human-ql 1.0.1 → 1.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 51ffc03488bc280dbdb361e86b029a56c6deac74
-  data.tar.gz: d02aac690117c9849b08a89558dece5c06db8a6c
+  metadata.gz: 98be83fa893f858ada46ea6938e4bf217662acc2
+  data.tar.gz: 56ae63c63848dca766c6ea66091e87409588710f
 SHA512:
-  metadata.gz: c5f2be15113a044056f332b114c8ce768e71c3a3db4699812df2b46334446a68ceaec2dbd081044f707cf208b4722c565bd2bfaa0ccf24b7601bb65be0ee7679
-  data.tar.gz: c4a773aedd6a5aeb7a5517115f1a266be27edab6a1741db739cfcc7585bda4594b30f824e4f19e8c2b8d54c3beb61038933bfa580a1da83d86d95406bbf139c7
+  metadata.gz: cb36220c0a286185221e5b21a3592797ab72ce5924991a6f08376e96507c6f7e64d69945baa4a0260c5bea49f8da3ea9e91332409eb68f68523814fc32de992c
+  data.tar.gz: 560f9b4ce26140065beb8c2456e8227d34076e766eec8dddc447f8ece34c5dbe96de95f8041d8840c2bcd8d3ded1b86223fba091a0dd42fc8ec337d516c2bd50

data/History.rdoc

@@ -1,3 +1,13 @@
+=== 1.0.2 (2017-3-28)
+
+* HumanQL::PostgreSQLCustomParser now treats backslashes and embedded
+  nulls as whitespace since these characters were found to be
+  significant to the PG tsquery parser.
+
+* Additional phrase and text match tests in TestPostgreSQLGenerator.
+
+* Travis CI config including PostgreSQL 9.6
+
 === 1.0.1 (2017-3-21)
 
 * Drop more tokens, keep more characters in PostgreSQLCustomParser,

data/README.rdoc

@@ -2,6 +2,7 @@
 
 * http://github.com/dekellum/human-ql
 * http://rdoc.gravitext.com/human-ql/
+* {<img src="https://travis-ci.org/dekellum/human-ql.svg?branch=dev" />}[https://travis-ci.org/dekellum/human-ql]
 
 == Description
 

data/Rakefile

@@ -12,3 +12,10 @@ task :publish_rdoc => [ :clean, :rerdoc ] do
     aws s3 sync --acl public-read doc/ s3://rdoc.gravitext.com/human-ql/
   SH
 end
+
+task :rdoc do
+  sh <<-SH
+    rm -rf doc/fonts
+    cp rdoc_css/*.css doc/css/
+  SH
+end

data/lib/human-ql/base.rb

@@ -16,6 +16,6 @@
 
 module HumanQL
 
-  VERSION='1.0.1'
+  VERSION='1.0.2'
 
 end

data/lib/human-ql/postgresql_custom_parser.rb

@@ -58,13 +58,13 @@ module HumanQL
         # whitespace. This can't include anything part of HumanQL,
         # e.g. ':' as used for scopes, so deal with the remainder
         # below.
-        self.spaces = /[[:space:]*!<>]+/.freeze
+        self.spaces = /[[:space:]*!<>\0\\]+/.freeze
       else
         # Disable quote tokens
         self.lquote = nil
         self.rquote = nil
         # As above but add DQUOTE as well.
-        self.spaces = /[[:space:]*!<>"]+/.freeze
+        self.spaces = /[[:space:]*!<>\0\\"]+/.freeze
       end
 
       # Use by custom #norm_phrase_tokens as a superset of the

data/test/test_postgresql_fuzz.rb

@@ -64,8 +64,8 @@ class TestPostgresqlFuzz < Minitest::Test
   # Starting point query
   GENERIC_Q = 'ape | ( boy -"cat dog" )'.freeze
 
-  # Characters which are likely to cause trouble
-  RANDOM_C = '({"\'a !:* ,^#:/-0.123e-9)<>'.freeze
+  # Additional characters which could conceivably cause trouble
+  RANDOM_C = ( '({"a !:*\n\t ,^#:/-0.123e-9)&<>' + "\0\'\\" ).freeze
 
   PASSES.times do |i|
     define_method( "test_fuzz_#{i}" ) do

data/test/test_postgresql_generator.rb

@@ -61,6 +61,22 @@ class TestPostgresqlGenerator < Minitest::Test
     end
   end
 
+  def assert_tsq_match( tsq, text )
+    if DB
+      rt = DB[ "select to_tsvector(?) @@ to_tsquery(?) as m",
+               text, tsq ].first[ :m ]
+      assert_equal( true, rt )
+    end
+  end
+
+  def refute_tsq_match( tsq, text )
+    if DB
+      rt = DB[ "select to_tsvector(?) @@ to_tsquery(?) as m",
+               text, tsq ].first[ :m ]
+      assert_equal( false, rt )
+    end
+  end
+
   def test_gen_term
     assert_gen( 'ape', 'ape' )
     assert_tsq( "'ape'", 'ape' )
@@ -87,6 +103,65 @@ class TestPostgresqlGenerator < Minitest::Test
     assert_tsq( "'boy'", '": boy"' )
   end
 
+  def test_phrase_or
+    skip( "For postgresql 9.6+" ) unless pg_gte_9_6?
+    # '<->' has precendence over '|'
+    assert_gen( "(ape <-> boy | dog <-> girl)",
+                '"ape boy" or "dog girl"' )
+
+    assert_tsq( tsq = "'ape' <-> 'boy' | 'dog' <-> 'girl'",
+                '"ape boy" or "dog girl"' )
+
+    refute_tsq_match( tsq, 'boy girl', )
+    refute_tsq_match( tsq, 'girl dog', )
+
+    assert_tsq_match( tsq, 'ape boy', )
+    assert_tsq_match( tsq, 'ape boy cat', )
+    assert_tsq_match( tsq, 'dog girl', )
+    assert_tsq_match( tsq, 'ape dog girl', )
+  end
+
+  def test_phrase_stopword
+    assert_tsq( tsq = "'cat' <2> 'dog'", '"cat _ dog"' )
+
+    assert_tsq_match( tsq, 'cat goose dog' )
+
+    # '<2>' is exact, not up to...
+    refute_tsq_match( tsq, 'cat dog', )
+    refute_tsq_match( tsq, 'cat girl goose dog', )
+
+    # to_tsvector('cat a dog') -> 'cat':1 'dog':3
+    # to_tsvector('cat _ dog') -> 'cat':1 'dog':2
+
+    assert_tsq_match( tsq, 'cat a dog', )
+    refute_tsq_match( tsq, 'cat _ dog', )
+  end
+
+  def test_phrase_apostrophe
+    skip( "For postgresql 9.6+" ) unless pg_gte_9_6?
+
+    # As of 9.6.2 this phrase won't match itself.
+    assert_gen( "cat's <-> rat", %{"cat's rat"} )
+    assert_tsq( tsq = "'cat' <-> 'rat'", %{"cat's rat"} )
+
+    assert_tsq_match( tsq, "cat rat", )
+
+    # to_tsvector('cat''s rat') -> 'cat':1 'rat':3
+    refute_tsq_match( tsq, "cat's rat", )
+  end
+
+  def test_phrase_and
+    skip( "For postgresql 9.6+" ) unless pg_gte_9_6?
+    assert_gen( "johnson <-> johnson", '"johnson & johnson"' )
+    assert_tsq( tsq = "'johnson' <-> 'johnson'", '"johnson & johnson"' )
+
+    # to_tsvector('johnson & johnson') -> 'johnson':1,2
+    assert_tsq_match( tsq, "Johnson & Johnson", )
+
+    # to_tsvector('Johnson A Johnson') -> 'johnson':1,3
+    refute_tsq_match( tsq, "Johnson A Johnson", )
+  end
+
   def test_gen_empty
     assert_gen( nil, '' )
   end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: human-ql
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 1.0.2
 platform: ruby
 authors:
 - David Kellum
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-03-21 00:00:00.000000000 Z
+date: 2017-03-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: minitest
@@ -24,6 +24,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: 5.8.4
+- !ruby/object:Gem::Dependency
+  name: rdoc
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 4.3.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 4.3.0
 - !ruby/object:Gem::Dependency
   name: sequel
   requirement: !ruby/object:Gem::Requirement