transpose: parallelized.

w4it-for-port-open: -q (quiet) implemented.
This commit is contained in:
Ole Tange 2018-03-27 02:56:54 +02:00
parent 0a9b2b3503
commit f6a34e1200
4 changed files with 178 additions and 113 deletions

View file

@ -1,125 +1,160 @@
#!/usr/bin/perl -w #!/bin/bash
use English; # transpose [-d delimiter] [-b blocksize] table.csv > transposed.csv
use FileCache; # cat table.csv | transpose [-d delimiter] [-b blocksize] > transposed.csv
use File::Temp;
my $delimiter = shift; transpose_inner() {
my $buffer = shift; # simple in-memory transpose
# -d sep
# Input:
# data to be transposed
# Output:
# transposed data
perl <(cat <<'cut-here-UbsAqi0j6GoOuk5W5yWA'
use Text::CSV;
use Getopt::Long;
$delimiter ||= ","; Getopt::Long::Configure("bundling","require_order");
# Use at most 1000M before flushing my $retval = GetOptions("debug|D=s" => \$opt::debug,
$buffer ||= "1000M"; "delimiter|d=s" => \$opt::delimiter,
$buffer = multiply_binary_prefix($buffer); "verbose|v" => \@opt::verbose,
# Perl makes the buffer baloon to 10 times the requested value "simple|s" => \$opt::simple,
$buffer /= 10; );
# max_col_size will be lowered after first line read.
$max_col_size = $buffer; if(defined $opt::delimiter) {
my $delimiter_regexp = $delimiter; simple();
$delimiter_regexp =~ s/(\W)/\\$1/g; } else {
my @current; die("-d must be set");
my $col_no_last_line = 0; }
my $lineno = 0;
my %col; sub simple {
while(<>) { my (@table);
chomp; my $col = 0;
my $col_no = 0; my $csv_setting = { binary => 1, sep_char => $opt::delimiter };
my @to_be_flushed = (); my $sep = $csv_setting->{sep_char};
map { my $csv = Text::CSV->new($csv_setting)
push(@{$col{$col_no}},$_); or die "Cannot use CSV: ".Text::CSV->error_diag ();
$col_size{$col_no} += length $_;
if($col_size{$col_no} > $max_col_size) { while(my $l = <>) {
push @to_be_flushed, $col_no; if(not $csv->parse($l)) {
$col_size{$col_no} = 0; die "CSV has unexpected format";
} }
$col_no++; # append to each row
} split /$delimiter_regexp/o, $_; # This should do de-csv'ing my $row = 0;
if(@to_be_flushed) { for($csv->fields()) {
flush(\%col,@to_be_flushed); $table[$row][$col] = defined($_) ? $_ : '';
} $row++;
if($col_no != $col_no_last_line) {
if(0 == $col_no_last_line) {
# This is first time around
$col_no_last_line = $col_no;
$max_col_size = $buffer/$col_no_last_line;
} else {
warning("Number of columns in line $NR: $col_no != $col_no_last_line\n");
} }
$col++;
} }
print map { join($sep,@$_),"\n" } @table;
} }
flush(\%col); cut-here-UbsAqi0j6GoOuk5W5yWA
merge(); ) "$@"
}
export -f transpose_inner
sub flush { stdin_to_paste_files() {
my $col_ref = shift; # Run transpose_inner on blocks from stdin
my @cols_to_flush = @_; # output each block as file name
if(not @cols_to_flush) { local block_size
@cols_to_flush = keys %$col_ref; local sep
} block_size="$1"
for my $c (@cols_to_flush) { sep="$2"
$Global::tempfile{$c} ||= tmpnam(); PARALLEL="-k --files --block $block_size" \
my $fh = cacheout $Global::tempfile{$c}; parallel --pipe transpose_inner -d "'$sep'"
# This will print one delimiter too much, which we will deal with later
print $fh map { $_,$delimiter } @{$col_ref->{$c}};
delete $col_ref->{$c};
}
} }
sub merge { file_to_paste_files() {
for my $c (sort keys %Global::tempfile) { # Run transpose_inner on blocks from $file
my $fh = cacheout $Global::tempfile{$c}; # output each block as file name
# truncate by length of delimiter to get rid of the last $delimiter local block_size
seek $fh,-length($delimiter),SEEK_END; local sep
truncate $fh, tell $fh; block_size="$1"
# Make sure the file is closed of writing sep="$2"
close $fh; file="$3"
open($fh, "<", $Global::tempfile{$c}) || die; PARALLEL="-k --files --block $block_size" \
my $buf; parallel --pipe-part -a "$file" transpose_inner -d "'$sep'"
while(sysread($fh,$buf,1000_000)) {
print $buf;
}
print "\n";
unlink $Global::tempfile{$c};
}
} }
sub warning { super_paste() {
my @w = @_; # Like 'paste' up to 1000000 files
print STDERR "transpose: Warning: ", @w; # The files are read from stdin
local sep
local paste_files
local fifo
sep="$1"
paste_files=`tempfile`
# basename
fifo=`tempfile`
rm $fifo
cat > $paste_files
# Define replacement string {0#} to 0-pad job number
PARALLEL="--rpl "\''{0#} $f=1+int("".(log(total_jobs())/log(10)));
$_=sprintf("%0${f}d",seq())'\'
# Make fifos that can be read from
cat $paste_files | parallel -n1000 "rm -f $fifo{0#}; mkfifo $fifo{0#}"
# Start a paste process for every 1000 files
cat $paste_files | parallel -n1000 -j0 "paste -d '$sep' {} > $fifo{0#}" &
# Paste all the fifos
eval paste -d "'$sep'" $fifo*
# Cleanup
cat $paste_files | parallel -n1000 "rm -f {} $fifo{0#}"
rm $paste_files
} }
sub error { stdin_detect_sep() {
my @w = @_; # Read the first 3 lines and detect the separator
print STDERR "transpose: Error: ", @w; # Save the read input to file
local file
file="$1"
# TODO
echo "$d"
} }
sub multiply_binary_prefix { usage() {
# Evalualte numbers with binary prefix echo "Usage: $0 [-d delimiter] [-b blocksize]" 1>&2; exit 1;
# k=10^3, m=10^6, g=10^9, t=10^12, p=10^15, e=10^18, z=10^21, y=10^24
# K=2^10, M=2^20, G=2^30, T=2^40, P=2^50, E=2^70, Z=2^80, Y=2^80
# Ki=2^10, Mi=2^20, Gi=2^30, Ti=2^40, Pi=2^50, Ei=2^70, Zi=2^80, Yi=2^80
# ki=2^10, mi=2^20, gi=2^30, ti=2^40, pi=2^50, ei=2^70, zi=2^80, yi=2^80
# 13G = 13*1024*1024*1024 = 13958643712
my $s = shift;
$s =~ s/k/*1000/g;
$s =~ s/M/*1000*1000/g;
$s =~ s/G/*1000*1000*1000/g;
$s =~ s/T/*1000*1000*1000*1000/g;
$s =~ s/P/*1000*1000*1000*1000*1000/g;
$s =~ s/E/*1000*1000*1000*1000*1000*1000/g;
$s =~ s/Z/*1000*1000*1000*1000*1000*1000*1000/g;
$s =~ s/Y/*1000*1000*1000*1000*1000*1000*1000*1000/g;
$s =~ s/X/*1000*1000*1000*1000*1000*1000*1000*1000*1000/g;
$s =~ s/Ki?/*1024/gi;
$s =~ s/Mi?/*1024*1024/gi;
$s =~ s/Gi?/*1024*1024*1024/gi;
$s =~ s/Ti?/*1024*1024*1024*1024/gi;
$s =~ s/Pi?/*1024*1024*1024*1024*1024/gi;
$s =~ s/Ei?/*1024*1024*1024*1024*1024*1024/gi;
$s =~ s/Zi?/*1024*1024*1024*1024*1024*1024*1024/gi;
$s =~ s/Yi?/*1024*1024*1024*1024*1024*1024*1024*1024/gi;
$s =~ s/Xi?/*1024*1024*1024*1024*1024*1024*1024*1024*1024/gi;
$s = eval $s;
return $s;
} }
block_size=10M
while getopts ":b:d:" o; do
case "${o}" in
d)
d="$(printf "${OPTARG}")"
if [ "'" = "${d}" ] ; then
echo "Delimiter cannot be '"
usage
exit
fi
;;
b)
block_size="${OPTARG}"
;;
*)
usage
;;
esac
done
shift $((OPTIND-1))
if [ -z "${d}" ] ; then
d="$(printf "\t")"
fi
# Sep cannot be '
file="$@"
first_lines=`tempfile`
if [ -z "$file" ]; then
sep="$(stdin_detect_sep $first_lines)"
(cat $first_lines; rm $first_lines; cat) |
stdin_to_paste_files $block_size "$sep" | super_paste "$sep"
else
sep="$(stdin_detect_sep < "$file" $first_lines)"
rm $first_lines
file_to_paste_files $block_size "$sep" "$file" | super_paste "$sep"
fi

12
treeoflife/treeoflife Executable file
View file

@ -0,0 +1,12 @@
#!/bin/bash
# Download tree-of-life.jpg
# from http://www.open.edu/openlearn/nature-environment/natural-history/tree-life
base=www2.open.ac.uk/openlearn/treeoflife/treeOfLifePoster/TileGroup
parallel -j30 wget -c $base{3}/6-{1}-{2}.jpg ::: {0..28} ::: {4..42} ::: {1..7}
parallel eval convert +append 6-{0..28}-{}.jpg line{}.jpg ::: {4..42}
convert -append line{4..42}.jpg tree-of-life.jpg
echo Cleanup with
echo rm line*.jpg 6-*-*.jpg

View file

@ -1,5 +1,19 @@
#!/bin/bash #!/bin/bash
QUIET=false
while getopts ":q" opt; do
case $opt in
q)
QUIET=true
shift
;;
\?)
echo "Invalid option: -$OPTARG" >&2
;;
esac
done
HOST=$1 HOST=$1
PORT=$2 PORT=$2
@ -13,7 +27,11 @@ usage () {
} }
print_not_reachable () { print_not_reachable () {
echo -n . if $QUIET; then
true skip
else
echo -n .
fi
} }
is_port_open () { is_port_open () {

View file

@ -70,7 +70,7 @@ if tty -s ; then
# timeout 12 forever dmesg | puniq # timeout 12 forever dmesg | puniq
fi fi
sudo bash -c 'cat >> /etc/resolv.conf' < /etc/resolvconf/resolv.conf.d/head #sudo bash -c 'cat >> /etc/resolv.conf' < /etc/resolvconf/resolv.conf.d/head
sudo iwconfig wls1 essid Turris #sudo iwconfig wls1 essid Turris
sudo dhclient wls1 & #sudo dhclient wls1 &
sudo wpa_supplicant -Dwext -c/etc/wpa_supplicant.conf -iwls1 -d & #sudo wpa_supplicant -Dwext -c/etc/wpa_supplicant.conf -iwls1 -d &