Re^2: ordering array of sentences

[welle]

in reply to Re: ordering array of sentences
in thread ordering array of sentences

Thanks,it works (couple of changes can be necessary for taking into account case, commas, etc. BUT: as I am a newbe in perl, could you "split" (i.e. putting it in simpler blocks) for me the lines

say for map $_->[1],
    sort { $a->[0] cmp $b->[0] }
    map { /(\w+)\s+searching/; [$1, $_] }
    @sentences;

so that I better understand its functioning. For example, I am trying to put the result of the sorting again in an array, but I can't do it. That would be great!

[choroba]

As usual in functional programming, read from right to left (or bottom to top):

The map changes the sentences to arrays of arrays. Each such an array has two elements: the word to sort on and the original sentence.

sort just sorts the arrays according to their first element, i.e. the sort word.

map transforms the arrays back to sentences.

[welle]

I got it. Thank you!

In a very ugly Perl, I translated it so:

    my @sentences1 = map { /(\w+)\s+searching/; [$1, $_] }  @sentences
+;
    my @sentences2 = sort { $a->[0] cmp $b->[0] } @sentences1;
    my @sentences3 = map {$_->[1]} @sentences2;
    
    foreach (@sentences3) {
        print $_;
    }

...just to lear Perl

[welle]

What if my @sentences has a more complicated structure? I get @sentences from querying a Sqlite database with the following:

$sentences = $dbh->selectall_arrayref("SELECT ID, sentence FROM texts 
+WHERE sentence REGEXP '(?i:$query)'");

As one can see, the result is made of two elements (ID and sentence). How can I read/manipulate in the code provided by choroba only the second element "sentence", preserving at the end of the sorting operation the same structure ID/sentence?

[afoken]

1. use ORDER BY in SQL to have the data delivered to perl in the expected order
2. Schwartzian transform

Alexander