counting number of lines in file using dcl script

[Jaimin Parikh]

Friends:

Can anybody help me in writing a script to count number of lines for a ascii file? I do not want to use open /read loop thing.

If anybody can help me, I would be highly oblige.

[Craig A Berry]

You can use various options on the search command to see how many lines are in a file:

$ search/statistics/nooutput sys$login:login.com ""

Files searched: 1 Buffered I/O count: 5
Records searched: 48 Direct I/O count: 2
Characters searched: 1162 Page faults: 26
Records matched: 48 Elapsed CPU time: 0 00:00:00.00
Lines printed: 0 Elapsed time: 0 00:00:00.02

but that doesn't store the result in a symbol -- is that something you need to do? If so, you probably do need to read in a loop, which isn't all that bad. Here's an example:

$ open/read a 'p1
$ count = 0
$ loop:
$ read/end=all_done a line
$ count = count + 1
$ goto loop
$ all_done:
$ close a
$ write sys$output "There are ",count," records in ", f$parse("''p1'")
$ exit

[Lokesh_2 · Accepted solution]

and how about this:

1. Below is my text file:

DCCDT2> typ a.txt
this is test file
this is test file
this is test file

2. Here is my command procedure:

DCCDT2> type d.com
$set noon
$pipe sear/stat a.txt "" | sear sys$pipe lines > c.dat
$open/read x c.dat
$read x y
$d :== 'F$extract(14,19,y)
$sho sym d
$close x
$exit

3. And here I execute my command procedure . It give me no. of lines in file :

DCCDT2> @d
D == "3"
DCCDT2>

[Hein van den Heuvel]

When you get that count, what are you going to do with it? If you use a tool like AWK or PERL then you can potentially immediatly use the number and add more value to it.

An other example of a tool telling you how many records there were:

$ app/log tmp.tmp nl:
%APPEND-S-APPENDED, U$1:[HEIN]TMP.TMP;4 appended to NL: (275 records)

Up to you to 'catch' that output and parse it.

A trivial perl example to set a logical name (NR) to the number of lines. It just does nothing reading the files and uses the perl built-in $. to report the count.


$ perl -E "while (<>){}; $ENV{""NR""}=$." < tmp.tmp
$ show log NR
"NR" = "275" (LNM$PROCESS_TABLE)

And an awk example:
$ gawk /comm="END {print NR}" tmp.tmp
275

What I could not figure out quickly enough, maybe a problem with the awk I have, is to return the numnber of lines encodeded in the exit status. Something like:

$ gawk /comm="END {exit 2*NR+1}" your-file

You could then maybe convert back to lines,
but it only returns 'abort' status (%X1000002C)

$ write sys$output "lines: ", ($STATUS - %X10000001)/2


hth,
Hein.

[John Gillings]

You can do it in one "line" using PIPE. There is no avoiding "open read loop", but you can avoid the loop being DCL. I'm using SEARCH/STATISTICS.

(beware wrapping...)

$ file="your-file-name"
$ PIPE SEARCH/STATISTICS/NOOUTPUT 'file' " " | -
SEARCH SYS$PIPE "records searched" | -
(READ SYS$PIPE line ; -
lines=F$ELEMENT(2," ",F$EDIT(line,"COMPRESS")) ; -
DEFINE/JOB lines &lines)

On completion, the JOB logical name "LINES" will contain the line count of your file. If the file doesn't exist, the count will be 0.

You'd be much better off writing a trivial program, 2 parameters, file name and symbol name:

$ countlines 'file' lines
$ WRITE SYS$OUTPUT "line count=''lines'"

[Hein van den Heuvel]

and a little more as a one-liner also with pipes:

$pipe sea/log file xxx |( read sys$pipe l ; l=f$ele(3," ",l) ; defi/job l &l
)


Hein.