text_rank 1.1.7 → 1.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 1625933f78441107094a4306a488cf0d68f45e53
-  data.tar.gz: 5e605f2ac2210c6e44e21a1c088620c5d491314c
+  metadata.gz: c7c140fafc459b538cdc9fcc84f639e2155c6fbb
+  data.tar.gz: c2e8b24f80414a113ba9d2d93ca26b6a7e1c38a8
 SHA512:
-  metadata.gz: 1703773a7d6d1391ec81f26bac6aa409344d8f036352b24e0acf37511e141da88cc560c0bbc525563f5b2bd277ea528c9d30bb41f9768afb1987ae9acee42af6
-  data.tar.gz: fc23bc9f6d63a0361d88d804bc64448613de5aa6c22d1f5b223630bee18c485d3d35b7a12876bb5b4c25dda5efa537aad83ed2e686e12a04a4fed33d88e2147d
+  metadata.gz: ab7b1875f82d42a51243f74b827f90dc0e23b3ad68479b7dca4a78058423bf63e74d50361c952809e1396ddc70423fbafcee090344bbb47cd5095f94cd09e435
+  data.tar.gz: 7fb37752476eb9f0fefab815af32522eb3a748bb39fdeace39af08446c9a6e544c96fe187bcf541c27947493d6c7aafd69351af1c6a6438f987926396189adde

data/lib/text_rank/fingerprint.rb

@@ -10,35 +10,35 @@ module TextRank
   # significant keywords.  But to prevent less significant keywords from being
   # completely ignored we apply an inverse log linear transformation to each of the
   # N prefixes.
-  # 
+  #
   # For example, consider the following comparison:
-  # 
+  #
   #   town man empty found
   #   vs.
   #   general empty found jar
-  # 
+  #
   # The first pass considers just the first keywords: town vs. general.  As these
   # are different, they contribute 0.
-  # 
+  #
   # The second pass considers the first two keywords: town man vs general empty.
   # Again, no overlap, so they contribute 0.
-  # 
+  #
   # The third pass considers the first three keywords: town man empty vs general
   # empty found.  Here we have one overlap: empty. This contributes 1.
-  # 
+  #
   # The fourth pass considers all, and there is two overlaps:  empty & found.  This
   # contributes 2.
-  # 
+  #
   # We can represent the overlaps as the vector [0, 0, 1, 2].  Then we will apply
   # the inverse log linear transformation defined by:
-  # 
+  #
   #   f(x_i) = x_i / ln(i + 1)
   #          = [0, 0, 1 / ln(4), 2 / ln(5)]
   #          = [0, 0, 0.7213475204444817, 1.2426698691192237]
-  # 
+  #
   # Finally we take the average of the transformed vector and normalize it (to
   # ensure a final value between 0.0 and 1.0):
-  # 
+  #
   #   norm(avg(SUM f(x_i))) = norm( avg(1.9640173895637054) )
   #                         = norm( 0.49100434739092635 )
   #                         = 0.49100434739092635 / avg(SUM f(1, 2, 3, 4))

data/lib/text_rank/keyword_extractor.rb

@@ -41,7 +41,7 @@ module TextRank
     # @option options [Array<Class, Symbol, #filter!>]  :rank_filters A list of filters to be applied to the keyword ranks after keyword extraction
     def initialize(**options)
       @page_rank_options = {
-        strategy: options[:strategy] || :dense,
+        strategy: options[:strategy] || :sparse,
         damping: options[:damping],
         tolerance: options[:tolerance],
       }

data/lib/text_rank/rank_filter/collapse_adjacent.rb

@@ -105,7 +105,8 @@ module TextRank
           # until all of the top N final keywords (single or collapsed) have been
           # considered.
           loop do
-            single_tokens_to_consider = @tokens.keys.first(@ranks_to_collapse + @to_remove.size - @to_collapse.size) - @to_remove.to_a
+            regexp_safe_tokens = @tokens.keys.select { |s| Regexp.escape(s) == s }
+            single_tokens_to_consider = regexp_safe_tokens.first(@ranks_to_collapse + @to_remove.size - @to_collapse.size) - @to_remove.to_a
             scan_text_for_all_permutations_of(single_tokens_to_consider) or break
             decide_what_to_collapse_and_what_to_remove
           end

data/lib/text_rank/version.rb

@@ -1,4 +1,4 @@
 module TextRank
   # Current gem version
-  VERSION = '1.1.7'
+  VERSION = '1.2.0'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: text_rank
 version: !ruby/object:Gem::Version
-  version: 1.1.7
+  version: 1.2.0
 platform: ruby
 authors:
 - David McCullars
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2016-07-05 00:00:00.000000000 Z
+date: 2016-10-19 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -183,7 +183,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.5.1
+rubygems_version: 2.6.7
 signing_key: 
 specification_version: 4
 summary: Implementation of TextRank solution to ranked keyword extraction