Conform lexer rules to specification

lib/rouge/lexers/csvs.rb

@@ -11,33 +11,33 @@ class CSVS < RegexLexer
 
       state :root do
         rule %r/\s+/m, Text
-        rule %r(//.*), Comment::Single
-        rule %r(#.*), Comment::Single
-        rule %r(/(\\\n)?[*].*?[*](\\\n)?/)m, Comment::Multiline
-        rule %r(/[+]), Comment::Multiline, :nested_comment
 
-        rule %r/"[^"]*"/, Str::Double
-        rule %r/'[^\r\n\f']'/, Str::Char
-
-        rule %r(:?:=), Keyword
-        rule %r/[()]/, Punctuation
+        rule %r(//[\S\t ]*), Comment::Single
+        rule %r(/\*[^*]*\*/)m, Comment::Multiline
 
-        rule %r([-=;,*+><!/|^.%&\[\]{}]), Operator
+        rule %r/(version)( )(\d+\.\d+)/ do
+          groups Keyword, Text::Whitespace, Num::Float
+        end
 
-        rule %r/[A-Z]\w*/, Name::Class
-
-        rule %r/[a-z_]\w*/, Name
+        rule %r/T?\d{2}:\d{2}:\d{2}(\.\d{5})?(Z|(?:[-+]\d{2}:\d{2}))?/, Literal::Date
+        rule %r/\d{4}-\d{2}-\d{2}/, Literal::Date
+        rule %r/\d{2}\/\d{2}\/\d{4}/, Literal::Date
 
         rule %r((\d+[.]?\d*|\d*[.]\d+)(e[+-]?[0-9]+)?)i, Num::Float
         rule %r/\d+/, Num::Integer
 
-        rule %r/@@?|\'|\:/, Keyword
-      end
+        rule %r/@\w+/, Keyword::Pseudo
+
+        rule %r/[-.\w]+:/, Name::Variable
+        rule %r/^"[^"]+"/, Name::Variable
+        rule %r/\$([-.\w]+|("[^"]+"))\/?/, Name::Variable
+
+        rule %r/[A-Z]+/i, Keyword
+
+        rule %r/"[^"]*"/, Str::Double
+        rule %r/'[^\r\n\f']'/, Str::Char
 
-      state :nested_comment do
-        rule %r([^/+]+)m, Comment::Multiline
-        rule %r(/[+]), Comment::Multiline, :nested_comment
-        rule %r([+]/), Comment::Multiline, :pop!
+        rule %r/[,()*]/, Punctuation
       end
     end
   end

spec/visual/samples/csvs

@@ -1,22 +1,32 @@
-version 1.0
-@totalColumns 19
-@separator ','
-Customer_Name: notEmpty
-Customer_Date_of_Birth: xDate
-Customer_Place_of_Birth: notEmpty
-Customer_Address: notEmpty
-Customer_Account_number: notEmpty
-Customer_Legal_Entity_Type: notEmpty
-Customer_Industry: notEmpty
-Adverse_Information_Search: xDate
-Customer_Product_Usage: notEmpty
-Customer_Net_Worth: notEmpty
-Customer_Cash_Balance: notEmpty
-Account_Open_Date: notEmpty
-CDD_Completion_Date: notEmpty
-System_Customer_Risk_Rating: notEmpty
-PEP_Status: any("yes","no")
-Associated_PEP: any("yes","no")
-EDD_Triggered: any("yes","no")
-Last_CDD_Renewal_Date: xDate
-Relationship_Manager_Employee_ID: notEmpty
+version 1.2
+@totalColumns 3
+name: notEmpty
+age: range(0, 120)
+gender: is("m") or is("f") or is("t") or is("n")
+
+@separator TAB
+@quoted
+@totalColumns 21
+@permitEmpty
+@ignoreColumnNameCase
+
+//This Comment is a Single Line Comment it terminates at this line break
+a_column:
+
+/*This Comment is a Multi Line Comment:
+it can go on for as many lines as you like, until you type*/
+a_column:
+
+a_column: is("some string") and $b_column/starts("some string") //here two tests are combined on a single line, the second test here looks to the second column 
+b_column:                                                       //to check it's value starts with "some string"
+
+a_column: is("some string")    //the contents of a_column must be the string "some string"
+a_column: is($a_column)  //the contents of a_column must be the value held in a_column, treated as a string
+
+a_column: range(*, 10)
+
+a_column: xDateTime(2014-10-04T00:00:01Z,2015-12-03T23:59:59) //the value of a_column must be a valid xDateTime and between the two xDateTimes shown (inclusive)
+a_column: xDateTime(2014-10-04T00:00:01+02:00,2015-12-03T23:59:59+02:00) //the value of a_column must be a valid xDateTime and between the two xDateTimes shown (inclusive)
+a_column: xDateTime(2014-10-04,2015-12-03) //the value of a_column must be a valid xDate and between the two xDates shown (inclusive)
+a_column: xTime(00:00:01+02:00,2015-12-03T23:59:59+02:00) //the value of a_column must be a valid xTime and between the two xTimes shown (inclusive)
+a_column: ukDate(04/10/2014,03/12/2015) //the value of a_column must be a valid ukDate and between the two ukDates shown (inclusive)