# Todo: Shared Hair

Male and female teams need to complete shared hair, will need new preview thumbnails for male.

Easiest way I found to check for prior work is to have finished male/female hair open (glbs.html) side by side shared hair todo.html to see which hair mesh most closely resembles a todo shared trait:, like so:

- https://anata.dev/vrm/files/male/HAIR/glbs.html
- https://anata.dev/vrm/files/female/HAIR/glbs.html
- https://anata.dev/vrm/files/shared/HAIR/todo.html


forking female / male blend to get shared category started by filtering out the ones already done


# Female Shared Done
- Bun_Black-Red
- Mid_Green-Black
- Mid_Pink-Blue
- Mid_Gray-Red

## Retexturing

It's a start, going to export these glbs and upload blend - need help finishing this one

## Male Previews

Even though the male and female bases share the same metadata, the assets look drastically different. We need new previews featuring the male version of each trait.

We have all the previews and metadata in the github, and can get new previews from the repo.

1. exact pattern match the shared hair traits with all male metadata

First get the original filenames for each shared one

`awk -F',' 'NR==FNR{shared[$1];next} ($4 in shared)' shared_hair.csv male_hair.csv`

Example output:

Masculine,HAIR,Braided Black,Braided_Black
Masculine,HAIR,Braided Blonde,Braided_Blonde
Masculine,HAIR,Braided Blue,Braided_Blue
Masculine,HAIR,Braided Blue/Pink,Braided_Blue-Pink
Masculine,HAIR,Braided Brown,Braided_Brown
Masculine,HAIR,Braided Green,Braided_Green

2. now can get new reference images for each trait (62) by searching with original names vs all the JSON metadata

`while IFS= read -r folder; do jq -r --arg folder "$folder" '.attributes[] | select(.trait_type == "Hair" and .value == $folder) | .value + "," + input_filename' *.json; done < folders2.txt`

There's 871/1000 hits, which implies 87% of male hair are based on the shared traits. That is a lot to do.