Re: Form feeds in source code

[Blake McBride]

In the old days, with limited memory, editors would only read a file up to the formfeed.  The programmer would edit the text up to the formfeed and then read in the next block (up to the next formfeed).  This way the whole file need not be in memory at once.  Compilers ignore it.  Obviously they can be eliminated now.

Blake McBride

On Fri, Sep 18, 2020 at 11:08 AM Pete Dietl <petedietl@gmail.com> wrote:

Why are there so many form feeds in the source?

Running this from the root of the Make repo:

cd src && grep -c $'\f' *.[ch] | grep -v ':0$' | sort -t : -k 2 -n -r | awk -F : 'BEGIN {printf "%-15s Number of form feeds\n", "File"; print "------------------------------------" } {printf "%-15s %2d\n", $1, $2}'

Results in:

File            Number of form feeds
------------------------------------
variable.c      17
read.c          15
misc.c          12
job.c           11
function.c      10
dir.c           10
remake.c         9
main.c           8
file.c           8
expand.c         6
commands.c       6
vpath.c          5
rule.c           5
output.c         5
getopt.c         4
arscan.c         4
remote-stub.c    3
remote-cstms.c   3
ar.c             3
implicit.c       2
getopt1.c        2
version.c        1
makeint.h        1
hash.h           1
default.c        1