sportdb-parser 0.5.9 → 0.6.1

data/lib/sportdb/parser/racc_parser.rb

@@ -14,7 +14,7 @@ def initialize( txt,  debug: false )
 
     ### todo:
     ##  -  pass along debug flag
-    lexer = SportDb::Lexer.new( txt )
+    lexer = SportDb::Lexer.new( txt, debug: debug )
     ##  note - use tokenize_with_errors and add/collect tokenize errors
     @tokens, @errors = lexer.tokenize_with_errors
     ## pp @tokens
@@ -68,12 +68,14 @@ def initialize( txt,  debug: false )
 
 
   def on_error(error_token_id, error_value, value_stack)
-    args = [error_token_id, error_value, value_stack]
+    ## auto-add error_token (as string)
+    error_token = Racc_token_to_s_table[error_token_id] 
+    args = [error_token, error_token_id, error_value, value_stack]
     puts
     puts "!! on parse error:"
     puts "args=#{args.pretty_inspect}"
 
-    @errors << "parse error on token: #{error_token_id} with value: #{error_value}, stack: #{value_stack.pretty_inspect}"
+    @errors << "parse error on token: #{error_token} (#{error_token_id}) with value: #{error_value}, stack: #{value_stack.pretty_inspect}" 
     ## exit 1  ##   exit for now  -  get and print more info about context etc.!!
   end
 

data/lib/sportdb/parser/racc_tree.rb

@@ -71,10 +71,14 @@ RoundDef   = Struct.new( :name, :date, :duration )  do
   end
 end
 
-DateHeader = Struct.new( :date ) do
+DateHeader = Struct.new( :date, :time, :geo, :timezone ) do
   def pretty_print( printer )
     printer.text( "<DateHeader " )
-    printer.text( "#{self.date.pretty_inspect}>" )
+    printer.text( "#{self.date.pretty_inspect}" )
+    printer.text( " time=#{self.time.pretty_inspect}" )          if self.time
+    printer.text( " geo=#{self.geo.pretty_inspect}" )            if self.geo
+    printer.text( " timezone=#{self.timezone}")             if self.timezone
+    printer.text( ">")
   end
 end
 
@@ -85,14 +89,17 @@ GroupHeader = Struct.new( :name ) do
   end
 end
 
-RoundHeader = Struct.new( :names ) do
+RoundHeader = Struct.new( :names, :group ) do
   def pretty_print( printer )
     printer.text( "<RoundHeader " )
-    printer.text( "#{self.names.join(', ')}>" )
+    printer.text( "#{self.names.join(', ')}" )
+    printer.text( " group=#{self.group}")    if self.group
+    printer.text( ">" )
   end
 end
 
-MatchLine   = Struct.new( :ord, :date, :time,
+
+MatchLine   = Struct.new( :ord, :date, :time, :wday,
                           :team1, :team2, :score,
                           :status, 
                           :geo,

data/lib/sportdb/parser/token-date.rb

@@ -146,12 +146,29 @@ DATE_II_RE = %r{
 )}ix
 
 
+# e.g. iso-date  -  2011-08-25     
+##   note - allow/support ("shortcuts") e.g 2011-8-25  or 2011-8-3 / 2011-08-03 etc. 
+DATE_III_RE = %r{
+(?<date>
+  \b
+   (?<year>\d{4})
+       -
+   (?<month>\d{1,2})
+       -
+   (?<day>\d{1,2})
+  \b
+)}ix
+
+
+
+
 #############################################
 # map tables
 #  note: order matters; first come-first matched/served
 DATE_RE = Regexp.union(
    DATE_I_RE,
-   DATE_II_RE
+   DATE_II_RE,
+   DATE_III_RE,
 )
 
 
@@ -197,29 +214,36 @@ end
 #
 #  Sun Jun/23 - Wed Jun/26   -- YES
 #  Jun/23 - Jun/26           -- YES
-#  Tue Jun/25 + Wed Jun/26   -- YES
-#  Jun/25 + Jun/26           -- YES
-#
-#  Jun/25 - 26        - why? why not???
+#  Jun/25 - 26        - why? why not???  - YES - see blow variant iii!!!
+
+#  Tue Jun/25 + Wed Jun/26   -- NO
+#  Jun/25 + Jun/26           -- NO
 #  Jun/25 .. 26        - why? why not???
 #  Jun/25 to 26        - why? why not???
 #  Jun/25 + 26        - add - why? why not???
 #  Sun-Wed Jun/23-26  -  add - why? why not???
 #  Wed+Thu Jun/26+27 2024  -  add - why? why not???
 #
-#  maybe use comman and plus for list of dates
+#  maybe use comma and plus for list of dates
 #    Tue Jun/25, Wed Jun/26, Thu Jun/27  ??
 #    Tue Jun/25 + Wed Jun/26 + Thu Jun/27  ??
 #
 #   add back optional comma (before) year - why? why not?
+#
 
 
 ##
 #   todo add plus later on - why? why not?
+###   todo/fix  add optional comma (,) before year
+
+### regex note/tip/remindr -  \b () \b MUST always get enclosed in parantheses
+##                                     because alternation (|) has lowest priority/binding
+
 
 DURATION_I_RE =  %r{
 (?<duration>
     \b
+  (?:
    ## optional day name
    ((?<day_name1>#{DAY_NAMES})
       [ ]
@@ -228,12 +252,13 @@ DURATION_I_RE =  %r{
       (?: \/|[ ] )
    (?<day1>\d{1,2})
    ## optional year
-   ( [ ]
+   (  ,?   # optional comma
+      [ ]
       (?<year1>\d{4})
    )?
 
    ## support + and -  (add .. or such - why??)
-   [ ]*[-][ ]*
+   [ ]* - [ ]*
 
    ## optional day name
    ((?<day_name2>#{DAY_NAMES})
@@ -243,20 +268,28 @@ DURATION_I_RE =  %r{
       (?: \/|[ ] )
    (?<day2>\d{1,2})
    ## optional year
-   ( [ ]
+   (  ,?   # optional comma
+      [ ]
       (?<year2>\d{4})
    )?
+  )
    \b
 )}ix
 
 
+
+#   FIX - remove this variant 
+#         "standardize on month day [year]" !!!!
+
+=begin
 ###
 #   variant ii
 # e.g. 26 July - 27 July
-
-DURATION_II_RE =  %r{
+#      26 July, 
+XXX_DURATION_II_RE =  %r{
 (?<duration>
     \b
+  (?
    ## optional day name
    ((?<day_name1>#{DAY_NAMES})
       [ ]
@@ -265,7 +298,8 @@ DURATION_II_RE =  %r{
       [ ]
    (?<month_name1>#{MONTH_NAMES})
    ## optional year
-   ( [ ]
+   (  
+       [ ]
       (?<year1>\d{4})
    )?
 
@@ -283,16 +317,50 @@ DURATION_II_RE =  %r{
    ( [ ]
       (?<year2>\d{4})
    )?
+  )
+   \b
+)}ix
+=end
+
+
+#  variant ii
+#  add support for shorthand
+#     August 16-18, 2011     
+#     September 13-15, 2011
+#      October 18-20, 2011
+#      March/6-8, 2012
+#      March 6-8 2012
+#      March 6-8
+#     
+#   - add support for August 16+17 or such (and check 16+18)
+#       use <op> to check if day2 is a plus or range or such - why? why not?
+
+DURATION_II_RE =  %r{
+(?<duration>
+    \b
+   (?:
+       (?<month_name1>#{MONTH_NAMES})
+           [ /]
+        (?<day1>\d{1,2})
+             -
+        (?<day2>\d{1,2})
+          (?:
+            ,?     ## optional comma
+            [ ]
+            (?<year1>\d{4})
+          )?     ## optional year   
+   )
    \b
 )}ix
 
 
+
 #############################################
 # map tables
 #  note: order matters; first come-first matched/served
 DURATION_RE = Regexp.union(
    DURATION_I_RE,
-   DURATION_II_RE
+   DURATION_II_RE,
 )
 
 

data/lib/sportdb/parser/token-minute.rb

@@ -0,0 +1,45 @@
+
+module SportDb
+class Lexer
+
+#
+#  todo/check - move goal type regexes to goal or somewhere else?
+#
+
+##   goal types
+# (pen.) or (pen) or (p.) or (p)
+## (o.g.) or (og)
+##   todo/check - keep case-insensitive 
+##                   or allow OG or P or PEN or
+##                   only lower case - why? why not?
+GOAL_PEN_RE = %r{
+   (?<pen> \(
+           (?:pen|p)\.?
+           \)
+    )
+}ix
+GOAL_OG_RE = %r{
+   (?<og> \(
+          (?:og|o\.g\.)
+          \)
+   )
+}ix
+
+
+MINUTE_RE = %r{
+     (?<minute>
+       (?<=[ (])	 # positive lookbehind for space or opening ( e.g. (61') required
+                     #    todo - add more lookbehinds e.g.  ,) etc. - why? why not?
+           (?<value>\d{1,3})      ## constrain numbers to 0 to 999!!!
+        (?: \+
+            (?<value2>\d{1,3})   
+        )?
+        '     ## must have minute marker!!!!
+     )
+}ix
+
+
+
+    
+end   # module SportDb
+end   # class Lexer

data/lib/sportdb/parser/token-prop.rb

@@ -0,0 +1,133 @@
+###
+##  team prop mode e.g.
+##
+##
+##    Fri Jun 14 21:00  @ München Fußball Arena, München        
+##  (1)  Germany  v  Scotland   5-1 (3-0)     
+##  Wirtz 10' Musiala 19' Havertz 45+1' (pen.) Füllkrug 68' Can 90+3'; Rüdiger 87' (o.g.)
+## 
+## Germany:    Neuer - Kimmich, Rüdiger, Tah [Y], Mittelstädt - Andrich [Y] (Groß 46'),
+##       Kroos (Can 80') - Musiala (Müller 74'), Gündogan, Wirtz (Sane 63') - 
+##       Havertz (Füllkrug 63')
+## Scotland:   Gunn - Porteous [R 44'], Hendry, Tierney (McKenna 78') - Ralston [Y],
+##       McTominay, McGregor (Gilmour 67'), Robertson - Christie (Shankland 82'),
+##       Adams (Hanley 46'), McGinn (McLean 67')
+
+
+module SportDb
+class Lexer
+
+
+## name different from text (does NOT allow number in name/text)
+
+PROP_NAME_RE = %r{
+                 (?<prop_name> \b
+                   (?<name>
+                      \p{L}+       
+                        \.?    ## optional dot
+                      (?: 
+                          [ ]?    # only single spaces allowed inline!!!
+                          (?:
+                              (?:
+                                (?<=\p{L})   ## use lookbehind
+                                 [/'-]   ## must be surrounded by letters
+                                       ## e.g. One/Two NOT
+                                       ##      One/ Two or One / Two or One /Two etc.
+                                (?=\p{L})      ## use lookahead        
+                              )
+                                 |   
+                              (?:
+                                (?<=[ ])   ## use lookbehind  -- add letter (plus dot) or such - why? why not?
+                                 [']   ## must be surrounded by leading space and
+                                       ## traling letters  (e.g. UDI 'Beter Bed)
+                                (?=\p{L})      ## use lookahead        
+                              )   
+                                 |
+                              (?:
+                                (?<=\p{L})   ## use lookbehind
+                                 [']   ## must be surrounded by leading letter and
+                                       ## trailing space PLUS letter  (e.g. UDI' Beter Bed)
+                                (?=[ ]\p{L})      ## use lookahead (space WITH letter         
+                              )   
+                                 |   ## standard case with letter(s) and optinal dot
+                              (?: \p{L}+
+                                    \.?  ## optional dot
+                              )
+                          )+
+                     )*
+                   )
+               ## add lookahead - must be non-alphanum 
+                  (?=[ ,;\]\)]|$)
+                  )
+}ix
+
+
+
+
+##############
+#  add support for props/ attributes e.g.
+# 
+#    Germany:    Neuer - Kimmich, Rüdiger, Tah [Y], Mittelstädt - Andrich [Y] (46' Groß),
+#      Kroos (80' Can) - Musiala (74' Müller), Gündogan,
+#      Wirtz (63' Sane) - Havertz (63' Füllkrug)
+#    Scotland:   Gunn - Porteous [R 44'], Hendry, Tierney (78' McKenna) - Ralston [Y],
+#      McTominay, McGregor (67' Gilmour), Robertson - Christie (82' Shankland),
+#      Adams (46' Hanley), McGinn (67' McLean)
+#
+## note:  colon (:) MUST be followed by one (or more) spaces
+##      make sure mon feb 12 18:10 will not match
+##        allow 1. FC Köln etc.
+##               Mainz 05:
+##           limit to 30 chars max
+##          only allow  chars incl. intl but (NOT ()[]/;)
+##
+## todo/fix:
+##   check if   St. Pölten     works; with starting St. ???
+
+
+  PROP_KEY_RE = %r{ 
+                 (?<prop_key> \b
+                   (?<key>
+                       (?:\p{L}+
+                           |
+                           \d+  # check for num lookahead (MUST be space or dot)
+                        ## MUST be followed by (optional dot) and
+                        ##                      required space !!!
+                        ## MUST be follow by a to z!!!!
+                         \.?     ## optional dot
+                         [ ]?   ## make space optional too  - why? why not?
+                             ##  yes - eg. 1st, 2nd, 5th etc.
+                         \p{L}+
+                        )
+                        [\d\p{L}'/° -]*?   ## allow almost anyting 
+                                          ## fix - add negative lookahead 
+                                          ##         no space and dash etc.
+                                          ##    only allowed "inline" not at the end
+                                          ## must end with latter or digit!
+                   )
+                    [ ]*?     # slurp trailing spaces
+                     :
+                    (?=[ ]+)  ## possitive lookahead (must be followed by space!!)
+                   )
+                 }ix
+
+
+
+PROP_BASICS_RE = %r{
+    (?<spaces> [ ]{2,}) |
+    (?<space>  [ ])
+        |
+    (?<sym>  
+        [;,\(\)\[\]-] 
+    )   
+}ix
+
+PROP_RE = Regexp.union(
+   PROP_BASICS_RE, 
+   MINUTE_RE,
+   PROP_NAME_RE,
+)
+
+    
+end  # class Lexer
+end  # module SportDb

data/lib/sportdb/parser/token-score.rb

@@ -17,7 +17,7 @@ class Lexer
     ##      3-4 pen.   2-2 a.e.t.
     ##               2-2 a.e.t.
     SCORE__P_ET__RE = %r{
-        (?<score>
+        (?<score_more>
            \b
             (?:
                (?<p1>\d{1,2}) - (?<p2>\d{1,2})
@@ -34,7 +34,7 @@ class Lexer
     ##  note: allow SPECIAL with penalty only
     ##      3-4 pen.
     SCORE__P__RE = %r{
-        (?<score>
+        (?<score_more>
            \b
               (?<p1>\d{1,2}) - (?<p2>\d{1,2})
                 [ ]* #{P_EN}
@@ -52,7 +52,7 @@ class Lexer
     ##               2-2 a.e.t. (1-1)
 
     SCORE__P_ET_FT_HT__RE = %r{
-          (?<score>
+          (?<score_more>
                \b
                (?:
                 (?<p1>\d{1,2}) - (?<p2>\d{1,2})
@@ -79,7 +79,7 @@ class Lexer
     ##   special case for case WITHOUT extra time!!
     ##     same as above (but WITHOUT extra time and pen required)
     SCORE__P_FT_HT__RE = %r{
-             (?<score>
+             (?<score_more>
                 \b
      (?<p1>\d{1,2}) - (?<p2>\d{1,2})
         [ ]* #{P_EN} [ ]+
@@ -99,36 +99,47 @@ class Lexer
             ## note: \b works only after non-alphanum e.g. )
 
 
-
-    ## e.g. 2-1 (1-1) or
-    ##      2-1
-
+    ##########
+    ## e.g. 2-1 (1-1)
     SCORE__FT_HT__RE = %r{
-            (?<score>
+            (?<score_more>
               \b
               (?<ft1>\d{1,2}) - (?<ft2>\d{1,2})
-               (?:
                    [ ]+ \( [ ]*
                 (?<ht1>\d{1,2}) - (?<ht2>\d{1,2})
                    [ ]* \)
-               )?   # note: make half time (HT) score optional for now
              (?=[ ,\]]|$)
              )}ix    ## todo/check:  remove loakahead assertion here - why require space?
                     ## note: \b works only after non-alphanum e.g. )
 
-
+    #####
+    ##      2-1
+    SCORE__FT__RE = %r{
+            (?<score>
+              \b
+              (?<ft1>\d{1,2}) - (?<ft2>\d{1,2})
+              \b
+             )}ix  
 
 #############################################
 # map tables
 #  note: order matters; first come-first matched/served
+#
+## check - find a better name for SCORE_MORE - SCORE_EX, SCORE_BIG, or ___ - why? why not?
 
-SCORE_RE = Regexp.union(
+SCORE_MORE_RE = Regexp.union(
   SCORE__P_ET_FT_HT__RE,  # e.g. 5-1 pen. 2-2 a.e.t. (1-1, 1-0)
   SCORE__P_FT_HT__RE,     # e.g. 5-1 pen. (1-1)
   SCORE__P_ET__RE,        # e.g. 2-2 a.e.t.  or  5-1 pen. 2-2 a.e.t.
   SCORE__P__RE,           # e.g. 5-1 pen.
-  SCORE__FT_HT__RE,        # e.g. 1-1 (1-0) or 1-1  -- note - must go last!!!
+  SCORE__FT_HT__RE,        # e.g. 1-1 (1-0)
+  ##  note - keep basic score as its own token!!!!
+  ##   that is, SCORE & SCORE_MORE
+  ### SCORE__FT__RE,           # e.g. 1-1  -- note - must go last!!!
 )
 
+SCORE_RE   =   SCORE__FT__RE
+  
+
 end  #  class Lexer
 end  # module SportDb

data/lib/sportdb/parser/token-text.rb

@@ -24,6 +24,13 @@ class Lexer
 #  allow Cote'd Ivoir or such
 ##   e.g. add '
 
+## note:
+##  make sure these do NOT match!!!
+## TEXT  =>  "Matchday 1 / Group A"
+## TEXT  =>  "Matchday 2 / Group A"
+## TEXT  =>  "Matchday 3 / Group A"
+
+
 
 TEXT_RE = %r{
     ## must start with alpha (allow unicode letters!!)
@@ -59,11 +66,11 @@ TEXT_RE = %r{
                                ##    AND switch to case-sensitive (via -i!!!)
                       )
                       |     # only single spaces allowed inline!!!
-                     [-]
+                     [-/]
                   )?
                 (?:
                   \p{L} |
-                  [&/'°]
+                  [&'°]
                     |
                  (?:
                    \d+