script/cfi_learn_model.pl untested

filter: Statementstmt 0.0% 0 / 35 Branchbran 0.0% 0 / 2 Conditioncond n/a MC/DCmcdc n/a Subroutinesub 0.0% 0 / 4 Podpod n/a totaltotal 0.0% 0 / 41 CC 1 SCAR 6.9
CC
1
Cov
0%
BEGIN
2.0
BEGIN
2.0
BEGIN
2.0
CRAP 2.0
1 #!/usr/local/bin/perl
2 # (c) 2020 by Andreas Romeyke
3 # licensed via GPL v3.0 or later
4 # call ./cfi_lear_model.pl CSV-file
5 # outputs perl code for a model of File::FormatIdentification::RandomSampling::Model
6 # PODNAME: cfi_learn_model.pl
7 use strict;
8 use warnings FATAL => 'all';
9 use feature qw(say);
10 use AI::DecisionTree;
11 my $dtree = AI::DecisionTree->new( noise_mode => 'pick_best', verbose => 1, max_depth => 8);
12 my $training_file = shift @ARGV;
13 open (my $fh, "<", $training_file);
14 my $header_line = <$fh>;
15 #chomp $header_line;
16 #my @header = split(/,/, $header_line);
17 my @header = qw(onegram1 onegram2 onegram3 onegram4 onegram5 onegram6 onegram7 onegram8 bigram1 bigram2 bigram3 bigram4 bigram5 bigram6 bigram7 bigram8 mimetype);
18 foreach my $line (<$fh>) {
19 chomp $line;
20 if ($line =~ m/^onegram/) { next;}
21 my @values = split(/, */, $line);
22 $dtree->add_instance(
23 attributes => {
24 $header[0] => $values[0],
25 $header[1] => $values[1],
26 $header[2] => $values[2],
27 $header[3] => $values[3],
28 $header[4] => $values[4],
29 $header[5] => $values[5],
30 $header[6] => $values[6],
31 $header[7] => $values[7],
32 $header[8] => $values[8],
33 $header[9] => $values[9],
34 $header[10] => $values[10],
35 $header[11] => $values[11],
36 $header[12] => $values[12],
37 $header[13] => $values[13],
38 $header[14] => $values[14],
39 $header[15] => $values[15],
40 },
41 result => $values[16],
42 name => $header[16],
43 );
44 #p($dtree);
45  
46 }
47 $dtree->train();
48 #use Data::Printer;
49 #my $ruletree = $dtree->rule_tree();
50 #p($ruletree);
51 #p( $dtree);
52 my @rule_statements = $dtree->rule_statements();
53 say <<'HEADER';
54 package File::FormatIdentification::RandomSampling::Model;
55 # ABSTRACT: methods to identify files using random sampling
56 # VERSION:
57 # (c) 2020 by Andreas Romeyke
58 # licensed via GPL v3.0 or later
59 use strict;
60 use warnings;
61 use feature qw(say);
62 use Moose;
63 use List::Util qw( any );
64  
65 sub calc_mimetype {
66 my $self = shift;
67 my $histogram = shift;
68 my @bigrams = @{$histogram->{bigram}};
69 my @onegrams = @{$histogram->{onegram}};
70 HEADER
71  
72  
73 say join("\n", map {
74 s/if ([^-]*)-> ('[^']*')/\tif ($1) { return $2 ; }/;
75 s/bigram[0-9]*='([^']*)'/(any {\$_ == $1 } \@bigrams)/g;
76 s/onegram[0-9]*='([^']*)'/(any {\$_ == $1 } \@onegrams)/g;
77 $_;
78 } @rule_statements);
79  
80 say <<'FOOTER';
81 return 'unknown';
82 }
83  
84 1;
85 FOOTER
86  
87 1;
88  
89 __END__