ['fair','light warm','light olive']),
 ('african_american','African-American',['medium brown','deep brown','dark brown']),
 ('arab','Arab',['light olive','warm tan','medium brown']),
 ('latino','Latino',['light tan','warm medium brown','olive']),
 ('indian','Indian',['warm brown','medium brown','deep brown']),
 ('japanese','Japanese',['light warm','medium light','light tan']),
 ('southeast_asian','Southeast Asian',['warm tan','medium brown','light golden brown']),
 ('indigenous_american','Indigenous American',['warm tan','medium brown','light brown'])]
HAIR_M=['neatly cropped straight hair','short naturally curly hair','soft wavy hair swept to one side',
        'close-cropped textured hair','short tousled hair','a tidy side-part hairstyle']
HAIR_F=['a short practical bob','shoulder-length naturally wavy hair','a neatly tied ponytail',
        'short naturally curly hair','a loose braid over one shoulder','a cropped textured hairstyle']
COLORS=['navy','burgundy','rust red','ochre','cream','plum','brown','charcoal','royal blue','soft lavender']

def seed(text):return int.from_bytes(hashlib.sha256(('vivarium-people-2026-10-06:'+text).encode()).digest()[:4],'big')

def write(path,value):
    path.parent.mkdir(parents=True,exist_ok=True);path.write_text(json.dumps(value,indent=2,ensure_ascii=False)+'\n')

def outfit(age,gender,rng,serial):
    color=rng.choice(COLORS)
    if age<=2:
        options=[('baby_everyday',f'a fully covering {color} cotton romper with long sleeves and soft booties'),
          ('baby_layers',f'a {color} knitted cardigan over a cream long-sleeved top, soft trousers and booties'),
          ('baby_sporty',f'a {color} long-sleeved sweatshirt, loose cotton trousers and soft socks')]
    elif age<=12:
        options=[('child_sporty',f'a {color} sweatshirt, loose navy joggers and closed sneakers'),
          ('child_everyday',f'a {color} cardigan over a cream T-shirt, comfortable brown trousers and closed shoes'),
          ('child_outdoors',f'a {color} zip jacket, full-length charcoal trousers and sturdy closed shoes'),
          ('child_school',f'a {color} cardigan over a white collared shirt, navy school trousers and closed shoes'),
          ('child_creative',f'a {color} long-sleeved shirt, brown cotton dungarees and closed canvas shoes'),
          ('child_formal',f'a {color} buttoned cardigan, cream collared shirt, full-length navy trousers and neat closed shoes')]
    else:
        options=[('sporty',f'a {color} sweatshirt, loose navy athletic trousers and closed sneakers'),
          ('casual_layers',f'a {color} cardigan over a cream crew-neck top, brown trousers and closed shoes'),
          ('outdoors',f'a {color} practical jacket, charcoal hiking trousers and closed walking shoes'),
          ('smart_casual',f'a {color} collared shirt under a cream cardigan, navy trousers and neat closed shoes'),
          ('formal',f'a {color} tailored jacket over a cream buttoned shirt, matching full-length trousers and closed dress shoes'),
          ('workwear',f'a {color} long-sleeved work shirt, practical brown carpenter trousers and closed work boots'),
          ('rock_inspired',f'a dark {color} denim jacket over a cream crew-neck T-shirt, loose dark trousers and closed boots'),
          ('crafts',f'a {color} cotton overshirt over a cream top, navy canvas trousers and closed shoes'),
          ('simple_everyday',f'a {color} knitted pullover, full-length brown trousers and plain closed shoes')]
        if gender=='female':
            options+=[('everyday_dress',f'a {color} long-sleeved knee-length day dress over opaque dark leggings with closed shoes'),
                      ('formal_skirt',f'a {color} buttoned jacket, a cream high-neck blouse, a knee-length skirt over opaque leggings and closed shoes')]
    kind,clothes=options[serial%len(options)]
    return kind,clothes

def description(person):
    age=person['age_years'];gender=person['gender'];eth=person['ethnicity_prompt_label'];skin=person['skin_tone']
    if person['age_group']=='baby':
        subject=f"one fictional {person['age_months']}-month-old baby {'boy' if gender=='male' else 'girl'}"
        anatomy='natural infant proportions, a softly rounded face and gentle proportionate eyes'
    elif age<13:
        subject=f"one fictional {age}-year-old {'boy' if gender=='male' else 'girl'}"
        anatomy='natural school-child proportions, modest almond-shaped eyes, a proportionate head and restrained facial features' if age>=6 else 'natural preschool-child proportions and restrained facial features'
    elif age<16:
        subject=f"one fictional {age}-year-old {'boy' if gender=='male' else 'girl'}, an early adolescent"
        anatomy='natural adolescent proportions, modest almond-shaped eyes and a proportionate head'
    elif age<=18:
        subject=f"one fictional mature-looking {age}-year-old {'young man' if gender=='male' else 'young woman'}, a late adolescent"
        anatomy='a near-adult elongated oval face, modest almond-shaped eyes, a subtly defined jaw, a proportionate head and long adolescent limbs'
    else:
        subject=f"one fictional {age}-year-old {'man' if gender=='male' else 'woman'}"
        anatomy='natural adult proportions and a proportionate face'
        if age>=50:anatomy+=', age-appropriate facial lines'
        if age>=65:anatomy+=', clearly visible natural wrinkles'
    return f"{subject}, {eth} heritage, {skin} skin, {person['hair']}, {person['build']}, wearing {person['outfit']}, fully clothed and age-appropriate. {anatomy}"

def prompt(person,emotion,reference=False):
    expression=dict(EMOTIONS)[emotion]
    text=CHARACTER_PROMPT_TEMPLATE.format(description=description(person),style=CHARACTER_STYLE)
    text=text.replace('Calm friendly expression, facing the viewer.',f'{expression.capitalize()}, facing the viewer.')
    if person['age_group']=='baby':
        text=text.replace('Standing three-quarter portrait, head to thighs, hands and shoulders inside the frame, margin around the silhouette.',
            'Sitting upright in a safe natural infant posture, entire clothed body inside the frame with a margin around the silhouette.')
    if reference:
        age=f"{person['age_months']}-month-old baby" if person['age_group']=='baby' else f"{person['age_years']}-year-old person"
        text=(f'vvar_char, edit the reference portrait. Make this exact same fictional {age} show {expression}. '
              'Change the mouth, eyebrows and eyelids so this mild natural emotion is clearly visible. '
              'Preserve exact identity, face shape, skin tone, hair, clothing, pose and framing. '
              f'{CHARACTER_STYLE} Fully clothed, age-appropriate. Perfectly uniform solid bright chroma green background. '
              'No scenery, no text, no logo, no watermark.')
    return text

def main():
    other_index=0;persons=[]
    # Eight cells get24people; the other14cells get22: total500.
    priority={'older_schoolchild','younger_teen','older_teen','young_adult'}
    for group,low,high in AGES:
        for gender in ['male','female']:
            count=24 if group in priority else 22
            for j in range(count):
                ident=f'person-{len(persons)+1:04d}';rng=random.Random(seed(ident))
                if j%2==0:eth=ETHNICITIES[0]
                else:eth=ETHNICITIES[1+other_index%7];other_index+=1
                age=rng.randint(max(low,1),high)
                months=rng.choice([9,12,15,18,21,24]) if group=='baby' else 0
                if group=='baby':age=months//12
                hair_color=rng.choice(['dark brown','black','medium brown','chestnut'])
                if eth[0]=='white_caucasian':hair_color=rng.choice(['dark brown','light brown','blond','auburn','black'])
                if age>=65:hair_color=rng.choice(['silver','white','gray'])
                elif age>=50:hair_color+=' with natural gray strands'
                style=rng.choice(HAIR_M if gender=='male' else HAIR_F)
                hair=f'{hair_color} {style}'
                if group=='baby':hair=f'soft sparse {hair_color} baby hair'
                body=rng.choice(['a slim build','an average build','a sturdy build','a compact build']) if age>=16 else 'a natural age-appropriate build'
                style_id,clothes=outfit(age,gender,rng,j)
                p=dict(personality_id=ident,character_id=ident,age_group=group,age_years=age,age_months=months,
                    gender=gender,ethnicity=eth[0],ethnicity_prompt_label=eth[1],skin_tone=rng.choice(eth[2]),hair=hair,
                    build=body,outfit_id='outfit-01',outfit_style=style_id,outfit=clothes,seed=seed(ident),default_emotion='neutral')
                p['description']=description(p);persons.append(p)
    assert len(persons)==500
    assert Counter(p['gender'] for p in persons)==dict(male=250,female=250)
    assert sum(p['ethnicity']=='white_caucasian' for p in persons)==250
    jobs=[]
    for p in persons:
        negative='extra people, changed clothing, different hairstyle, exaggerated expression, text, logo, watermark, background scenery'
        if p['age_years']>=6:negative+=', oversized doll eyes, chibi proportions'
        expressions=[dict(id=p['personality_id']+'-'+e,emotion=e,prompt=prompt(p,e,e!='neutral'),negative_prompt=negative if e=='neutral' else negative.replace('exaggerated expression, ',''),
                          seed=p['seed'],reference_id='' if e=='neutral' else p['personality_id']+'-neutral') for e,cue in EMOTIONS]
        jobs.append(dict(**p,expressions=expressions))
    assert len({j['id'] for p in jobs for j in p['expressions']})==5000
    write(STATE/'person-plan.json',jobs)
    write(STATE/'planning/allocation.json',dict(persons=500,images=5000,emotions_per_person=10,
        by_age_and_gender={f'{a}/{g}':sum(p['age_group']==a and p['gender']==g for p in persons) for a,l,h in AGES for g in ['male','female']},
        ethnicity_counts=dict(Counter(p['ethnicity'] for p in persons)),gender_counts=dict(Counter(p['gender'] for p in persons)),
        race_labels_are_prompt_attributes=True,ethnicity_not_inferred_from_appearance=True,
        default_emotion='neutral',reference_policy='Neutral image reused for9 expression edits; same seeded noise per identity',
        reference_width=416,reference_height=624,output_width=832,output_height=1248,steps=50,cfg=4.0))
    write(STATE/'planning/taxonomy.json',dict(age_groups=[dict(id=a,min_age=l,max_age=h) for a,l,h in AGES],
        emotions=[dict(id=e,expression=v) for e,v in EMOTIONS],ethnicity_groups=[dict(id=e,prompt_label=v) for e,v,s in ETHNICITIES],
        default_emotion='neutral',minor_policy='Fully clothed,age-appropriate everyday outfits; mild platonic expressions',
        baby_pose='Safely seated upright,full body visible; actual sampled infant ages9–24months',
        school_child_pose='Standing head-to-thigh portrait; modest natural facial proportions'))
    print(json.dumps(json.loads((STATE/'planning/allocation.json').read_text()),indent=2))

if __name__=='__main__':main()