Count number of names starts with particular character in file

i have the following file::

FirstName, FamilyName, Address, Pho开发者_如何学PythonneNo

the file is sorted according to the family name, how can i count the number of family names starts with a particular character ??

output should look like this ::

A: 2
B: 1
...

??

---

With awk:

awk '{print substr($2, 1, 1)}' file|
  uniq -c|
  awk '{print $2 ": " $1}'

OK, no awk. Here's with sed:

sed s'/[^,]*, \(.\).*/\1/' file|
  uniq -c|
  sed 's/.*\([0-9]\)\+ \([a-zA-Z]\)\+/\2: \1/'

OK, no sed. Here's with python:

import csv
r = csv.reader(open(file_name, 'r'))
d = {}
for i in r:
    d[i[1][1]] = d.get(i[1][1], 0) + 1
for (k, v) in d.items():
    print "%s: %s" % (k, v)

---

while read -r f l r; do echo "$l"; done < inputfile | cut -c 1 | sort | uniq -c

---

Just the Shell

#! /bin/bash

##### Count occurance of familyname initial

#FirstName, FamilyName, Address, PhoneNo
exec <<EOF
Isusara, Ali,      Someplace,    022-222
Rat,     Fink,     Some Hole,    111-5555
Louis,   Frayser,  whaterver,  123-1144
Janet,   Hayes,    whoever St,      111-5555
Mary,    Holt,     Henrico VA,   222-9999
Phillis, Hughs,    Some Town,   711-5525
Howard,  Kingsley, ahahaha,   222-2222
EOF



while read first family rest
do
  init=${family:0:1}
  [ -n "$oinit" -a $init != "$oinit" ]  && {
      echo $oinit : $count
      count=0
  }
  oinit=$init
  let count++
done

echo $oinit : $count

Running

frayser@gentoo ~/doc/Answers/src/SH/names $ sh names.sh
A : 1
F : 2
H : 3
K : 1
frayser@gentoo ~/doc/Answers/src/SH/names $ 

To read from a file, remove the here document, and run:

chmod +x names.sh
./names.sh <file

---

The "hard way" — no use of awk or sed, exactly as asked for. If you're not sure what any of these commands mean, you should definitely look at the man page for each one.

INTERMED=`mktemp`       # Creates a temporary file
COUNTS_L=`mktemp`       # A second...
COUNTS_R=`mktemp`       # A third...

cut -d , -f 2 |         # Extracts the FamilyName field only

tr -d '\t ' |           # Deletes spaces/tabs

cut -c 1 |          # Keeps only the first character
                # on each line

tr '[:lower:]' '[:upper:]' |    # Capitalizes all letters

sort |              # Sorts the list

uniq -c > $INTERMED     # Counts how many of each letter
                # there are

cut -c1-7 $INTERMED |       # Cuts out the LHS of the temp file
tr -d ' ' > $COUNTS_R       # Must delete the padding spaces though


cut -c9- $INTERMED > $COUNTS_L  # Cut out the RHS of the temp file

# Combines the two halves into the final output in reverse order
paste -d ' ' /dev/null $COUNTS_R | paste -d ':' $COUNTS_L -

rm $INTERMED $COUNTS_L $COUNTS_R    # Cleans up the temp files

---

awk one-liner:

awk '
    {count[substr($2,1,1)]++} 
    END {for (init in count) print init ": " count[init]}
' filename

---

Prints the how many words start with each letter:

for i in {a..z}; do echo -n "$i:"; find path/to/folder -type f -exec sed "s/ /\n/g" {} \; | grep ^$i | wc -c | awk '{print $0}'; done