On 24 Feb 2020 at 08:53p, Avon pondered and said...
I wonder how best to automate things such that Zone, Region and Host are excluded along with Down but the rest are counted?
Here ya' go:
: fat-dragon; cat scr
#!/bin/sh
grep -v '^;' "$@" |
tr -C -d '[[:graph:][:space:]]' |
awk -F, '
$1 ~ /^(Down|Host|Zone|Region)$/ {next}
$4 != "" && $4 !~ /AnyCountry/ {print $4}' |
sed 's/.*_//' |
sort |
uniq -c |
sort -nr |
awk '
BEGIN {
while (getline<"cc.wiki") {
k=$1
$1=""
sub(" ", "")
cc[k]=$0
}
cc["GER"]="Germany"
}
{printf "%4d %s\n", $1, !cc[$2]?$2:cc[$2]}'
: fat-dragon; sh scr FSXNET.059
139 United States
25 Canada
20 Australia
10 New Zealand
7 Sweden
7 United Kingdom
6 Germany
4 Argentina
3 Netherlands
3 Brazil
2 Italy
2 Spain
1 Taiwan
1 Slovakia
1 Singapore
1 Romania
1 Philippines
1 Greece
1 France
1 Finland
1 Czech Republic
1 Costa Rica
1 Austria
: fat-dragon;
A couple of quick notes.... First, I downloaded
the ISO-3166-1 Alpha 3 list from Wikipedia and
massaged it into a plain 'ol text file containing
the 3-letter country code first and then the name
of country for the remainder of the line. A little
'awk' invocation reads that in a `BEGIN` block and
creates an array so we can map country code to name.
Note that "GER" for (I presume) Germany doesn't
exist in ISO-3166-1. However, it does exist in the
list of United Nations development program country
codes. This appears to date from the cold war and
pre-reunification.
Anyway, after we read the file we manually add an
entry to the array. The use of the tertiary operator
in the `printf` invocation is so we print out the
abbreviation if we don't have a country name. It's
not relevant now, but who knows what political
divisions will exist in the future?
Also, the `tr` is unnecessary. It just strips out
non-printing characters like the '^Z' at the end of
the file, but that doesn't appear in the location
field so it can be ignored.
The 'awk-reads-a-file' thing is a little too cute.
But note that `sed` can be made to read a script
from a file. So instead of having a mapping of
abbreviation to name in `cc`, we can have a list of
`sed` commands instead. Here's the first few lines:
: fat-dragon; head cc.sed
s/ ABW$/ Aruba/
s/ AFG$/ Afghanistan/
s/ AGO$/ Angola/
s/ AIA$/ Anguilla/
s/ ALA$/ _land Islands/
s/ ALB$/ Albania/
s/ AND$/ Andorra/
s/ ARE$/ United Arab Emirates/
s/ ARG$/ Argentina/
s/ ARM$/ Armenia/
: fat-dragon;
Note the spaces and the dollar signs ('$'):
these anchor the regular expressions for the
abbreviations so that they actually match what
we think we're matching (as opposed to some
substring of some weirdly named country, for
example). Again, probably unnecessary, but if
we can be a little defensive and avoid bugs,
why not? CPU cycles are cheap.
We still have that pesky Germany case, but we can
take care of that on the command line. Here's our
new script and it's output:
: fat-dragon; cat scr2
#!/bin/sh
grep -v '^;' "$@" |
awk -F, '
$1 ~ /^(Down|Host|Zone|Region)$/ {next}
$4 != "" && $4 !~ /AnyCountry/ {print $4}' |
sed 's/.*_//' |
sort |
uniq -c |
sort -nr |
sed -f cc.sed -e 's/ GER$/ Germany/'
: fat-dragon; sh scr2 FSXNET.059
139 United States
25 Canada
20 Australia
10 New Zealand
7 Sweden
7 United Kingdom
6 Germany
4 Argentina
3 Netherlands
3 Brazil
2 Italy
2 Spain
1 Taiwan
1 Slovakia
1 Singapore
1 Romania
1 Philippines
1 Greece
1 France
1 Finland
1 Czech Republic
1 Costa Rica
1 Austria
: fat-dragon;
I'd consider changing the nodelist to update Germany's
country code to `DEU` as in ISO-3166-1, though.
--- Mystic BBS v1.12 A45 2020/02/18 (Windows/32)
* Origin: Agency BBS | Dunedin, New Zealand | agency.bbs.nz (21:1/101)