| … | |
… | |
| 7 | and $self->selection (map { y/A-Za-z/N-ZA-Mn-za-m/; $_ } $self->selection); |
7 | and $self->selection (map { y/A-Za-z/N-ZA-Mn-za-m/; $_ } $self->selection); |
| 8 | |
8 | |
| 9 | () |
9 | () |
| 10 | } |
10 | } |
| 11 | |
11 | |
|
|
12 | sub on_init { |
|
|
13 | my ($self) = @_; |
|
|
14 | |
|
|
15 | for (my $idx = 0; defined (my $res = $self->x_resource ("selection.pattern-$idx")); $idx++) { |
|
|
16 | no re 'eval'; # just to be sure |
|
|
17 | $res = utf8::encode $self->locale_decode ($res); |
|
|
18 | push @{ $self->{patterns} }, qr/$res/; |
|
|
19 | } |
|
|
20 | |
|
|
21 | () |
|
|
22 | } |
|
|
23 | |
|
|
24 | # "find interetsing things"-patterns |
| 12 | my @mark_patterns = ( |
25 | my @mark_patterns = ( |
|
|
26 | # common types of "parentheses" |
|
|
27 | qr{ (?<![^[:space:]]) ‘ ([^‘’]+) ’ (?![^[:space]]) }x, |
|
|
28 | qr{ (?<![^[:space:]]) ` ([^`']+) ' (?![^[:space]]) }x, |
|
|
29 | qr{ (?<![^[:space:]]) (" [^[:space:]] [^"]* ") }x, |
|
|
30 | qr{ (" [^"]* [^[:space:]] ") (?![^[:space]]) }x, |
|
|
31 | qr{ \< ([^<>[:space:]]+) \> }x, |
|
|
32 | qr{ \{ ([^{}[:space:]]+) \} }x, |
|
|
33 | qr{ \[ ([^{}[:space:]]+) \] }x, |
|
|
34 | qr{ \( ([^()[:space:]]+) \) }x, |
|
|
35 | |
| 13 | # urls, just a heuristic |
36 | # urls, just a heuristic |
| 14 | qr{( |
37 | qr{( |
| 15 | (?:https?|ftp|news|mailto|file)://[ab-zA-Z0-9\-\@;\/?:&=%\$_.+!*\x27(),~]+ |
38 | (?:https?|ftp|news|mailto|file)://[ab-zA-Z0-9\-\@;\/?:&=%\$_.+!*\x27(),~]+ |
| 16 | [ab-zA-Z0-9\-\@;\/?:&=%\$_+!*\x27(),~] # do not include a trailing dot, its wrong too often |
39 | [ab-zA-Z0-9\-\@;\/?:&=%\$_+!*\x27()~] # exclude some trailing characters (heuristic) |
| 17 | )}x, |
40 | )}x, |
| 18 | |
41 | |
| 19 | # common forms of quoting |
42 | # shell-like argument quoting, basically always matches |
| 20 | qr{(?:^|\s) [‘`] ([^‘`’']+) [’'] (?:\s|$)}x, |
|
|
| 21 | |
|
|
| 22 | # shell-like argument quoting |
|
|
| 23 | qr{\G [\ \t|&;<>()] *( |
43 | qr{\G [\ \t|&;<>()] *( |
| 24 | (?: |
44 | (?: |
| 25 | [^\\"'\ \t|&;<>()]+ |
45 | [^\\"'\ \t|&;<>()]+ |
| 26 | | \\. |
46 | | \\. |
| 27 | | " (?: [^\\"]+ | \\. )* " |
47 | | " (?: [^\\"]+ | \\. )* " |
| 28 | | ' [^']* ' |
48 | | ' [^']* ' |
| 29 | )+ |
49 | )+ |
| 30 | )}x, |
50 | )}x, |
| 31 | ); |
51 | ); |
| 32 | |
52 | |
|
|
53 | # "correct obvious? crap"-patterns |
| 33 | my @simplify_patterns = ( |
54 | my @simplify_patterns = ( |
| 34 | qr{^"([^\\"'\ \t|&;<>()*?]+)"$}, # "simple" => simple |
55 | qr{^"([^\\"'\ \t|&;<>()*?]+)"$}, # "simple" => simple |
| 35 | qr{^(.*)[,\-]$}, |
56 | qr{^(.*)[,\-]$}, # strip off trailing , and - |
| 36 | ); |
57 | ); |
| 37 | |
58 | |
| 38 | sub on_sel_extend { |
59 | sub on_sel_extend { |
| 39 | my ($self) = @_; |
60 | my ($self, $time) = @_; |
| 40 | |
61 | |
| 41 | my ($row, $col) = $self->selection_mark; |
62 | my ($row, $col) = $self->selection_mark; |
| 42 | my $line = $self->line ($row); |
63 | my $line = $self->line ($row); |
| 43 | my $offset = $line->offset_of ($row, $col); |
|
|
| 44 | my $text = $line->t; |
64 | my $text = $line->t; |
|
|
65 | my $markofs = $line->offset_of ($row, $col); |
|
|
66 | my $curlen = $line->offset_of ($self->selection_end) |
|
|
67 | - $line->offset_of ($self->selection_beg); |
| 45 | |
68 | |
| 46 | for my $regex (@mark_patterns) { |
69 | my @matches; |
|
|
70 | |
|
|
71 | # not doing matches in unicode mode helps speed |
|
|
72 | # enourmously here. working in utf-8 should be |
|
|
73 | # equivalent due to the magic of utf-8 encoding. |
|
|
74 | utf8::encode $text; |
|
|
75 | study $text; # _really_ helps, too :) |
|
|
76 | |
|
|
77 | for my $regex (@mark_patterns, @{ $self->{patterns} }) { |
| 47 | while ($text =~ /$regex/g) { |
78 | while ($text =~ /$regex/g) { |
| 48 | if ($-[1] <= $offset and $offset <= $+[1]) { |
79 | if ($-[1] <= $markofs and $markofs <= $+[1]) { |
|
|
80 | my $ofs = $-[1]; |
| 49 | my $match = $1; |
81 | my $match = $1; |
| 50 | my ($ofs1, $ofs2) = ($-[1], $+[1]); |
|
|
| 51 | |
82 | |
| 52 | for my $regex (@simplify_patterns) { |
83 | for my $regex (@simplify_patterns) { |
| 53 | if ($match =~ $regex) { |
84 | if ($match =~ $regex) { |
| 54 | $match = $1; |
85 | $match = $1; |
| 55 | $ofs1 += $-[1]; |
86 | $ofs += $-[1]; |
| 56 | $ofs2 = $ofs1 + length $match; |
|
|
| 57 | } |
87 | } |
| 58 | } |
88 | } |
| 59 | |
89 | |
| 60 | $self->selection_beg ($line->coord_of ($ofs1)); |
90 | push @matches, [$ofs, length $match]; |
| 61 | $self->selection_end ($line->coord_of ($ofs2)); |
|
|
| 62 | return 1; |
|
|
| 63 | } |
91 | } |
| 64 | } |
92 | } |
| 65 | } |
93 | } |
| 66 | |
94 | |
|
|
95 | # whole line |
|
|
96 | push @matches, [0, ($line->end - $line->beg + 1) * $self->ncol]; |
|
|
97 | |
|
|
98 | for (sort { $a->[1] <=> $b->[1] or $b->[0] <=> $a->[0] } @matches) { |
|
|
99 | my ($ofs, $len) = @$_; |
|
|
100 | |
|
|
101 | next if $len <= $curlen; |
|
|
102 | |
|
|
103 | $self->selection_beg ($line->coord_of ($ofs)); |
|
|
104 | $self->selection_end ($line->coord_of ($ofs + $len)); |
|
|
105 | return 1; |
|
|
106 | } |
|
|
107 | |
| 67 | () |
108 | () |
| 68 | } |
109 | } |
