| … | |
… | |
| 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 | $res = $self->locale_decode ($res); |
|
|
17 | utf8::encode $res; |
|
|
18 | push @{ $self->{patterns} }, qr/$res/; |
|
|
19 | } |
|
|
20 | |
|
|
21 | () |
|
|
22 | } |
|
|
23 | |
| 12 | # "find interetsing things"-patterns |
24 | # "find interesting things"-patterns |
| 13 | 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 | |
| 14 | # urls, just a heuristic |
36 | # urls, just a heuristic |
| 15 | qr{( |
37 | qr{( |
| 16 | (?:https?|ftp|news|mailto|file)://[ab-zA-Z0-9\-\@;\/?:&=%\$_.+!*\x27(),~]+ |
38 | (?:https?|ftp|news|mailto|file)://[ab-zA-Z0-9\-\@;\/?:&=%\$_.+!*\x27(),~]+ |
| 17 | [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) |
| 18 | )}x, |
40 | )}x, |
| 19 | |
41 | |
| 20 | # common "parentheses" |
|
|
| 21 | qr{(?:^|\s) ‘ ([^‘’]+?) ’ (?:\s|\)|$)}x, |
|
|
| 22 | qr{(?:^|\s) ` ([^`']+?) ' (?:\s|\)|$)}x, |
|
|
| 23 | qr{ \{ ([^{}]+?) \} }x, |
|
|
| 24 | qr{ \[ ([^{}]+?) \] }x, |
|
|
| 25 | |
|
|
| 26 | # shell-like argument quoting, basically always matches |
42 | # shell-like argument quoting, basically always matches |
| 27 | qr{\G [\ \t|&;<>()] *( |
43 | qr{\G [\ \t|&;<>()]* ( |
| 28 | (?: |
44 | (?: |
| 29 | [^\\"'\ \t|&;<>()]+ |
45 | [^\\"'\ \t|&;<>()]+ |
| 30 | | \\. |
46 | | \\. |
| 31 | | " (?: [^\\"]+ | \\. )* " |
47 | | " (?: [^\\"]+ | \\. )* " |
| 32 | | ' [^']* ' |
48 | | ' [^']* ' |
| … | |
… | |
| 39 | qr{^"([^\\"'\ \t|&;<>()*?]+)"$}, # "simple" => simple |
55 | qr{^"([^\\"'\ \t|&;<>()*?]+)"$}, # "simple" => simple |
| 40 | qr{^(.*)[,\-]$}, # strip off trailing , and - |
56 | qr{^(.*)[,\-]$}, # strip off trailing , and - |
| 41 | ); |
57 | ); |
| 42 | |
58 | |
| 43 | sub on_sel_extend { |
59 | sub on_sel_extend { |
| 44 | my ($self) = @_; |
60 | my ($self, $time) = @_; |
| 45 | |
61 | |
| 46 | my ($row, $col) = $self->selection_mark; |
62 | my ($row, $col) = $self->selection_mark; |
| 47 | my $line = $self->line ($row); |
63 | my $line = $self->line ($row); |
| 48 | my $offset = $line->offset_of ($row, $col); |
|
|
| 49 | 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); |
| 50 | |
68 | |
| 51 | for my $regex (@mark_patterns) { |
69 | my @matches; |
| 52 | while ($text =~ /$regex/g) { |
|
|
| 53 | if ($-[1] <= $offset and $offset <= $+[1]) { |
|
|
| 54 | my $ofs = $-[1]; |
|
|
| 55 | my $match = $1; |
|
|
| 56 | |
70 | |
|
|
71 | if ($markofs < $line->l) { |
|
|
72 | # convert markofs form character to UTF-8 offset space |
|
|
73 | { |
|
|
74 | my $prefix = substr $text, 0, $markofs; |
|
|
75 | utf8::encode $prefix; |
|
|
76 | $markofs = length $prefix; |
|
|
77 | } |
|
|
78 | |
|
|
79 | # not doing matches in unicode mode helps speed |
|
|
80 | # enourmously here. working in utf-8 should be |
|
|
81 | # equivalent due to the magic of utf-8 encoding. |
|
|
82 | utf8::encode $text; |
|
|
83 | study $text; # _really_ helps, too :) |
|
|
84 | |
|
|
85 | for my $regex (@mark_patterns, @{ $self->{patterns} }) { |
|
|
86 | while ($text =~ /$regex/g) { |
|
|
87 | if ($-[1] <= $markofs and $markofs <= $+[1]) { |
|
|
88 | my $ofs = $-[1]; |
|
|
89 | my $match = $1; |
|
|
90 | |
| 57 | for my $regex (@simplify_patterns) { |
91 | for my $regex (@simplify_patterns) { |
| 58 | if ($match =~ $regex) { |
92 | if ($match =~ $regex) { |
| 59 | $match = $1; |
93 | $match = $1; |
| 60 | $ofs += $-[1]; |
94 | $ofs += $-[1]; |
|
|
95 | } |
| 61 | } |
96 | } |
|
|
97 | |
|
|
98 | push @matches, [$ofs, length $match]; |
| 62 | } |
99 | } |
| 63 | |
|
|
| 64 | $self->selection_beg ($line->coord_of ($ofs)); |
|
|
| 65 | $self->selection_end ($line->coord_of ($ofs + length $match)); |
|
|
| 66 | return 1; |
|
|
| 67 | } |
100 | } |
| 68 | } |
101 | } |
| 69 | } |
102 | } |
| 70 | |
103 | |
|
|
104 | # whole line |
|
|
105 | push @matches, [0, ($line->end - $line->beg + 1) * $self->ncol]; |
|
|
106 | |
|
|
107 | for (sort { $a->[1] <=> $b->[1] or $b->[0] <=> $a->[0] } @matches) { |
|
|
108 | my ($ofs, $len) = @$_; |
|
|
109 | |
|
|
110 | next if $len <= $curlen; |
|
|
111 | |
|
|
112 | # convert back from UTF-8 offset space to character space |
|
|
113 | { |
|
|
114 | my $length = substr "$text ", $ofs, $len; |
|
|
115 | utf8::decode $length; |
|
|
116 | $len = length $length; |
|
|
117 | } |
|
|
118 | { |
|
|
119 | my $prefix = substr $text, 0, $ofs; |
|
|
120 | utf8::decode $prefix; |
|
|
121 | $ofs = length $prefix; |
|
|
122 | } |
|
|
123 | |
|
|
124 | $self->selection_beg ($line->coord_of ($ofs)); |
|
|
125 | $self->selection_end ($line->coord_of ($ofs + $len)); |
|
|
126 | return 1; |
|
|
127 | } |
|
|
128 | |
| 71 | () |
129 | () |
| 72 | } |
130 | } |
