Ndtyjky

After a long morning, I have given up!

I have the following text file: StationLog.txt

Containing the following:

Version      = 2.0

StationName  = STN67_P70

BeginTime    = 2017-10-06.03:25:00

EndTime      = 2017-10-06.03:55:00

IgnoreNo     = 5000

PumpedVolume = 0

I need to extract the BeginTime, EndTime and StationName, those headings are fine, as values which feeds into another bit of code.

The idea is I don't have to do it manually as there will be lots of these files in time.

Following various other guides I have to this far:

a <- read.fwf("StationLog.txt", c(37,100), stringsAsFactors=FALSE)

a <- a[grep("=", a$V1), ]

a <- cbind(

  do.call( rbind, strsplit(a$V1, "=\s+") )

But hit a bit of a wall, any help would be greatly appreciated!

Based on the read.table function, you can use the arguments to do what you want.

The following proposal will work if you have only one = per row, and if the BeginTime, EndTime and StationName have the same writing in all files:

read.table(

  file            ="StationLog.txt",

  header          =FALSE, # No column names

  sep             ="=",   # separator character

  strip.white     =TRUE,  # remove multiple white character

  row.names       =1,     # the first column contains the rownames

  stringsAsFactors=FALSE

)[c("BeginTime", "EndTime", "StationName"), # extract the 3 infos based on their names corresponding to the rownames

  ,drop=FALSE] # keep the data.frame format

And the result:

                             V2

BeginTime   2017-10-06.03:25:00

EndTime     2017-10-06.03:55:00

StationName           STN67_P70

If you read the whole thing as a multiline string:

data:

txt_string <- "Version      = 2.0

StationName  = STN67_P70

BeginTime    = 2017-10-06.03:25:00

EndTime      = 2017-10-06.03:55:00

IgnoreNo     = 5000

PumpedVolume = 0"



stationName<- regmatches(txt_string, gregexpr("StationName\s+=\s+\K\S+" ,txt_string, perl = T))

beginTime  <- regmatches(txt_string, gregexpr("BeginTime\s+=\s+\K\S+" ,txt_string, perl = T))

endTime    <- regmatches(txt_string, gregexpr("EndTime\s+=\s+\K\S+" ,txt_string, perl = T))



do.call(cbind, c(stationName, beginTime, endTime))



#     [,1]        [,2]                  [,3]                 

#[1,] "STN67_P70" "2017-10-06.03:25:00" "2017-10-06.03:55:00"

Alternate method:

# use the filename vs this embedded example

station_info <-   readLines(textConnection("Version      = 2.0

StationName  = STN67_P70

BeginTime    = 2017-10-06.03:25:00

EndTime      = 2017-10-06.03:55:00

IgnoreNo     = 5000

PumpedVolume = 0"))

base:

as.list(sapply(

  strsplit(station_info, split = "[[:space:]]*=[[:space:]]*"),

  function(x) setNames(x[2], x[1])

)) -> station_info



str(station_info, 1)

## List of 6

##  $ Version     : chr "2.0"

##  $ StationName : chr "STN67_P70"

##  $ BeginTime   : chr "2017-10-06.03:25:00"

##  $ EndTime     : chr "2017-10-06.03:55:00"

##  $ IgnoreNo    : chr "5000"

##  $ PumpedVolume: chr "0"

tidyverse:

library(tidyverse)



# use the filename vs this embedded example

station_info <-   readLines(textConnection("Version      = 2.0

StationName  = STN67_P70

BeginTime    = 2017-10-06.03:25:00

EndTime      = 2017-10-06.03:55:00

IgnoreNo     = 5000

PumpedVolume = 0"))



str_split(station_info, pattern = "[[:space:]]*=[[:space:]]*") %>% 

  map(~set_names(.x[2], .x[1])) %>% 

  flatten() %>% 

  str(1)

Wrapping the base version (to avoid dependencies) into a function so you can re-use it for other stations:

read_station_metadata <- function(path) {



  path <- path.expand(path)

  stopifnot(file.exists(path))



  station_info <- readLines(path, warn = FALSE)



  as.list(sapply(

    strsplit(station_info, split = "[[:space:]]*=[[:space:]]*"),

    function(x) setNames(x[2], x[1])

  ))



}