lita-markov 1.1.0 → 1.1.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 34357a93d6cd917570538738c72f1bc0d301b78b
-  data.tar.gz: 4fa6a56d9c25baf00d141bb9d593bc11bf3fc469
+  metadata.gz: 899ea3cf061e0e9e60046fa630ecc85ac0c62d3c
+  data.tar.gz: c9c47683d42548b71f6b9adcf17130786d0e7ae3
 SHA512:
-  metadata.gz: a139b7b1e4618b420addacdaf0732885d02defcbc0c741186c892efb85b7d9dba0ea230b6dffd544903cac599f257f73624158340cf7baf10cf1cb745f84189a
-  data.tar.gz: 7737301aa793730960cc142892586d64d6a8e5310760fe4ad868d792610d04240ba253f2bd69c24148c3c5be2cd842efa4d9bce7fbef8a698c111c3aac2c682a
+  metadata.gz: f52a9534465220eadd3b53cee38d98169f04640696deb168a4c7dcc549b3f19587c6d508046c950b99438894d962f527a8c4f8affbf36bdd8b58b92ecb4e4548
+  data.tar.gz: ee9904cfdaef57ca3757407511e90763fac96173b0897973cd6616a21915b3a8a41da460f56fd9767c85e35f56069d4c77774a7afa1ac400be8274393470d31a

data/lib/lita/handlers/markov/engine.rb

@@ -36,8 +36,8 @@ class Lita::Handlers::Markov
 
       return if words.length == 0
 
-      # Capitalize the first word
-      words = [words[0].capitalize] + words.slice(1..-1)
+      # Capitalize the first word and add a period at the end
+      words = [words[0].capitalize] + words.slice(1..-1) + ['.']
 
       # Iterate over it one step at a time in sets of `@depth + 1`
       words.each_cons(@depth + 1) do |words|
@@ -99,10 +99,6 @@ class Lita::Handlers::Markov
       state.split(' ').last
     end
 
-    def is_punctuation?(string)
-      PUNCTUATION.any? { |p| string == p }
-    end
-
     def get_next_state(user, current_state)
       states = @db[:dictionary]
         .where(user: user, current_state: current_state)
@@ -133,7 +129,7 @@ class Lita::Handlers::Markov
 
         sentence << next_state
 
-        if is_punctuation? next_state
+        if next_state == '.'
           ended_with_punctuation = true
           break
         end
@@ -146,7 +142,7 @@ class Lita::Handlers::Markov
       chain
     end
 
-    STRING_SEPARATOR = /([.!?])|\s+/
+    STRING_SEPARATOR = /\s+/
 
     def separate_string string
       # Including the punctuation in group so they'll be included in the
@@ -157,33 +153,22 @@ class Lita::Handlers::Markov
         .select { |w| !w.empty? }
     end
 
-    PUNCTUATION = [',', '.', '!', '?']
-
     # Don't allow anything besides letters, digits, whitespace, and puncutation
-    ILLEGAL_CHARACTERS = /[^\w\d\s:;,.!?#@]/
+    NON_WORD_CHARACTERS = /[^\w\d'"“”’:+-]/
 
     HYPERLINKS          = /http[^\s]+/
     SIMPLE_CODE_BLOCK   = /`[^`]+`/
     EXTENDED_CODE_BLOCK = /```.+```/m
-
-    REPEATED_PUNCTUATION = /([.!?])[.!?]+/
-    BASIC_PUNCTUATION    = /([;,.!?])/
-
+    REPEATED_WHITESPACE = /\s+/
 
     def sanitize_string string
       string = string
         .gsub(HYPERLINKS, ''.freeze)             # Remove any hyperlinks
         .gsub(SIMPLE_CODE_BLOCK, ''.freeze)      # Remove code blocks and illegal characters
         .gsub(EXTENDED_CODE_BLOCK, ''.freeze)
-        .gsub(ILLEGAL_CHARACTERS, ''.freeze)
-        .gsub(REPEATED_PUNCTUATION, '\1'.freeze) # Trim down repeated punctuation
-        .gsub(BASIC_PUNCTUATION, '\1 '.freeze)   # Put whitespace after punctuation for proper separation
+        .gsub(NON_WORD_CHARACTERS, ' '.freeze)   # Convert non-word characters into whitespace
+        .gsub(REPEATED_WHITESPACE, ' '.freeze)   # Convert repeated whitespace into just single spaces
         .strip()
-
-      ends_with_punctuation = PUNCTUATION.any? { |p| string.end_with? p }
-      string = string+'.'.freeze unless ends_with_punctuation
-
-      string
     end
   end
 end

data/lita-markov.gemspec

@@ -1,6 +1,6 @@
 Gem::Specification.new do |spec|
   spec.name          = "lita-markov"
-  spec.version       = "1.1.0"
+  spec.version       = "1.1.1"
   spec.authors       = ["Dirk Gadsden"]
   spec.email         = ["dirk@dirk.to"]
   spec.description   = "Markov chains for Lita."

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: lita-markov
 version: !ruby/object:Gem::Version
-  version: 1.1.0
+  version: 1.1.1
 platform: ruby
 authors:
 - Dirk Gadsden