parse_igotu2gpx_txt.R

# Primarily developed by Tom Evans: thomas.jude.evans AT gmail.com
# You are welcome to use parts of this code, but please give credit when using it extensively.

# This script is to extract data from RAW text files produced
# by igotu2gpx. Parsing the text and then output to file.

# Function for parsing a single GPS record
parse.record <- function(sub.text){
    # Get record number -----
    rec.num <- strsplit(sub.text[1], split = "Record " )[[1]][2]
    rec.num <- as.numeric(rec.num)
    
    # Get date_time   ------
    # Get index for line in sub.text with date_time
    i.date <- grep("Date", sub.text)
    # Just date line
    date.txt <- sub.text[i.date]
    # Date only
    date.only <- strsplit(date.txt, split = "  Date " )[[1]][2]
    # Split into date and time
    date.only.split <- strsplit(date.only, split = "T" )[[1]]
    # Time only part without "Z"
    time.only <- strsplit(date.only.split[2], split = "Z" )[[1]][1]
    # Put date and time back together into single character unit vector
    date.time.only <- paste(date.only.split[1],time.only)
    # Change to date_time object with UTC time-zone
    # A bit of a hack for occasional line without a date_time value
    if(date.only.split[1] == "Z"){date_time <- NA}else{
      date_time <- as.POSIXct(date.time.only, tz = "UTC")}
    
    # If you wanted character with full date_time including
    # fractional seconds
    # x <- strftime(date_time,'%Y-%m-%d %H:%M:%OS3')
    # str(x)
    
    
    # Get latitude ----
    # Get index for line in sub.text with latitude
    i.lat <- grep("Latitude", sub.text)
    # Just lat line
    lat.txt <- sub.text[i.lat]
    # lat only
    lat.only <- strsplit(lat.txt, split = "  Latitude " )[[1]][2]
    lat <- as.numeric(lat.only)
    
    
    # Get longitude ----
    # Get index for line in sub.text with longitude
    i.long <- grep("Longitude", sub.text)
    # Just long line
    long.txt <- sub.text[i.long]
    # long only
    long.only <- strsplit(long.txt, split = "  Longitude " )[[1]][2]
    long <- as.numeric(long.only)
    
    
    # Get elevation ----
    # Get index for line in sub.text with elevation
    i.elev <- grep("Elevation", sub.text)
    # Just elev line
    elev.txt <- sub.text[i.elev]
    # elev only
    elev.only <- strsplit(elev.txt, split = "  Elevation " )[[1]][2]
    elev.only <- strsplit(elev.only, split = " m" )[[1]][1]
    elev <- as.numeric(elev.only)
    
    
    
    # Get speed ----
    # Get index for line in sub.text with speedation
    i.speed <- grep("Speed", sub.text)
    # Just speed line
    speed.txt <- sub.text[i.speed]
    # speed only
    speed.only <- strsplit(speed.txt, split = "  Speed " )[[1]][2]
    speed.only <- strsplit(speed.only, split = " km/h" )[[1]][1]
    speed <- as.numeric(speed.only)
    
    
    # Get course ----
    # Get index for line in sub.text with course
    i.course <- grep("Course", sub.text)
    # Just course line
    course.txt <- sub.text[i.course]
    # course only
    course.only <- strsplit(course.txt, split = "  Course " )[[1]][2]
    course.only <- strsplit(course.only, split = " degrees" )[[1]][1]
    course <- as.numeric(course.only)
    
    
    # Get ehpe ----
    # Get index for line in sub.text with ehpe
    i.ehpe <- grep("EHPE", sub.text)
    # Just ehpe line
    ehpe.txt <- sub.text[i.ehpe]
    # ehpe only
    ehpe.only <- strsplit(ehpe.txt, split = "  EHPE " )[[1]][2]
    ehpe.only <- strsplit(ehpe.only, split = " m" )[[1]][1]
    ehpe <- as.numeric(ehpe.only)
    
    
    # Get Timeout -----
    # Possibly can get times of diving by looking at this, if timeout is
    # 12s with no fix appears to probably be diving.
    # Get index for line in sub.text with timeout
    i.timeout <- grep("Timeout", sub.text)
    # Just timeout line
    timeout.txt <- sub.text[i.timeout]
    # timeout only
    timeout.only <- strsplit(timeout.txt, split = "  Timeout " )[[1]][2]
    timeout.only <- strsplit(timeout.only, split = " s" )[[1]][1]
    timeout <- as.numeric(timeout.only)
    
    
    # Get MSVs_QCN ----
    # Get index for line in sub.text with ehpe
    i.MSVs_QCN <- grep("MSVs_QCN", sub.text)
    # Just MSVs_QCN line
    MSVs_QCN.txt <- sub.text[i.MSVs_QCN]
    # MSVs_QCN only
    MSVs_QCN.only <- strsplit(MSVs_QCN.txt, split = "  MSVs_QCN " )[[1]][2]
    MSVs_QCN <- as.numeric(MSVs_QCN.only)
    
    
    # Satellites:
    # Get index for line in sub.text with Satellites
    i.sat <- grep("Satellites", sub.text)
    # Just MSVs_QCN line
    sat.txt <- sub.text[i.sat]
    # sat only
    sat.only <- strsplit(sat.txt, split = "  Satellites:" )[[1]][2]
    
    if(is.na(sat.only)) {sat_n <- 0} else {
      sats <- unlist(strsplit(sat.only, "\\ "))  
      sats <- sats[-1]  
      sat_n <- length(sats)
    }
    
    # Convert speed to ms-1
    speed_ms <- speed*1000/3600
    
    # Put this all together -----
    data.c <- list(rec.num = rec.num, date_time = date_time,
                   lat = lat, long = long,
                   elev_m = elev,
                   speed_ms = speed_ms,
                   course = course,
                   ehpe = ehpe,
                   sat_n = sat_n,
                   timeout = timeout,
                   MSVs_QCN = MSVs_QCN
                   )
    
    # Output above
    return(data.c)
}


# Convert file function
parse.file <- function(file = NULL){
 
  # Read in the file
  text.samp <- readLines(file, encoding = "UTF-8")
    
  # Get an index for lines including 'Record'
  index <- grep("Record", text.samp)
  
  # View first 10 lines indexed
#   text.samp[index[1:10]]
  
  # Number of records
  n <- length(index)

  # analysed
  data.list <- list()

  # For 1: n-1 element
  for(i in 1:(n-1)){
    sub.text  <-  text.samp[index[i]:(index[i+1]-1)]
    data.list[[i]] <- parse.record(sub.text)
  }

  # For n element
  sub.text  <-  text.samp[index[n]:length(text.samp)]
  data.list[[n]] <- parse.record(sub.text)
  

  # Get to data frame -----
  df <- data.frame(matrix(unlist(data.list),
                          nrow = n, byrow = T))
  
  # Names
  names(df) <- c("rec.num", "date_time",
                 "lat", "long",
                 "elev",
                 "speed_ms",
                 "course",
                 "ehpe",
                 "sat_n",
                 "timeout",
                 "MSVs_QCN")
  
  # Get date_time back to date_time
  df$date_time <- as.POSIXct(df$date_time,
                             origin = "1970-01-01",
                             tz = "UTC")

  return(df)
}