用具有匹配键的哈希值替换文件中的文本

Question

I would like to replace all words in a file matching the keys of my hash with corresponding values.

Hash:

$VAR1 = {
    'asmbl_1'  => 'TCONS_00000046',
    'asmbl_2'  => 'TCONS_00000014',
    'asmbl_16' => 'MELO3C000012',
}

File:

CM3.6.1_CONTIG30890 assembler   transcript  187 1568    .   -   .   gene_id "PASA_cluster_1"; transcript_id "align_id:184317|asmbl_1";
CM3.6.1_CONTIG30890 assembler   exon    187 251 .   -   .   gene_id "PASA_cluster_1"; transcript_id "align_id:184317|asmbl_1";
CM3.6.1_CONTIG30898 assembler   exon    1339    2793    .   -   .   gene_id "PASA_cluster_2"; transcript_id "align_id:184318|asmbl_2";

Desired output:

CM3.6.1_CONTIG30890 assembler   transcript  187 1568    .   -   .   gene_id "PASA_cluster_1"; transcript_id "align_id:184317|TCONS_00000046";
CM3.6.1_CONTIG30890 assembler   exon    187 251 .   -   .   gene_id "PASA_cluster_1"; transcript_id "align_id:184317|TCONS_00000046";
CM3.6.1_CONTIG30898 assembler   exon    1339    2793    .   -   .   gene_id "PASA_cluster_2"; transcript_id "align_id:184318|TCONS_00000014";

I'm looking for a straightforward way to do this, preferably in Perl, since I'm writing a script in Perl.

Approaches:

• Read the file line by line, extract the key from the file, match this key in hash and replace it by the value.
• Read hash pair by pair, open file, read line by line and replace matches.

(What is the difference between these both methods?)

• Read hash pair by pair and call bash " sed -i '/key/value/' ". A bit ugly, I would prefer to do all in Perl.

Answer 1

There's a nice trick I like, that basically involves building a regex and using that to capture and match your regex:

use strict;
use warnings;

my %replace = (
    'asmbl_1'  => 'TCONS_00000046',
    'asmbl_2'  => 'TCONS_00000014',
    'asmbl_16' => 'MELO3C000012',
);

my $search = join( "|", map {quotemeta} sort { length ($b) <=> length ($a) } keys %replace );
$search = qr/\b($search)\b/;

while (<>) {
    s/$search/$replace{$1}/g;
    print;
}

Something like that produces the desired output. (Diamond operators to read the content off STDIN or invocation via myscript.pl <some_File_To_process>

Answer 2

This is all that is necessary

use strict;
use warnings;

my %map = (
    asmbl_1  => 'TCONS_00000046',
    asmbl_2  => 'TCONS_00000014',
    asmbl_16 => 'MELO3C000012',
);

my $re = join '|', map quotemeta, keys %map;

while ( <DATA> ) {
    s/\b($re)\b/$map{$1}/g;
    print;
}

__DATA__
CM3.6.1_CONTIG30890 assembler   transcript  187 1568    .   -   .   gene_id "PASA_cluster_1"; transcript_id "align_id:184317|asmbl_1";
CM3.6.1_CONTIG30890 assembler   exon    187 251 .   -   .   gene_id "PASA_cluster_1"; transcript_id "align_id:184317|asmbl_1";
CM3.6.1_CONTIG30898 assembler   exon    1339    2793    .   -   .   gene_id "PASA_cluster_2"; transcript_id "align_id:184318|asmbl_2";

output

CM3.6.1_CONTIG30890 assembler   transcript  187 1568    .   -   .   gene_id "PASA_cluster_1"; transcript_id "align_id:184317|TCONS_00000046";
CM3.6.1_CONTIG30890 assembler   exon    187 251 .   -   .   gene_id "PASA_cluster_1"; transcript_id "align_id:184317|TCONS_00000046";
CM3.6.1_CONTIG30898 assembler   exon    1339    2793    .   -   .   gene_id "PASA_cluster_2"; transcript_id "align_id:184318|TCONS_00000014";