… | |
… | |
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 = $self->locale_decode ($res); |
|
|
18 | utf8::encode $res; |
|
|
19 | push @{ $self->{patterns} }, qr/$res/; |
|
|
20 | } |
|
|
21 | |
|
|
22 | () |
|
|
23 | } |
|
|
24 | |
|
|
25 | # "find interetsing things"-patterns |
12 | my @mark_patterns = ( |
26 | my @mark_patterns = ( |
|
|
27 | # common types of "parentheses" |
|
|
28 | qr{ (?<![^[:space:]]) ‘ ([^‘’]+) ’ (?![^[:space]]) }x, |
|
|
29 | qr{ (?<![^[:space:]]) ` ([^`']+) ' (?![^[:space]]) }x, |
|
|
30 | qr{ (?<![^[:space:]]) (" [^[:space:]] [^"]* ") }x, |
|
|
31 | qr{ (" [^"]* [^[:space:]] ") (?![^[:space]]) }x, |
|
|
32 | qr{ \< ([^<>[:space:]]+) \> }x, |
|
|
33 | qr{ \{ ([^{}[:space:]]+) \} }x, |
|
|
34 | qr{ \[ ([^{}[:space:]]+) \] }x, |
|
|
35 | qr{ \( ([^()[:space:]]+) \) }x, |
|
|
36 | |
13 | # urls, just a heuristic |
37 | # urls, just a heuristic |
14 | qr{( |
38 | qr{( |
15 | (?:https?|ftp|news|mailto|file)://[ab-zA-Z0-9\-\@;\/?:&=%\$_.+!*\x27(),~]+ |
39 | (?: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 |
40 | [ab-zA-Z0-9\-\@;\/?:&=%\$_+!*\x27()~] # exclude some trailing characters (heuristic) |
17 | )}x, |
41 | )}x, |
18 | |
42 | |
19 | # common forms of quoting |
43 | # shell-like argument quoting, basically always matches |
20 | qr{(?:^|\s) [‘`] ([^‘`’']+) [’'] (?:\s|$)}x, |
|
|
21 | |
|
|
22 | # shell-like argument quoting |
|
|
23 | qr{\G [\ \t|&;<>()] *( |
44 | qr{\G [\ \t|&;<>()] *( |
24 | (?: |
45 | (?: |
25 | [^\\"'\ \t|&;<>()]+ |
46 | [^\\"'\ \t|&;<>()]+ |
26 | | \\. |
47 | | \\. |
27 | | " (?: [^\\"]+ | \\. )* " |
48 | | " (?: [^\\"]+ | \\. )* " |
28 | | ' [^']* ' |
49 | | ' [^']* ' |
29 | )+ |
50 | )+ |
30 | )}x, |
51 | )}x, |
31 | ); |
52 | ); |
32 | |
53 | |
|
|
54 | # "correct obvious? crap"-patterns |
33 | my @simplify_patterns = ( |
55 | my @simplify_patterns = ( |
34 | qr{^"([^\\"'\ \t|&;<>()*?]+)"$}, # "simple" => simple |
56 | qr{^"([^\\"'\ \t|&;<>()*?]+)"$}, # "simple" => simple |
35 | qr{^(.*)[,\-]$}, |
57 | qr{^(.*)[,\-]$}, # strip off trailing , and - |
36 | ); |
58 | ); |
37 | |
59 | |
38 | sub on_sel_extend { |
60 | sub on_sel_extend { |
39 | my ($self) = @_; |
61 | my ($self, $time) = @_; |
40 | |
62 | |
41 | my ($row, $col) = $self->selection_mark; |
63 | my ($row, $col) = $self->selection_mark; |
42 | my $line = $self->line ($row); |
64 | my $line = $self->line ($row); |
43 | my $offset = $line->offset_of ($row, $col); |
|
|
44 | my $text = $line->t; |
65 | my $text = $line->t; |
|
|
66 | my $markofs = $line->offset_of ($row, $col); |
|
|
67 | my $curlen = $line->offset_of ($self->selection_end) |
|
|
68 | - $line->offset_of ($self->selection_beg); |
45 | |
69 | |
46 | for my $regex (@mark_patterns) { |
70 | my @matches; |
47 | while ($text =~ /$regex/g) { |
|
|
48 | if ($-[1] <= $offset and $offset <= $+[1]) { |
|
|
49 | my $match = $1; |
|
|
50 | my ($ofs1, $ofs2) = ($-[1], $+[1]); |
|
|
51 | |
71 | |
|
|
72 | if ($markofs < $line->l) { |
|
|
73 | # convert markofs form character to UTF-8 offset space |
|
|
74 | { |
|
|
75 | my $prefix = substr $text, 0, $markofs; |
|
|
76 | utf8::encode $prefix; |
|
|
77 | $markofs = length $prefix; |
|
|
78 | } |
|
|
79 | |
|
|
80 | # not doing matches in unicode mode helps speed |
|
|
81 | # enourmously here. working in utf-8 should be |
|
|
82 | # equivalent due to the magic of utf-8 encoding. |
|
|
83 | utf8::encode $text; |
|
|
84 | study $text; # _really_ helps, too :) |
|
|
85 | |
|
|
86 | for my $regex (@mark_patterns, @{ $self->{patterns} }) { |
|
|
87 | while ($text =~ /$regex/g) { |
|
|
88 | if ($-[1] <= $markofs and $markofs <= $+[1]) { |
|
|
89 | my $ofs = $-[1]; |
|
|
90 | my $match = $1; |
|
|
91 | |
52 | for my $regex (@simplify_patterns) { |
92 | for my $regex (@simplify_patterns) { |
53 | if ($match =~ $regex) { |
93 | if ($match =~ $regex) { |
54 | $match = $1; |
94 | $match = $1; |
55 | $ofs1 += $-[1]; |
95 | $ofs += $-[1]; |
56 | $ofs2 = $ofs1 + length $match; |
96 | } |
57 | } |
97 | } |
|
|
98 | |
|
|
99 | push @matches, [$ofs, length $match]; |
58 | } |
100 | } |
59 | |
|
|
60 | $self->selection_beg ($line->coord_of ($ofs1)); |
|
|
61 | $self->selection_end ($line->coord_of ($ofs2)); |
|
|
62 | return 1; |
|
|
63 | } |
101 | } |
64 | } |
102 | } |
65 | } |
103 | } |
66 | |
104 | |
|
|
105 | # whole line |
|
|
106 | push @matches, [0, ($line->end - $line->beg + 1) * $self->ncol]; |
|
|
107 | |
|
|
108 | for (sort { $a->[1] <=> $b->[1] or $b->[0] <=> $a->[0] } @matches) { |
|
|
109 | my ($ofs, $len) = @$_; |
|
|
110 | |
|
|
111 | next if $len <= $curlen; |
|
|
112 | |
|
|
113 | # convert back from UTF-8 offset space to character space |
|
|
114 | { |
|
|
115 | my $length = substr $text, $ofs, $len; |
|
|
116 | utf8::decode $length; |
|
|
117 | $len = length $length; |
|
|
118 | } |
|
|
119 | { |
|
|
120 | my $prefix = substr $text, 0, $ofs; |
|
|
121 | utf8::decode $prefix; |
|
|
122 | $ofs = length $prefix; |
|
|
123 | } |
|
|
124 | |
|
|
125 | $self->selection_beg ($line->coord_of ($ofs)); |
|
|
126 | $self->selection_end ($line->coord_of ($ofs + $len)); |
|
|
127 | return 1; |
|
|
128 | } |
|
|
129 | |
67 | () |
130 | () |
68 | } |
131 | } |