Reading .DAT file with odd tab-delimited structure in r

I am trying to read in a .DAT file that is supposed to be tab delimited, however it is a bit messier than anticipated. The data should have five columns:

the first is a numeric value between 1 and infinity
the second is a date formatted as "MM/DD/YYYY"
the third contains time formatted as "HH:MM:SS" or "HH:MM"
the fourth contains free text
the fifth contains free text

c("12315\t01/01/1999\t22:31\tOther, specify - Test\tBILE TUCT", 
"UNKNOWN", 
"CONTRAINDICATION, STOP", 
"75\t01/28/2021\t19:34\tct\tBilateral (unable to cannulate), normal", 
"75\t07/01/2014\t15:01:02\tCT/MRI\tCT chest shows collapse...otherwise OKAY. ETT okay. ", 
"CT neg.  Left s w/1mm cyst.", 
"75\t06/13/2018\t14:30\tCardiac cath\tNormal A, EF 66%, no AS/MR"
)

I pulled a piece of the file from readLines.

Ideally, the output would be:

V1	V2	V3	V4	V5
12315	01/01/1999	22:31	Other, specify - Test	BILE TUCT, UNKNOWN, CONTRAINDICATION, STOP
75	01/28/2021	19:34	ct	Bilateral (unable to cannulate), normal
75	07/01/2014	15:01:02	CT/MRI	CT chest shows collapse...otherwise OKAY. ETT okay. CT neg. Left s w/1mm cyst.
75	06/13/2018	14:30	Cardiac cath	Normal A, EF 66%, no AS/MR

Solution

Assuming that tab is not used in free text fields and dat being the result of readLines() / readr::read_lines(), we could

create a frame from lines,
use grepl("\t", lines) |> cumsum() for grouping, to collect consecutive related lines (one with tabs optionally followed by those without tabs)
and collapse lines with desired separator, ", ";
from there we can use readr::read_tsv()

library(dplyr, warn.conflicts = FALSE)

dat |> 
  tibble(lines = _) |> 
  group_by(record = grepl("\t", lines) |> cumsum()) |> 
  #>   lines                                                                   record
  #>   <chr>                                                                    <int>
  #> 1 "12315\t01/01/1999\t22:31\tOther, specify - Test\tBILE TUCT"                 1
  #> 2 "UNKNOWN"                                                                    1
  #> 3 "CONTRAINDICATION, STOP"                                                     1
  #> 4 "75\t01/28/2021\t19:34\tct\tBilateral (unable to cannulate), normal"         2
  #> 5 "75\t07/01/2014\t15:01:02\tCT/MRI\tCT chest shows collapse...otherwise…      3
  #> 6 "CT neg.  Left s w/1mm cyst."                                                3
  #> 7 "75\t06/13/2018\t14:30\tCardiac cath\tNormal A, EF 66%, no AS/MR"            4
  summarise(lines = paste0(lines, collapse = ", ")) |> 
  #>   record lines                                                                  
  #>    <int> <chr>                                                                  
  #> 1      1 "12315\t01/01/1999\t22:31\tOther, specify - Test\tBILE TUCT, UNKNOWN, …
  #> 2      2 "75\t01/28/2021\t19:34\tct\tBilateral (unable to cannulate), normal"   
  #> 3      3 "75\t07/01/2014\t15:01:02\tCT/MRI\tCT chest shows collapse...otherwise…
  #> 4      4 "75\t06/13/2018\t14:30\tCardiac cath\tNormal A, EF 66%, no AS/MR"
  pull(lines) |> 
  I() |> 
  readr::read_tsv(col_names = FALSE) 

#> # A tibble: 4 × 5
#>      X1 X2         X3       X4                    X5                            
#>   <dbl> <chr>      <time>   <chr>                 <chr>                         
#> 1 12315 01/01/1999 22:31:00 Other, specify - Test BILE TUCT, UNKNOWN, CONTRAIND…
#> 2    75 01/28/2021 19:34:00 ct                    Bilateral (unable to cannulat…
#> 3    75 07/01/2014 15:01:02 CT/MRI                CT chest shows collapse...oth…
#> 4    75 06/13/2018 14:30:00 Cardiac cath          Normal A, EF 66%, no AS/MR

Example data:

dat <- c("12315\t01/01/1999\t22:31\tOther, specify - Test\tBILE TUCT", 
         "UNKNOWN", 
         "CONTRAINDICATION, STOP", 
         "75\t01/28/2021\t19:34\tct\tBilateral (unable to cannulate), normal", 
         "75\t07/01/2014\t15:01:02\tCT/MRI\tCT chest shows collapse...otherwise OKAY. ETT okay. ", 
         "CT neg.  Left s w/1mm cyst.", 
         "75\t06/13/2018\t14:30\tCardiac cath\tNormal A, EF 66%, no AS/MR"
)