Regex to identify unique words in a dictionary database

Shell Programming and Scripting
1

Hello,
I have a dictionary which I am building for the Open Source Community. The data structure is as under

HEADWORD=PARTOFSPEECH=ENGLISH MEANING

as shown in the example below

=m=Prefix signifying negation.
=ind=Interjection expressing disapprobation.
=int=An interjection expressing contempt,unconcern,disbelief.
=m=A figure;a mark.The thigh.An act of a play.
=n=Arithmetic.
=m=A number,figure.A hook.
=f=A pole with a hook at the extremity.
=f=A ruler,a division.
=vt=Mark;rule;sketch.
=n=Marking;numbering.
=f=A ticket,label showing price of an article.
=m=Permutations.
=f=Operations with figures.
=f=Operations with figures.
=f=Figure-writing.
 =vi=To be completely subject to the authority of.
=a=Marked;defined.Circumscribed,limited.
=a=Figured,numbered.
=a=Unhesitating;unstopped.
 =vi=To give indications of future character.
=m=A sprout.Germination.
=m=An elephant goad.
=m=Alangium Lamarku.
=f=A ruler.Marking;a division.
=f=A ruler.Marking;a division.
 =E=To have a hand in.
 =vt=tremble,shiver,quake
 =E=Withdraw one's self from.
 =E=Contract one's self,evade.
 =vi=To spare one's self;to work lazily.
 =E=Decline,disallow vehemently.
 =E=Lose flesh;lie down carelessly.
 =vt=Be seized with rheumatic affection;gain flesh.
 =vt=To gain flesh,to have cramps.
 =n=Bathing,ablution-used by,of women.
 =n=Turning over from one side to the other.
   =E=To be unstinted in effort,work strenuously.
  =E=Have the aching premonitory symptoms of fever.
=n=Body.Limb.Side.Concern in.Ability.Support.

I need to identify within the headwords, only those which are single headwords and not those where a Headword is made of more than a single headword
Thus

=m=Prefix signifying negation.
=ind=Interjection expressing disapprobation.
=int=An interjection expressing contempt,unconcern,disbelief.
=m=A figure;a mark.The thigh.An act of a play.
=n=Arithmetic.
=m=A number,figure.A hook.
=f=A pole with a hook at the extremity.
=f=A ruler,a division.
=vt=Mark;rule;sketch.
=n=Marking;numbering.
=f=A ticket,label showing price of an article.
=m=Permutations.
=f=Operations with figures.
=f=Operations with figures.
=f=Figure-writing.
=n=Body.Limb.Side.Concern in.Ability.Support.

should be identified but the following which have more than one word are not valid and should not be identified

 =E=To have a hand in.
 =vt=tremble,shiver,quake
 =E=Withdraw one's self from.
 =E=Contract one's self,evade.
 =vi=To spare one's self;to work lazily.
 =E=Decline,disallow vehemently.
 =E=Lose flesh;lie down carelessly.
 =vt=Be seized with rheumatic affection;gain flesh.
 =vt=To gain flesh,to have cramps.
 =n=Bathing,ablution-used by,of women.
 =n=Turning over from one side to the other.
   =E=To be unstinted in effort,work strenuously.
  =E=Have the aching premonitory symptoms of fever.

A regex in Perl or Unix would be really useful. Thanks a lot.

2

The regular expression (RE) you need depends on what tool you're using and what you want the RE to do.

If you were using awk and wanted an ERE to select lines from your file that just have one headword, you might try:

awk '/^[^ =]*=/' file

As always, if you want to try this on a Solaris/SunOS system, change awk to /usr/xpg4/bin/awk or nawk .

3

Many thanks. I am sorry, I should have specified that I work in a Windows environment.
Playing round with the AWK regex, I found that it worked beautifully as a Unix regex. My text-editor allows me to choose Unix/Perl as Regexes. All I had to was strip off the forward slashes and I could identify all the singletons.