Perl regex extract parts of string with multiple condition

Question

What's the single regex that enables me to capture all the text that goes after are genes and is gene from this text

The closest human genes of best are genes A B C
The closest human gene of best is gene A

Hence I hope to extract $1 that contain

A B C
A

Tried this but fail:

$line =~ /The closest .* gene[s] (.*)$/;

Do you also need to avoid illegal strings like "... gene of best are A"? — tiftik
– tiftik, Commented Apr 14, 2010 at 11:38

SilentGhost · Accepted Answer · 2010-04-14 11:09:50Z

4

$line =~ /The closest .* genes? (.*)$/;

answered Apr 14, 2010 at 11:09

SilentGhost

322k67 gold badges312 silver badges294 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

kbenson Over a year ago

+1 for matching requester's example as close as possible, but this could benefit from some information explaining that [s] is the same as s, [s ] would have been what he was trying to accomplish with that, and that s? is equivalent.

Axeman · Accepted Answer · 2010-04-14 12:44:51Z

3

I think the most explicit is:

$line =~ m/best \s (?:is \s gene|are \s genes) \s ([\p{IsUpper}](?: \s [\p{IsUpper} ])*)/x;

Of course if you know that all sentences are going to be grammatical, then you can do the (?:are|is) thing. And if you know that you're only going to have genes A-N or something, you can forget the \p{IsUpper} and use [A-N].

answered Apr 14, 2010 at 12:44

Axeman

29.9k2 gold badges50 silver badges104 bronze badges

Comments

ghostdog74 · Accepted Answer · 2010-04-14 11:15:51Z

2

$ perl -F/genes*/ -ane 'print $F[-1];' file
 A B C
 A

answered Apr 14, 2010 at 11:15

ghostdog74

346k62 gold badges264 silver badges349 bronze badges

Comments

Donal Fellows · Accepted Answer · 2010-04-14 11:24:35Z

2

Use non-greedy at the beginning to reduce the opportunities for surprises. Use non-capturing parens to group alternatives that you don't care about. Append ? to a letter to make it optional. Hence, try this:

$line =~ /The closest .*? (?:is|are) genes? (.*)$/;

To see where you were going wrong BTW, just compare the above with what you were originally trying.

answered Apr 14, 2010 at 11:24

Donal Fellows

139k19 gold badges161 silver badges222 bronze badges

5 Comments

Donal Fellows Over a year ago

It captures some cases that are bad grammar too (“The closest ... is genes ..”) but that's hardly important, yes? :-)

SilentGhost Over a year ago

if it's not important why bother with that non-capturing group at all?

Donal Fellows Over a year ago

@SilentGhost: Without it, you'll capture from the first instance of the word "gene" to the end, e.g., “of best are genes A B C”.

SilentGhost Over a year ago

that's only because of using non-greedy quantifier

Donal Fellows Over a year ago

There's not really enough input data samples in the question to be able to work out what is wanted. I personally prefer to match more in the fixed proportion to reduce the number of landmines^Wsurprises in the matched text.

Space · Accepted Answer · 2010-04-14 11:20:05Z

0

With the other suggestions, I would like to suggest to have a look at the perllre for Regular Expressions

answered Apr 14, 2010 at 11:20

Space

7,2997 gold badges51 silver badges68 bronze badges

Collectives™ on Stack Overflow

Perl regex extract parts of string with multiple condition

5 Answers 5

1 Comment

Comments

Comments

5 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

5 Answers 5

1 Comment

Comments

Comments

5 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related