# ==============================================================================
# SmartCane Script 10: Create Per-Field TIFFs
# ==============================================================================
#
# PURPOSE:
#   Split full-farm satellite TIFFs into per-field file structure across TWO phases:
#
#   PHASE 1 - MIGRATION (Legacy Data):
#     Input:  merged_final_tif/{DATE}.tif (5-band: R,G,B,NIR,CI - with CI calculated)
#     Output: field_tiles_CI/{FIELD}/{DATE}.tif
#     Status: One-time reorganization of existing data; will be removed after 2-3 weeks
#
#   PHASE 2 - PROCESSING (New Downloads):
#     Input:  merged_tif/{DATE}.tif (4-band: R,G,B,NIR - raw from Planet API)
#     Output: field_tiles/{FIELD}/{DATE}.tif
#     Status: Ongoing for all new downloads; always runs (not conditional)
#
#   INTEGRATION WITH DOWNSTREAM SCRIPTS:
#   - Script 20 (CI Extraction):
#     Reads from field_tiles/{FIELD}/{DATE}.tif
#     Adds CI calculation → outputs to field_tiles_CI/{FIELD}/{DATE}.tif (5-band)
#   - Script 40 (Mosaic Creation):
#     Reads from field_tiles_CI/{FIELD}/{DATE}.tif (via per-field weekly aggregation)
#     Creates weekly_mosaic/{FIELD}/week_{WW}.tif
#
#   ARCHITECTURE:
#   This script uses field/date folder organization:
#     field_tiles/
#       ├── field_1/
#       │   ├── 2024-01-15.tif
#       │   └── 2024-01-16.tif
#       └── field_2/
#           ├── 2024-01-15.tif
#           └── 2024-01-16.tif
#
#   Benefits: Upstream scripts iterate per-field → per-date, enabling clean
#   aggregation for mosaics (Script 40) and KPIs (Script 80/90).
#
# ==============================================================================


library(terra)
library(sf)

# ============================================================================
# HELPER FUNCTIONS (DEFINE FIRST)
# ============================================================================

smartcane_log <- function(msg) {
  cat(paste0("[", Sys.time(), "] ", msg, "\n"))
}

# Load field boundaries from GeoJSON
load_field_boundaries <- function(geojson_path) {
  smartcane_log(paste("Loading field boundaries from:", geojson_path))
  
  if (!file.exists(geojson_path)) {
    stop("GeoJSON file not found:", geojson_path)
  }
  
  fields <- st_read(geojson_path, quiet = TRUE)
  
  # Standardize field name property
  if (!"field_name" %in% names(fields)) {
    if ("field" %in% names(fields)) {
      fields$field_name <- fields$field
    } else if ("FIELD_ID" %in% names(fields)) {
      fields$field_name <- fields$FIELD_ID
    } else if ("Name" %in% names(fields)) {
      fields$field_name <- fields$Name
    } else {
      # Default: use first non-geometry column
      field_col <- names(fields)[!names(fields) %in% c("geometry", "geom")]
      if (length(field_col) > 0) {
        fields$field_name <- fields[[field_col[1]]]
      } else {
        stop("No suitable field name column found in GeoJSON")
      }
    }
  }
  
  smartcane_log(paste("Loaded", nrow(fields), "field(s)"))
  return(fields)
}

# ============================================================================
# PROJECT SETUP
# ============================================================================

# Get project parameter
args <- commandArgs(trailingOnly = TRUE)
if (length(args) == 0) {
  PROJECT <- "angata"
} else {
  PROJECT <- args[1]
}

# Construct paths directly (avoid complex parameter initialization)
base_path <- file.path(getwd(), "laravel_app", "storage", "app", PROJECT)
data_dir <- file.path(base_path, "Data")

smartcane_log(paste("Project:", PROJECT))
smartcane_log(paste("Base path:", base_path))
smartcane_log(paste("Data dir:", data_dir))

# Unified function to crop TIFF to field boundaries
# Called by both migration and processing phases
crop_tiff_to_fields <- function(tif_path, tif_date, fields, output_base_dir) {
  
  created <- 0
  skipped <- 0
  errors <- 0
  
  # Load raster
  if (!file.exists(tif_path)) {
    smartcane_log(paste("ERROR: TIFF not found:", tif_path))
    return(list(created = 0, skipped = 0, errors = 1))
  }
  
  rast <- tryCatch({
    rast(tif_path)
  }, error = function(e) {
    smartcane_log(paste("ERROR loading raster:", e$message))
    return(NULL)
  })
  
  if (is.null(rast)) {
    return(list(created = 0, skipped = 0, errors = 1))
  }
  
  # Create raster bounding box in raster CRS
  rast_bbox <- st_as_sfc(st_bbox(rast))
  st_crs(rast_bbox) <- st_crs(rast)
  
  # Reproject fields to match raster CRS
  fields_reprojected <- st_transform(fields, st_crs(rast_bbox))
  
  # Find which fields intersect this raster (CRITICAL: raster bbox first, then fields)
  overlapping_indices <- st_intersects(rast_bbox, fields_reprojected, sparse = TRUE)
  overlapping_indices <- unique(unlist(overlapping_indices))
  
  if (length(overlapping_indices) == 0) {
    smartcane_log(paste("No fields intersect TIFF:", basename(tif_path)))
    return(list(created = 0, skipped = 0, errors = 0))
  }
  
  # Process each overlapping field
  for (field_idx in overlapping_indices) {
    field_name <- fields$field_name[field_idx]
    field_geom <- fields_reprojected[field_idx, ]
    
    # Create field directory
    field_dir <- file.path(output_base_dir, field_name)
    if (!dir.exists(field_dir)) {
      dir.create(field_dir, recursive = TRUE, showWarnings = FALSE)
    }
    
    # Output file path
    output_path <- file.path(field_dir, paste0(tif_date, ".tif"))
    
    # Check if file already exists (idempotent)
    if (file.exists(output_path)) {
      skipped <- skipped + 1
      next
    }
    
    # Crop raster to field boundary
    tryCatch({
      field_rast <- crop(rast, field_geom)
      writeRaster(field_rast, output_path, overwrite = TRUE)
      created <- created + 1
    }, error = function(e) {
      smartcane_log(paste("ERROR cropping field", field_name, ":", e$message))
      errors <<- errors + 1
    })
  }
  
  return(list(created = created, skipped = skipped, errors = errors))
}

# Migrate legacy 5-band TIFFs with CI from merged_final_tif
migrate_old_merged_final_tif <- function(merged_final_dir, field_tiles_ci_dir, fields) {
  
  smartcane_log("\n========================================")
  smartcane_log("PHASE 1: MIGRATING LEGACY DATA")
  smartcane_log("========================================")
  
  # Check if legacy directory exists
  if (!dir.exists(merged_final_dir)) {
    smartcane_log("No legacy merged_final_tif/ directory found - skipping migration")
    return(list(total_created = 0, total_skipped = 0, total_errors = 0))
  }
  
  # Create output directory
  if (!dir.exists(field_tiles_ci_dir)) {
    dir.create(field_tiles_ci_dir, recursive = TRUE, showWarnings = FALSE)
  }
  
  # Find all date-pattern TIFFs in root of merged_final_tif
  tiff_files <- list.files(
    merged_final_dir,
    pattern = "^[0-9]{4}-[0-9]{2}-[0-9]{2}\\.tif$",
    full.names = TRUE
  )
  
  smartcane_log(paste("Found", length(tiff_files), "legacy TIFF(s) to migrate"))
  
  if (length(tiff_files) == 0) {
    smartcane_log("No legacy TIFFs found - skipping migration")
    return(list(total_created = 0, total_skipped = 0, total_errors = 0))
  }
  
  # Process each legacy TIFF
  total_created <- 0
  total_skipped <- 0
  total_errors <- 0
  
  for (tif_path in tiff_files) {
    tif_date <- gsub("\\.tif$", "", basename(tif_path))
    
    smartcane_log(paste("Migrating:", tif_date))
    
    result <- crop_tiff_to_fields(tif_path, tif_date, fields, field_tiles_ci_dir)
    total_created <- total_created + result$created
    total_skipped <- total_skipped + result$skipped
    total_errors <- total_errors + result$errors
  }
  
  smartcane_log(paste("Migration complete: created =", total_created,
                     ", skipped =", total_skipped, ", errors =", total_errors))
  
  return(list(total_created = total_created, total_skipped = total_skipped,
              total_errors = total_errors))
}

# Process new 4-band raw TIFFs from merged_tif
process_new_merged_tif <- function(merged_tif_dir, field_tiles_dir, fields) {
  
  smartcane_log("\n========================================")
  smartcane_log("PHASE 2: PROCESSING NEW DOWNLOADS")
  smartcane_log("========================================")
  
  # Check if download directory exists
  if (!dir.exists(merged_tif_dir)) {
    smartcane_log("No merged_tif/ directory found - no new data to process")
    return(list(total_created = 0, total_skipped = 0, total_errors = 0))
  }
  
  # Create output directory
  if (!dir.exists(field_tiles_dir)) {
    dir.create(field_tiles_dir, recursive = TRUE, showWarnings = FALSE)
  }
  
  # Find all date-pattern TIFFs in root of merged_tif
  tiff_files <- list.files(
    merged_tif_dir,
    pattern = "^[0-9]{4}-[0-9]{2}-[0-9]{2}\\.tif$",
    full.names = TRUE
  )
  
  smartcane_log(paste("Found", length(tiff_files), "TIFF(s) to process"))
  
  if (length(tiff_files) == 0) {
    smartcane_log("No new TIFFs found - nothing to process")
    return(list(total_created = 0, total_skipped = 0, total_errors = 0))
  }
  
  # Process each new TIFF
  total_created <- 0
  total_skipped <- 0
  total_errors <- 0
  
  for (tif_path in tiff_files) {
    tif_date <- gsub("\\.tif$", "", basename(tif_path))
    
    smartcane_log(paste("Processing:", tif_date))
    
    result <- crop_tiff_to_fields(tif_path, tif_date, fields, field_tiles_dir)
    total_created <- total_created + result$created
    total_skipped <- total_skipped + result$skipped
    total_errors <- total_errors + result$errors
  }
  
  smartcane_log(paste("Processing complete: created =", total_created,
                     ", skipped =", total_skipped, ", errors =", total_errors))
  
  return(list(total_created = total_created, total_skipped = total_skipped,
              total_errors = total_errors))
}

# ============================================================================
# MAIN EXECUTION
# ============================================================================

smartcane_log("========================================")
smartcane_log(paste("Script 10: Per-Field TIFF Creation for", PROJECT))
smartcane_log("========================================")

# Create necessary directories
dir.create(data_dir, recursive = TRUE, showWarnings = FALSE)

# Load field boundaries
geojson_path <- file.path(data_dir, "pivot.geojson")
fields <- load_field_boundaries(geojson_path)

# Define input and output directories
merged_final_dir <- file.path(base_path, "merged_final_tif")
merged_tif_dir <- file.path(base_path, "merged_tif")
field_tiles_dir <- file.path(base_path, "field_tiles")
field_tiles_ci_dir <- file.path(base_path, "field_tiles_CI")

# PHASE 1: Migrate legacy data (if exists)
migrate_result <- migrate_old_merged_final_tif(merged_final_dir, field_tiles_ci_dir, fields)

# PHASE 2: Process new downloads (always runs)
process_result <- process_new_merged_tif(merged_tif_dir, field_tiles_dir, fields)

smartcane_log("\n========================================")
smartcane_log("FINAL SUMMARY")
smartcane_log("========================================")
smartcane_log(paste("Migration: created =", migrate_result$total_created,
                   ", skipped =", migrate_result$total_skipped,
                   ", errors =", migrate_result$total_errors))
smartcane_log(paste("Processing: created =", process_result$total_created,
                   ", skipped =", process_result$total_skipped,
                   ", errors =", process_result$total_errors))
smartcane_log("Script 10 complete")
smartcane_log("========================================\n")