#### Load Libraries ####

library(tesseract)
library(magick)
library(lubridate)
library(tidyverse)
library(exifr)

# Search folder for all photos of domestic cats (files with "DOCA" in their name)

CAT <-
    read_exif(
        list.files(
            path = "F:\\WMP_MammalImages\\",
            full.names = T,
            recursive = T, # search subfiles as well
            pattern = "DOCA" # only include files with DOCA in their name
        ),
        tags = c("SourceFile", "Make") # only save the file path name
    )

# Extract relevant data from previous folder organization

CAT2 <- CAT %>%
    mutate(FILE = basename(CAT$SourceFile),
           WMP = basename(dirname(CAT$SourceFile)),
           YEAR = stringr::str_sub(CAT$SourceFile, start = 22, end = 25),
           NEW_FILE = as.character(NA)
    )

# Iterate through images 
## Copy images to new folder row by row (AKA DO NOT FUCK WITH ORIGINAL FILES)
### Rename images as they are copied


for (row in 1:nrow(CAT2)) {
  
  # Display iteration number
  
  print(row)
  
  ## Copy images to new folder row by row 
  
  file.copy(from = CAT2[row,1],
            to = "C:\\Users\\John\\Google Drive\\DOCA\\",
            overwrite = T)
 
  ### Create new name for copied images
  
  CAT2[row, 6] <- 
    stringr::str_c(CAT2[row, 5],
                   CAT2[row, 4],
                   row, # add the row number
                   "DOCA",
                   sep = "_")
  
  ### Rename images as they are copied

  file.rename(
    from = paste("C:\\Users\\John\\Google Drive\\DOCA\\",
                 CAT2[row, 3],
                 sep = ""),
    to = paste("C:\\Users\\John\\Google Drive\\DOCA\\",
               CAT2[row, 6],
               ".jpg", sep = "")
  )
  
    
}


# Check files to make sure they all have the time stamp within a similar location
# Separate images with different time stamp locations (i.e. different camera brands)
# For me, this was "Bushnell", "Cuddleback", and "Other"
# I had 315 Bushnell Images, 6 Cuddeback, and 6 Other
# For the rest of this script, I'm focusing on the Bushnell photos

# Now re-search your folder of interst

CAT3 <-
    read_exif(
        list.files(
            path = "C:\\Users\\John\\Google Drive\\DOCA\\Bushnell",
            full.names = T,
            recursive = T, 
            pattern = "DOCA"
        ),
        tags = c("SourceFile") # only save the file path name
    )


CAT4 <- CAT3 %>%
  mutate(FILE_NAME = basename(CAT3$SourceFile), #column 2
         WMP = NA, #column 3
         DATE = NA, #column 4
         YEAR = NA, #column 5
         MONTH = NA, #column 6
         DAY = NA, #column 7
         TIME = NA, #column 8
         HOUR = NA, #column 9
         MINUTE = NA, #column 10
         RENAMED_FILE = NA) #column 11

CAT4$WMP <- stringr::str_sub(CAT4$FILE_NAME, start = 6, end = 14)

# Create smaller dataset to practice on, if needed

CAT_TEST <- slice(CAT4, 1:3)

# Iterate through images

 for (row in 1:nrow(CAT4)) {
  
  # Display iteration
  
  print(row)
  
  # Load the first image to process
  
  CAT_IMAGE  <- magick::image_read(as.character(CAT4[row, 1]))
  print(as.character(CAT4[row, 1]))
  
  # Crop the first image
  # Notes on cropping: 
  
  CAT_IMAGE2  <- magick::image_crop(CAT_IMAGE, "774x96+2490+2352")

  # Extract Date-Time from cropped image using OCR
  
  DATE_TIME <- tesseract::ocr_data(CAT_IMAGE2)

  # Extract Date from Date-Time
  
  CAT4[row, 4] <- DATE_TIME[1, 1] # Date

  CAT4[row, 5] <- lubridate::year(lubridate::mdy(CAT4[row, 4])) # Year

  CAT4[row, 6] <- lubridate::month(lubridate::mdy(CAT4[row, 4])) # Month

  CAT4[row, 7] <- lubridate::day(lubridate::mdy(CAT4[row, 4])) # Day

  # Extract Time from Date-Time

  CAT4[row, 8] <- DATE_TIME[2, 1] # Time

  CAT4[row, 9] <- lubridate::hour(lubridate::hms(CAT4[row, 8])) # Hour

  CAT4[row, 10] <- lubridate::minute(lubridate::hms(CAT4[row, 8])) # Minute

  # Rename File using extracted Date-Time
  
  CAT4[row, 11] <- 
    stringr::str_c(as.character(CAT4[row, 5]), # YEAR
          as.character(CAT4[row, 3]), # WMP
          as.character(CAT4[row, 6]), # MONTH
          as.character(CAT4[row, 7]), # DAY
          as.character(CAT4[row, 9]), # HOUR
          as.character(CAT4[row, 10]), # MINUTE
          "DOCA.jpg",
          sep = "_")

    file.rename(
    from = as.character(CAT4[row, 1]), # SourceFile
    to = paste(
      "C:\\Users\\John\\Google Drive\\DOCA\\Bushnell\\",
      as.character(CAT4[row, 11]), # RENAMED_FILE
      sep = ""
    )
  )
  
  cat("image renamed","\n\n")


  
} 


write.table(CAT4, "clipboard-100000", row.names = F, sep = "\t")