copier.yml

_subdirectory: template
_min_copier_version: "9.0.0"
_envops:
  keep_trailing_newline: true
_templates_suffix: ""

# ---------------------------------------------------------------------------
# 1. What kind of thing is this?
# ---------------------------------------------------------------------------
project_type:
  type: str
  help: >-
    "team" = a real project that belongs on shared storage, with a raw-data
    landing zone under source_data/ and a working area under analysis_data/.
    "personal" = a one-off / exploratory thing that only needs a spot under
    your home/ directory. Nothing shared, nothing else is scaffolded.
  choices:
    Team project (source_data/ + analysis_data/): team
    Personal / exploratory (home/<you>/): personal
  default: team

# ---------------------------------------------------------------------------
# 2. Identity
# ---------------------------------------------------------------------------
project_name:
  type: str
  help: "Human-readable project name (e.g. 'Xenium remission cohort')"
  validator: >-
    {% if not project_name.strip() %}Project name cannot be empty{% endif %}

project_slug:
  type: str
  help: "Folder / repo name (lowercase, no spaces). Leave default to auto-derive."
  default: >-
    {{ project_name|lower|trim|replace(' ', '-')|replace('_', '-') }}
  validator: >-
    {% if not (project_slug | regex_search('^[a-z0-9][a-z0-9-]*[a-z0-9]$')) %}
    project_slug must be lowercase, alphanumeric and hyphens only, and not
    start or end with a hyphen (got: "{{ project_slug }}")
    {% endif %}

owner:
  type: str
  help: >-
    Your username as it appears under home/ (e.g. 'skrause'). Used as the
    contact/creator in the project manifest, and as the target folder for
    personal projects.

pi_or_group:
  type: str
  help: "Responsible PI / working group (for the manifest & README)"
  when: "{{ project_type == 'team' }}"

description:
  type: str
  help: "One or two sentences describing the project"
  default: "TBD"

data_sensitivity:
  type: str
  help: "Data sensitivity classification, for the manifest"
  choices:
    "Public / open data": public
    "Internal (no patient data)": internal
    "Patient-related / sensitive": sensitive
  default: internal
  when: "{{ project_type == 'team' }}"

uses_reference_data:
  type: str
  help: >-
    Does this project read from reference_data/ (genomes, databases, models)?
    Purely documentation — nothing is copied or linked, this just gets noted
    in the manifest so others know what the project depends on.
  default: ""
  when: "{{ project_type == 'team' }}"

data_type:
  type: str
  multiselect: true
  help: >-
    What kind of data does this project deal with? (pick any that apply)
    Each selection also becomes a modality folder: source_data/ gets a
    raw/processed skeleton per selection, analysis_data/ gets a
    01_preprocessing/02_quality_control skeleton per selection. Leave empty
    for a minimal project (just README + PROJECT.yml, nothing else).
  choices:
    "scRNA-seq / snRNA-seq": scrna-seq
    "Spatial transcriptomics (Visium / Xenium / MERFISH / CosMx)": spatial-transcriptomics
    "scATAC-seq / bulk ATAC-seq": atac-seq
    "Bulk RNA-seq": bulk-rna-seq
    "WGS / WES (genomic sequencing)": wgs-wes
    "Long-read sequencing (ONT / PacBio)": long-read-sequencing
    "Proteomics (mass spectrometry)": proteomics
    "Flow cytometry / CyTOF": flow-cytometry
    "DNA methylation (bisulfite / EM-seq)": methylation
    "Hi-C / chromatin conformation": hi-c
    "Whole-slide histology (H&E / IHC)": whole-slide-histology
    "Multiplex imaging (CODEX / PhenoCycler / Xenium / MERFISH imaging)": multiplex-imaging
    "Confocal / light-sheet microscopy": confocal-microscopy
    "MRI / CT / PET (DICOM)": radiology-dicom
    "Cryo-EM": cryo-em
    "Other / not listed": other
  default: []
  when: "{{ project_type == 'team' }}"

data_type_other:
  type: str
  help: "Briefly describe the 'other' data type(s)"
  default: ""
  when: "{{ project_type == 'team' and 'other' in data_type }}"

# ---------------------------------------------------------------------------
# 3. Version control & data versioning — both fully optional
# ---------------------------------------------------------------------------
init_git:
  type: bool
  help: "Initialize a git repository in the project's code/working folder?"
  default: true

init_dvc:
  type: bool
  help: >-
    Initialize DVC for data versioning? Works with or without git — if git
    is off, DVC is initialized with 'dvc init --no-scm'.
  default: false

dvc_remote_url:
  type: str
  help: >-
    DVC remote storage URL (e.g. an S3 bucket, or a path under
    reference_data/ or a dedicated dvc-store/ volume). Leave blank to
    configure this later with 'dvc remote add'.
  default: ""
  when: "{{ init_dvc }}"

# ---------------------------------------------------------------------------
# 4. Backup (Amanda)
# ---------------------------------------------------------------------------
backup:
  type: bool
  help: >-
    Should this directory be picked up by the group's Amanda backups?
    Recorded in PROJECT.yml and used by the group's exclude-list sync script
    to decide what Amanda skips. See the "Backup (Amanda)" section in the
    template repo's README for how this ties into the actual Amanda config.
  default: "{{ project_type == 'team' }}"

# ---------------------------------------------------------------------------
# Internal / computed — not asked, just derived
# ---------------------------------------------------------------------------
project_root:
  type: str
  when: false
  default: >-
    {% if project_type == 'team' %}analysis_data/{{ project_slug }}{% else %}home/{{ owner }}/{{ project_slug }}{% endif %}

_tasks:
  # --- 0. Sanity check: are we even at the root of the shared storage? ---
  # If NONE of the expected top-level folders exist here, this probably
  # isn't /data — warn, but don't block. Create the standard skeleton once
  # so the run (and any future ones) can proceed normally either way.
  - command: |
      found=0
      {% if project_type == 'team' %}
      for d in reference_data public_data home; do
      {% else %}
      for d in source_data analysis_data reference_data public_data; do
      {% endif %}
        [ -d "$d" ] && found=1
      done
      if [ "$found" -eq 0 ]; then
        RED='\033[1;31m'
        NC='\033[0m'
        printf "${RED}WARNING: none of source_data/, analysis_data/, reference_data/,${NC}\n"
        printf "${RED}public_data/, home/ exist in %s.${NC}\n" "$(pwd)"
        printf "${RED}This does NOT look like the root of the shared data storage.${NC}\n"
        printf "Continue anyway and create the project (and the standard skeleton) here? [y/N] "
        read ans
        case "$ans" in
          y|Y|yes|YES|Yes)
            mkdir -p source_data analysis_data public_data home
            mkdir -p reference_data/genomes reference_data/databases
            mkdir -p reference_data/models/classification \
                     reference_data/models/deconvolution \
                     reference_data/models/segmentation \
                     reference_data/models/llm
            ;;
          *)
            printf "${RED}Aborting - removing what was just created here.${NC}\n"
            {% if project_type == 'team' %}
            rm -rf "source_data/{{ project_slug }}" "analysis_data/{{ project_slug }}"
            rmdir source_data analysis_data 2>/dev/null || true
            {% else %}
            rm -rf "home/{{ owner }}/{{ project_slug }}"
            rmdir "home/{{ owner }}" home 2>/dev/null || true
            {% endif %}
            exit 1
            ;;
        esac
      fi
    working_directory: "."
  - command: |
      {% set extra_mod = [(data_type_other.strip() | lower | replace(' ', '-') | replace('_', '-'))] if (project_type == 'team' and 'other' in data_type and data_type_other.strip()) else [] %}
      {% set mods = ((data_type | reject('equalto', 'other') | list) + extra_mod) if project_type == 'team' else [] %}
      {% if mods %}
      set -e
      SLUG="{{ project_slug }}"
      echo "{{ mods | join(',') }}" | tr ',' '\n' | while IFS= read -r raw_m; do
        m=$(echo "$raw_m" | sed -e 's/^[[:space:]]*//' -e 's/[[:space:]]*$//')
        [ -z "$m" ] && continue

        SD="source_data/$SLUG/$m"
        mkdir -p "$SD/raw" "$SD/processed/scripts" "$SD/processed/logs" "$SD/processed/output"
        touch "$SD/metadata_table"
        cat > "$SD/README.md" <<EOF2
      # $m — source data

      Raw/incoming data for the "$m" modality of {{ project_name }}.

      - raw/            untouched incoming files, read-only once landed
      - processed/      scripts/, logs/, output/ from staging this modality's data
      - metadata_table  sample/metadata table for this modality
      EOF2

        AD="analysis_data/$SLUG/$m"
        mkdir -p "$AD/01_preprocessing/scripts" "$AD/01_preprocessing/notebooks" "$AD/01_preprocessing/output" "$AD/01_preprocessing/environment"
        touch "$AD/01_preprocessing/scripts/.gitkeep" "$AD/01_preprocessing/notebooks/.gitkeep" "$AD/01_preprocessing/output/.gitkeep" "$AD/01_preprocessing/environment/.gitkeep"
        cat > "$AD/01_preprocessing/README.md" <<EOF3
      # $m — 01_preprocessing

      Raw -> usable data for the "$m" modality.
      EOF3
        mkdir -p "$AD/02_quality_control"
        cat > "$AD/02_quality_control/README.md" <<EOF4
      # $m — 02_quality_control

      QC checks on the preprocessed "$m" data.
      EOF4
        cat > "$AD/README.md" <<EOF5
      # $m

      Analysis pipeline for the "$m" modality of {{ project_name }}.

      - 01_preprocessing/    raw -> usable data
      - 02_quality_control/  QC checks on the preprocessed data
      EOF5
      done

      mkdir -p "analysis_data/$SLUG/environment"
      echo "# add requirements.txt / environment.yml / pyproject.toml here" > "analysis_data/$SLUG/environment/README.md"
      {% else %}
      true
      {% endif %}
    working_directory: "."
  - command: sed -i "s/__CREATION_DATE__/$(date +%Y-%m-%d)/" PROJECT.yml 2>/dev/null || true
    working_directory: "{{ project_root }}"
  - command: >-
      {% if init_git %}git init -q{% else %}true{% endif %}
    working_directory: "{{ project_root }}"
  - command: >-
      {% if init_git %}git add -A && git commit -q -m "Bootstrap project from lab template" || true{% else %}true{% endif %}
    working_directory: "{{ project_root }}"
  - command: >-
      {% if init_dvc and init_git %}dvc init -q
      {% elif init_dvc and not init_git %}dvc init -q --no-scm
      {% else %}true{% endif %}
    working_directory: "{{ project_root }}"
  - command: >-
      {% if init_dvc and dvc_remote_url %}dvc remote add -d storage "{{ dvc_remote_url }}"{% else %}true{% endif %}
    working_directory: "{{ project_root }}"
  - command: >-
      {% if init_dvc and init_git %}git add -A && git commit -q -m "Initialize DVC" || true{% else %}true{% endif %}
    working_directory: "{{ project_root }}"
  - command: >-
      {% if backup %}touch .backup{% else %}true{% endif %}
    working_directory: "{{ project_root }}"

_message_after_copy: |
  Project scaffolded:
    {% if project_type == 'team' -%}
    source_data/{{ project_slug }}/    <- raw/incoming data lands here (immutable, not versioned)
    analysis_data/{{ project_slug }}/  <- your working project ({{ 'git' if init_git else 'no git' }}{{ ', dvc' if init_dvc else '' }})
    {%- else -%}
    home/{{ owner }}/{{ project_slug }}/   <- your personal working folder ({{ 'git' if init_git else 'no git' }}{{ ', dvc' if init_dvc else '' }})
    {%- endif %}

  Next steps:
    - Fill in the README / PROJECT.yml description if you left it as TBD.
    {%- if init_dvc and not dvc_remote_url %}
    - Set a DVC remote: cd {{ project_root }} && dvc remote add -d storage <url>
    {%- endif %}
    {%- if project_type == 'team' %}
    - Put raw/incoming files under source_data/{{ project_slug }}/ and treat them as read-only.
    {%- endif %}
    {%- if project_type == 'team' %}
    - Backup: {{ 'included' if backup else 'excluded' }} (recorded in PROJECT.yml). Run
      scripts/sync-amanda-excludes.sh at the /data root so Amanda's exclude
      list picks up this setting.
    {%- endif %}