Created
March 18, 2012 16:57
-
-
Save simonwistow/2077513 to your computer and use it in GitHub Desktop.
Infobot.pm.diff
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
--- Infobot.pm.orig 2012-03-18 09:13:26.000000000 -0700 | |
+++ Infobot.pm 2012-03-18 09:29:08.000000000 -0700 | |
@@ -14,6 +14,58 @@ | |
eval { require XML::Feed }; | |
our $HAS_XML_FEED = $@ ? 0 : 1; | |
+our %AUXILLARY_VERBS = ( | |
+ "am" => { tense => 'present' }, | |
+ "are" => { tense => 'present' }, | |
+ "aren't" => { tense => 'present', negative_of => "are" }, | |
+ "ain't" => { tense => 'present', negative_of => "are" }, | |
+ "can" => { tense => 'present' }, | |
+ "can't" => { tense => 'present', negative_of => "can" }, | |
+ "could" => { tense => 'past' }, | |
+ "couldn't" => { tense => 'past', negative_of => "could" }, | |
+ "dare" => { tense => 'present' }, | |
+ "did" => { tense => 'past' }, | |
+ "didn't" => { tense => 'past', negative_of => "did" }, | |
+ "do" => { tense => 'future' }, | |
+ "don't" => { tense => 'future', negative_of => "do" }, | |
+ "does" => { tense => 'present' }, | |
+ "doesn't" => { tense => 'present', negative_of => "does" }, | |
+ "had" => { tense => 'past' }, | |
+ "hadn't" => { tense => 'past', negative_of => "had" }, | |
+ "has" => { tense => 'present' }, | |
+ "hasn't" => { tense => 'present', negative_of => "has" }, | |
+ "have" => { tense => 'past' }, | |
+ "haven't" => { tense => 'past', negative_of => "have" }, | |
+ "is" => { tense => 'present' }, | |
+ "isn't" => { tense => 'present', negative_of => "is" }, | |
+ "may" => { tense => 'future' }, | |
+ "mayn't" => { tense => 'future', negative_of => "may" }, | |
+ "might" => { tense => 'future' }, | |
+ "mightn't" => { tense => 'future', negative_of => "might" }, | |
+ "must" => { tense => 'future' }, | |
+ "mustn't" => { tense => 'future', negative_of => "must" }, | |
+ "need" => { tense => 'future' }, | |
+ "needn't" => { tense => 'future', negative_of => "need" }, | |
+ "ought" => { tense => 'future' }, | |
+ "oughtn't" => { tense => 'future', negative_of => "ought" }, | |
+ "shall" => { tense => 'future' }, | |
+ "shan't" => { tense => 'future', negative_of => "shall" }, | |
+ "should" => { tense => 'future' }, | |
+ "shouldn't" => { tense => 'future', negative_of => "should" }, | |
+ "used to" => { tense => 'past' }, | |
+ "was" => { tense => 'past' }, | |
+ "wasn't" => { tense => 'past', negative_of => "was" }, | |
+ "were" => { tense => 'past' }, | |
+ "weren't" => { tense => 'past', negative_of => "were" }, | |
+ "will" => { tense => 'future' }, | |
+ "won't" => { tense => 'future', negative_of => "will" }, | |
+ "would" => { tense => 'future' }, | |
+ "wouldn't" => { tense => 'future', negative_of => "would" }, | |
+); | |
+ | |
+our $VERB_RE = "(".join("|", keys %AUXILLARY_VERBS).")" | |
+ | |
+ | |
sub init { | |
my $self = shift; | |
$self->config( | |
@@ -106,8 +158,8 @@ | |
# <word> (is|are) | |
# and then removes if if <word> is a stopword | |
# this means that we treat "what is foo?" as "foo?" | |
- if ( $body =~ /^(.*?)\s+(is|are)\s+(.*)$/i ) { | |
- $body =~ s/^(.*?)\s+(is|are)\s+//i if $stopwords{$1}; | |
+ if ( $body =~ /^(.*?)\s+($VERB_RE)\s+(.*)$/i ) { | |
+ $body =~ s/^(.*?)\s+($VERB_RE)\s+//i if $stopwords{$1}; | |
} | |
# answer a factoid. this is a crazy check which ensures we will ONLY answer | |
@@ -170,8 +222,7 @@ | |
# does it even look like a factoid? | |
return unless ( $mess->{address} or $self->get("user_passive_learn") ); | |
return | |
- unless ( $body =~ /^(.*?)\s+(is)\s+(.*)$/i | |
- or $body =~ /^(.*?)\s+(are)\s+(.*)$/i ); | |
+ unless ( $body =~ /^(.*?)\s+($VERB_RE)\s+(.*)$/i | |
my ( $object, $is_are, $description ) = ( $1, $2, $3 ); | |
my $literal = ( $object =~ s!^literal\s+!! ); | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment