# yearbook_processing_config.yaml
#
# Adapt this declaration in a personal/task copy before running any pipeline.
# No current template automatically loads this YAML or expands ${...} values.
# The caller maps inputs explicitly and routes all writes to output_root.

# --- Year range ---
yearbook_years:
  start: 2013       # First yearbook edition year (data year = yearbook_year - 1)
  end:   2026       # Last yearbook edition year

# --- Directory structure ---
paths:
  # Root directory containing yearbook folders named like: 中国工业统计年鉴{year}（Excel）
  yearbook_root: "${AIRMOSAIC_ASSETS_DIR}/data/01_socioeconomic_natural_change/IndustrialYearbook/<version>"

  output_root: "${AIRMOSAIC_ASSETS_DIR}/results/<project_id>/<run_id>/01_socioeconomic_natural_change/industrial_yearbook"
  # Output subdirectories relative to output_root, never to original inputs.
  output:
    economic_indicators: "economic_indicators"
    physical_output: "physical_output"

  # Sector code lookup table (GB/T 4754-2012)
  sector_code_lookup: "${AIRMOSAIC_ASSETS_DIR}/data/01_socioeconomic_natural_change/IndustrialYearbook/<version>/industry_code_lookup.csv"

# --- Region whitelist ---
regions:
  - "全国"
  - "北京"
  - "天津"
  - "河北"
  - "山西"
  - "内蒙古"
  - "辽宁"
  - "吉林"
  - "黑龙江"
  - "上海"
  - "江苏"
  - "浙江"
  - "安徽"
  - "福建"
  - "江西"
  - "山东"
  - "河南"
  - "湖北"
  - "湖南"
  - "广东"
  - "广西"
  - "海南"
  - "重庆"
  - "四川"
  - "贵州"
  - "云南"
  - "西藏"
  - "陕西"
  - "甘肃"
  - "青海"
  - "宁夏"
  - "新疆"

# Region name aliases (normalized → standard)
region_aliases:
  总计: "全国"
  合计: "全国"

# --- Table exclusion patterns ---
# Files whose names match any of these patterns are treated as
# non-industry grouping tables and excluded from processing.
exclude_patterns:
  - "国有控股"
  - "有限责任公司"
  - "股份有限公司"
  - "私营"
  - "港澳台商"
  - "外商投资"
  - "大型"
  - "中型"
  - "小型"

# --- Cleaning report ---
cleaning_report:
  enabled: true
  format: "markdown"
  suffix: "清洗说明.md"

# --- Missing data imputation ---
imputation:
  partial_missing_fill: "zero"         # Fill missing provinces as 0 when others exist
  full_missing_method: "interpolate"   # Interpolate national series, allocate by shares
  endpoint_extrapolation: "growth_rate" # Use 2-year growth rate for tails

# --- MEIC supplementation (product pipeline only) ---
meic:
  enabled: false
  # Paths to MEIC data files
  other_products: ""
  formal_energy_physical: ""
  formal_energy_standard_coal: ""
  sinter_ratio_file: ""
  cement_workbook: ""