How To: Convert Text Following Title Case Rules in Bash

问题

How to convert string to title case while following rules, not just simply capitalizing every first letter of the word?

Sample rule:

Capitalize all words, with exception to:
Lowercase all articles (a, the), prepositions (to, at, in, with), and coordinating conjunctions (and, but, or)
Capitalize the first and last word in a title, regardless of part of speech

Any easy way to do this in bash? One-liners appreciated.

(And just as an additional note, this is to be used in parcellite actions.)

回答1:

$ cat titles.txt
purple haze
Somebody To Love
fire on the mountain
THE SONG REMAINS THE SAME
Watch the NorthWind rise
eight miles high
just dropped in
strawberry letter 23

$ cat cap.awk
BEGIN { split("a the to at in on with and but or", w)
        for (i in w) nocap[w[i]] }

function cap(word) {
    return toupper(substr(word,1,1)) tolower(substr(word,2))
}

{
  for (i=1; i<=NF; ++i) {
      printf "%s%s", (i==1||i==NF||!(tolower($i) in nocap)?cap($i):tolower($i)),
                     (i==NF?"\n":" ")
  }
}

$ awk -f cap.awk titles.txt
Purple Haze
Somebody to Love
Fire on the Mountain
The Song Remains the Same
Watch the Northwind Rise
Eight Miles High
Just Dropped In
Strawberry Letter 23

EDIT (as a one liner):

$ echo "the sun also rises" | awk 'BEGIN{split("a the to at in on with and but or",w); for(i in w)nocap[w[i]]}function cap(word){return toupper(substr(word,1,1)) tolower(substr(word,2))}{for(i=1;i<=NF;++i){printf "%s%s",(i==1||i==NF||!(tolower($i) in nocap)?cap($i):tolower($i)),(i==NF?"\n":" ")}}'
The Sun Also Rises

回答2:

Thanks @jas for giving a nice answer to this one. Eventually, what I needed for parcellite is this one-long-liner in the shell: (For the love of the pipe!)

echo '%s' | sed 's/\<./\u&/g' | sed 's/\ The\ /\ the\ /' | sed 's/\ A\ /\ a\ /' | sed 's/\ An\ /\ an\ /' | sed 's/\ As\ /\ as\ /' | sed 's/\ At\ /\ at\ /' | sed 's/\ But\ /\ but\ /' | sed 's/\ By\ /\ by\ /' | sed 's/\ For\ /\ for\ /' | sed 's/\ In\ /\ in\ /' | sed 's/\ Of\ /\ of\ /' | sed 's/\ Off\ /\ off\ /' | sed 's/\ On\ /\ on\ /' | sed 's/\ Per\ /\ per\ /' | sed 's/\ To\ /\ to\ /' | sed 's/\ Up\ /\ up\ /' | sed 's/\ Via\ /\ via\ /' | sed 's/\ And\ /\ and\ /' | sed 's/\ Nor\ /\ nor\ /' | sed 's/\ Or\ /\ or\ /' | sed 's/\ So\ /\ so\ /' | sed 's/\ Yet\ /\ yet\ /' | parcellite

The seds were of course generated from a loop:

for word in {The,A,An,As,At,But,By,For,In,Of,Off,On,Per,To,Up,Via,And,Nor,Or,So,Yet}
do
    low=`echo "$word" | tr '[A-Z]' '[a-z]'`
    printf "sed 's/\ $word\ /\ $low\ /' | "
done

Thanks for those who tried. :-)

来源：https://stackoverflow.com/questions/35006611/how-to-convert-text-following-title-case-rules-in-bash

标签

string

bash

awk

sed