Count lines fast

Question

I'm running Windows 7 and have GnuWin32. I have a several-gigabyte text file with LF (\n) line endings. I want to know how many lines it has (or alternatively how many LFs it has, which is, let's say, one fewer, but I don't care). What's the fastest or least expensive way to get my answer?

Some possibilities (though I'm sure I'm missing some):

• wc -l foo
• grep -c $ foo (with -c, prints only the count of matching lines)
• grep -c ^^ foo (the first caret escapes the second)
• sed -n $= foo (-n prevents printing the line; $ restricts to the last line; = prints the line number)

(Those are the GnuWin32 utilities. I don't know of any native-to-Windows way.)

Answer 1

Windows command line solution

type foo | find "" /v /c

---

Powershell solutions

(get-content foo | measure-object -line).lines

(dir foo | select-string .).count

(type foo).count

(gc foo | measure-object | select count).count

---

perl solution

perl -pe '}{$_=$.' foo

---

awk solution

awk 'END { print NR }' foo

---

Further reading

• An A-Z Index of the Windows CMD command line - An excellent reference for all things Windows cmd line related.
• find - Search for a text string in a file & display all the lines where it is found.
• An A-Z Index of Windows PowerShell commands - An excellent reference for all things PowerShell related.

Answer 2

The fastest for me is the command line: find /v /c "" foo.txt

I am yet to do a proper benchmark but I tried type and (get-content foo | measure-object -line) for a 10GB file on a 2GB ram server, but they took far too long to complete. Find however returned fast enough