| 1 | #! perl |
1 | #! perl |
|
|
2 | |
|
|
3 | #:META:X_RESOURCE:%.pattern-0:string:first selection pattern |
|
|
4 | |
|
|
5 | =head1 NAME |
|
|
6 | |
|
|
7 | selection - more intelligent selection (enabled by default) |
|
|
8 | |
|
|
9 | =head1 DESCRIPTION |
|
|
10 | |
|
|
11 | This extension tries to be more intelligent when the user extends |
|
|
12 | selections (double-click and further clicks). Right now, it tries to |
|
|
13 | select words, urls and complete shell-quoted arguments, which is very |
|
|
14 | convenient, too, if your F<ls> supports C<--quoting-style=shell>. |
|
|
15 | |
|
|
16 | A double-click usually selects the word under the cursor, further clicks |
|
|
17 | will enlarge the selection. |
|
|
18 | |
|
|
19 | The selection works by trying to match a number of regexes and displaying |
|
|
20 | them in increasing order of length. You can add your own regexes by |
|
|
21 | specifying resources of the form: |
|
|
22 | |
|
|
23 | URxvt.selection.pattern-0: perl-regex |
|
|
24 | URxvt.selection.pattern-1: perl-regex |
|
|
25 | ... |
|
|
26 | |
|
|
27 | The index number (0, 1...) must not have any holes, and each regex must |
|
|
28 | contain at least one pair of capturing parentheses, which will be used for |
|
|
29 | the match. For example, the following adds a regex that matches everything |
|
|
30 | between two vertical bars: |
|
|
31 | |
|
|
32 | URxvt.selection.pattern-0: \\|([^|]+)\\| |
|
|
33 | |
|
|
34 | Another example: Programs I use often output "absolute path: " at the |
|
|
35 | beginning of a line when they process multiple files. The following |
|
|
36 | pattern matches the filename (note, there is a single space at the very |
|
|
37 | end): |
|
|
38 | |
|
|
39 | URxvt.selection.pattern-0: ^(/[^:]+):\ |
|
|
40 | |
|
|
41 | You can look at the source of the selection extension to see more |
|
|
42 | interesting uses, such as parsing a line from beginning to end. |
|
|
43 | |
|
|
44 | This extension also offers following bindable keyboard commands: |
|
|
45 | |
|
|
46 | =over 4 |
|
|
47 | |
|
|
48 | =item rot13 |
|
|
49 | |
|
|
50 | Rot-13 the selection when activated. Used via keyboard trigger: |
|
|
51 | |
|
|
52 | URxvt.keysym.C-M-r: perl:selection:rot13 |
|
|
53 | |
|
|
54 | =back |
|
|
55 | |
|
|
56 | =cut |
| 2 | |
57 | |
| 3 | sub on_user_command { |
58 | sub on_user_command { |
| 4 | my ($self, $cmd) = @_; |
59 | my ($self, $cmd) = @_; |
| 5 | |
60 | |
| 6 | $cmd eq "selection:rot13" |
61 | $cmd eq "selection:rot13" |
| … | |
… | |
| 17 | push @{ $self->{patterns} }, qr{\G [\Q$res\E[:space:]]* ([^\Q$res\E[:space:]]+) }x; |
72 | push @{ $self->{patterns} }, qr{\G [\Q$res\E[:space:]]* ([^\Q$res\E[:space:]]+) }x; |
| 18 | } |
73 | } |
| 19 | |
74 | |
| 20 | for (my $idx = 0; defined (my $res = $self->x_resource ("selection.pattern-$idx")); $idx++) { |
75 | for (my $idx = 0; defined (my $res = $self->x_resource ("selection.pattern-$idx")); $idx++) { |
| 21 | $res = $self->locale_decode ($res); |
76 | $res = $self->locale_decode ($res); |
| 22 | utf8::encode $res; |
|
|
| 23 | push @{ $self->{patterns} }, qr/$res/; |
77 | push @{ $self->{patterns} }, qr/$res/; |
| 24 | } |
78 | } |
| 25 | |
79 | |
| 26 | $self->{enabled} = 1; |
80 | $self->{enabled} = 1; |
| 27 | |
81 | |
| … | |
… | |
| 32 | () |
86 | () |
| 33 | } |
87 | } |
| 34 | |
88 | |
| 35 | # "find interesting things"-patterns |
89 | # "find interesting things"-patterns |
| 36 | my @mark_patterns = ( |
90 | my @mark_patterns = ( |
|
|
91 | # qr{ ([[:word:]]+) }x, |
|
|
92 | qr{ ([^[:space:]]+) }x, |
|
|
93 | |
| 37 | # common types of "parentheses" |
94 | # common types of "parentheses" |
|
|
95 | qr{ (?<![^[:space:]]) [`'] ([^`']+) [`'] (?![^[:space:]]) }x, |
| 38 | qr{ (?<![^[:space:]]) ‘ ([^‘’]+) ’ (?![^[:space]]) }x, |
96 | qr{ (?<![^[:space:]]) ‘ ([^‘’]+) ’ (?![^[:space:]]) }x, |
| 39 | qr{ (?<![^[:space:]]) ` ([^`']+) ' (?![^[:space]]) }x, |
97 | qr{ (?<![^[:space:]]) “ ([^“”]+) ” (?![^[:space:]]) }x, |
|
|
98 | |
|
|
99 | qr{ (?<![^[:space:]]) (' [^[:space:]] [^']* ') }x, |
|
|
100 | qr{ (' [^']* [^[:space:]] ') (?![^[:space:]]) }x, |
|
|
101 | qr{ (?<![^[:space:]]) (` [^[:space:]] [^']* ') }x, |
|
|
102 | qr{ (` [^']* [^[:space:]] ') (?![^[:space:]]) }x, |
| 40 | qr{ (?<![^[:space:]]) (" [^[:space:]] [^"]* ") }x, |
103 | qr{ (?<![^[:space:]]) (" [^[:space:]] [^"]* ") }x, |
| 41 | qr{ (" [^"]* [^[:space:]] ") (?![^[:space]]) }x, |
104 | qr{ (" [^"]* [^[:space:]] ") (?![^[:space:]]) }x, |
| 42 | qr{ \< ([^<>[:space:]]+) \> }x, |
105 | |
| 43 | qr{ \{ ([^{}[:space:]]+) \} }x, |
106 | qr{ \{ ([^\{\}]+) \} }x, |
| 44 | qr{ \[ ([^{}[:space:]]+) \] }x, |
|
|
| 45 | qr{ \( ([^()[:space:]]+) \) }x, |
107 | qr{ \( ([^\(\)]+) \) }x, |
|
|
108 | qr{ \[ ([^\[\]]+) \] }x, |
|
|
109 | qr{ \< ([^\<\>]+) \> }x, |
| 46 | |
110 | |
| 47 | # urls, just a heuristic |
111 | # urls, just a heuristic |
| 48 | qr{( |
112 | qr{( |
| 49 | (?:https?://|ftp://|news://|mailto:|file://)[ab-zA-Z0-9\-\@;\/?:&=%\$_.+!*\x27(),~#]+ |
113 | (?:https?://|ftp://|news://|mailto:|file://|\bwww\.)[ab-zA-Z0-9\-\@;\/?:&=%\$_.+!*\x27(),~#]+ |
| 50 | [ab-zA-Z0-9\-\@;\/?:&=%\$_+!*\x27()~] # exclude some trailing characters (heuristic) |
114 | [ab-zA-Z0-9\-\@;\/?:&=%\$_+*()~] # exclude some trailing characters (heuristic) |
| 51 | )}x, |
115 | )}x, |
| 52 | |
116 | |
| 53 | # shell-like argument quoting, basically always matches |
117 | # shell-like argument quoting, basically always matches |
| 54 | qr{\G [\ \t|&;<>()]* ( |
118 | qr{\G [\ \t|&;<>()]* ( |
| 55 | (?: |
119 | (?: |
| … | |
… | |
| 81 | - $line->offset_of ($self->selection_beg); |
145 | - $line->offset_of ($self->selection_beg); |
| 82 | |
146 | |
| 83 | my @matches; |
147 | my @matches; |
| 84 | |
148 | |
| 85 | if ($markofs < $line->l) { |
149 | if ($markofs < $line->l) { |
| 86 | # convert markofs form character to UTF-8 offset space |
|
|
| 87 | { |
|
|
| 88 | my $prefix = substr $text, 0, $markofs; |
|
|
| 89 | utf8::encode $prefix; |
|
|
| 90 | $markofs = length $prefix; |
|
|
| 91 | } |
|
|
| 92 | |
|
|
| 93 | # not doing matches in unicode mode helps speed |
|
|
| 94 | # enourmously here. working in utf-8 should be |
|
|
| 95 | # equivalent due to the magic of utf-8 encoding. |
|
|
| 96 | utf8::encode $text; |
|
|
| 97 | study $text; # _really_ helps, too :) |
150 | study $text; # _really_ helps, too :) |
| 98 | |
151 | |
| 99 | for my $regex (@mark_patterns, @{ $self->{patterns} }) { |
152 | for my $regex (@mark_patterns, @{ $self->{patterns} }) { |
| 100 | while ($text =~ /$regex/g) { |
153 | while ($text =~ /$regex/g) { |
| 101 | if ($-[1] <= $markofs and $markofs <= $+[1]) { |
154 | if ($-[1] <= $markofs and $markofs <= $+[1]) { |
| … | |
… | |
| 121 | for (sort { $a->[1] <=> $b->[1] or $b->[0] <=> $a->[0] } @matches) { |
174 | for (sort { $a->[1] <=> $b->[1] or $b->[0] <=> $a->[0] } @matches) { |
| 122 | my ($ofs, $len) = @$_; |
175 | my ($ofs, $len) = @$_; |
| 123 | |
176 | |
| 124 | next if $len <= $curlen; |
177 | next if $len <= $curlen; |
| 125 | |
178 | |
| 126 | # convert back from UTF-8 offset space to character space |
|
|
| 127 | { |
|
|
| 128 | my $length = substr "$text ", $ofs, $len; |
|
|
| 129 | utf8::decode $length; |
|
|
| 130 | $len = length $length; |
|
|
| 131 | } |
|
|
| 132 | { |
|
|
| 133 | my $prefix = substr $text, 0, $ofs; |
|
|
| 134 | utf8::decode $prefix; |
|
|
| 135 | $ofs = length $prefix; |
|
|
| 136 | } |
|
|
| 137 | |
|
|
| 138 | $self->selection_beg ($line->coord_of ($ofs)); |
179 | $self->selection_beg ($line->coord_of ($ofs)); |
| 139 | $self->selection_end ($line->coord_of ($ofs + $len)); |
180 | $self->selection_end ($line->coord_of ($ofs + $len)); |
| 140 | return 1; |
181 | return 1; |
| 141 | } |
182 | } |
| 142 | |
183 | |
