Using awk to count names in a column - awk

Question

Indicates the initials of people who have three proper names;
so, i have a column in this case it's
awk -F ';' '{print $1, $2}' users.txt
output:
xxx JoaoPedroVilar
xxa JoaoMiguel
RMF RitaPereira
....
My question is: i need with function count ( i guess) , so count only in the column $2 which names have more than 2 names... because i just want in the output Acronyms and names with 2 plus letter upppercases like this:
xxx JoaoPedroVilar
RAT RicardoAntonioPereira
Sample data:
awk -F ';' '{print $1, $2}' users.txt
Output:
xxx NunoAndr�Ferreira
xxx HugoFernandes
xxx HugoGomes
xxx In�sSilva
xxx Jo�oTeixeira
xxx JoaquimGon�alves
JAR JoaquimRibeiro
xxx Jos�PedroRafael
xxx Jos�Soares
xxx LuisFernandes
xxx MiguelMadeira
xxx NunoAndr�Ferreira
xxx PedroLucasFarinha
the answer is :
awk -F';' -b '$2~/[A-Z]{1}.*[A-Z]{1}.*[A-Z]{1}.*/{print $1, $2}' users.txt
So -b it's just awk interpret all caracthers

Answer 1

This regex can probably be refactored, but it does the trick (I think)
awk -F';' '$2~/[A-Z]{1}.*[A-Z]{1}.*[A-Z]{1}.*/{print $1, $2}' users.txt
That's just matching for 3 single upper case characters in your second column. Note that this is going to have false positives if you have names like ScottMcMasters or BobO'Neal but trying to seperate names that aren't already separated is never 100%.
Example:
 cat users.txt
xxx;JoaoPedroVilar
xxx;PedroAndrePereira
RAT;RicardoAntonioPereira
xxx;BobBob
xxx;SomeName
 awk -F";" '$2~/[A-Z]{1}.*[A-Z].*[A-Z].*/' users.txt
xxx;JoaoPedroVilar
xxx;PedroAndrePereira
RAT;RicardoAntonioPereira

Answer 2

Without testable input/output it's a guess but it sounds like you need:
awk -F';' '$2 ~ /([[:upper:]][^[:upper:]]+){2}[[:upper:]]/{print $1, $2}' file
e.g. using the output you posted elsewhere to create sample input:
$ cat file
xxx;NunoAndr�Ferreira
xxx;HugoFernandes
xxx;HugoGomes
xxx;In�sSilva
xxx;Jo�oTeixeira
xxx;JoaquimGon�alves
JAR;JoaquimRibeiro
xxx;Jos�PedroRafael
xxx;Jos�Soares
xxx;LuisFernandes
xxx;MiguelMadeira
xxx;NunoAndr�Ferreira
xxx;PedroLucasFarinha
.
$ awk -F';' '$2 ~ /([[:upper:]][^[:upper:]]+){2}[[:upper:]]/{print $1, $2}' file
xxx NunoAndr�Ferreira
xxx Jos�PedroRafael
xxx NunoAndr�Ferreira
xxx PedroLucasFarinha
If that doesn't work for you then try setting your locale to C first (and next try setting it to whatever local understands the "control chars" in your files as letters) as you're probably just having a locale problem:
LC_ALL=C awk -F';' '$2 ~ /([[:upper:]][^[:upper:]]+){2}[[:upper:]]/{print $1, $2}' file

Answer 3

another approach of counting upper case chars
$ awk -F';' 'gsub(/[A-Z]/,"&",$2)>2 {print $1,$2}'

Answer 4

with:
awk -F";" '$2~/[A-Z]{1}.*[A-Z].*[A-Z].*/' users.txt
Output:
PLF PedroLucasFarinha
but when i print just the names i have example:
awk -F ';' '{print $1, $2}' users.txt
Output:
xxx NunoAndr�Ferreira
xxx HugoFernandes
xxx HugoGomes
xxx In�sSilva
xxx Jo�oTeixeira
xxx JoaquimGon�alves
JAR JoaquimRibeiro
xxx Jos�PedroRafael
xxx Jos�Soares
xxx LuisFernandes
xxx MiguelMadeira
xxx NunoAndr�Ferreira
xxx PedroLucasFarinha
but with :
awk -F';' -b '$2~/[A-Z]{1}.*[A-Z]{1}.*[A-Z]{1}.*/{print $1, $2}' users.txt
the output it's
xxx Jos�PedroRafael
xxx NunoAndr�Ferreira
xxx PedroLucasFarinha