fbtxt-lexer 0.0.1

data/lib/fbtxt/lexer/token-date_duration.rb

@@ -0,0 +1,171 @@
+module Fbtxt
+class Lexer
+
+
+
+
+###
+#  date duration
+#   use - or + as separator
+#    in theory plus( +) only if dates
+#     are two days next to each other
+#
+#   otherwise  define new dates type in the future? why? why not?
+#
+#  check for plus (+) if dates are next to each other (t+1) - why? why not?
+
+#
+#  Sun Jun 23 - Wed Jun 26   -- YES
+#  Jun 23 - Jun 26           -- YES
+#  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 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})
+      [ ]
+   )?
+   (?<month_name1>#{MONTH_NAMES})
+      [ ]
+   (?<day1>\d{1,2})
+   ## optional year
+   (  ,?   # optional comma
+      [ ]
+      (?<year1>\d{4})
+   )?
+
+   ## support + and -  (add .. or such - why??)
+   [ ]* - [ ]*
+
+   ## optional day name
+   ((?<day_name2>#{DAY_NAMES})
+      [ ]
+   )?
+   (?<month_name2>#{MONTH_NAMES})
+      [ ]
+   (?<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
+#      26 July,
+XXX_DURATION_II_RE =  %r{
+(?<duration>
+    \b
+  (?
+   ## optional day name
+   ((?<day_name1>#{DAY_NAMES})
+      [ ]
+   )?
+   (?<day1>\d{1,2})
+      [ ]
+   (?<month_name1>#{MONTH_NAMES})
+   ## optional year
+   (
+       [ ]
+      (?<year1>\d{4})
+   )?
+
+   ## support + and -  (add .. or such - why??)
+   [ ]*[-][ ]*
+
+   ## optional day name
+   ((?<day_name2>#{DAY_NAMES})
+      [ ]
+   )?
+   (?<day2>\d{1,2})
+      [ ]
+   (?<month_name2>#{MONTH_NAMES})
+   ## optional year
+   ( [ ]
+      (?<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
+#
+#   - 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,
+)
+
+
+
+end  #   class Lexer
+end  # module Fbtxt

data/lib/fbtxt/lexer/token-geo.rb

@@ -0,0 +1,173 @@
+module Fbtxt
+class Lexer
+
+
+##
+#  allow Cote'd Ivoir or such
+##   e.g. add '
+
+##  allow Ta' Qali  !!!
+
+
+## todo/fix - make geo text regex more generic
+##               only care about two space rule
+
+
+##
+## before with optional space
+##   trying to use "hard" (joiner) space
+=begin
+                (?:
+                    [ ]?   # only single (inline) space allowed - double spaces are breaks!!!
+                    (?:
+                       \p{L} | \d  | [.&'°]
+                        |
+                       (?: (?<! [ ])  ## no space allowed before (but possible after)
+                            [-]
+                       )
+                         |
+                       (?: (?<! [ ])  ## no spaces allowed around these characters
+                           [_/]
+                          (?! [ ])
+                       )
+                    )+
+=end
+
+## positive lookbehind -  for now space (or beginning of line - for testing) only
+##  (MUST be fixed number of chars - no quantifier e.g. +? etc.)
+GEO_LEFT_BOUNDARY_  = '(?<= [ ,›>\[\]]|^)'
+## add lookahead/lookbehind
+##    must be space!!!
+##   (or comma or  start/end of string)
+##   kind of \b !!!
+ ## POSITIVE lookahead
+GEO_RIGHT_BOUNDARY_ = '(?= [ ,›>\[\]]|$)'
+
+
+
+GEO_TEXT_RE = %r{
+    ## must start with alpha (allow unicode letters!!)
+    (?<text>
+             #{GEO_LEFT_BOUNDARY_}
+          (?:
+              (?:
+                # opt 1 - start with alpha
+                #   note - allow (dots)
+                 \p{L}[\p{L}'.]*    ## all unicode letters (e.g. [a-z])
+                   |
+                # opt 2 - start with num!! -
+                     \d+  # check for num lookahead (MUST be space or dot)
+                      ## MAY be followed by (optional space) !
+                      ## MUST be follow by a to z!!!!
+                      [ ]?   ## make space optional too  - why? why not?
+                             ##  yes - eg. 1st, 2nd, 5th etc.
+                       \p{L}[\p{L}.]*
+                  |
+                ## opt 3 - add another weirdo case
+                ##   e.g.   's Gravenwezel-Schilde
+                ##   add more letters (or sequences here - why? why not?)
+                ##    or hard-code simly 's for now - why? why not?
+                    ' \p{L}[\p{L}.]*
+               )
+
+               ##
+               ## todo/check - find a different "more intuitive" regex/rule if possible?
+               ##    for single spaces only
+               ## | [_]              ## no space allowed around - keep (why? why not?)
+
+
+                (?:
+                  (?:
+                        [ ]? [&/-] [ ]?  ## note - yes, allow space before and after
+                     |  [ ]              # only single (inline) space allowed - double spaces are breaks!!!
+                   )
+
+                   (?:
+                       ##  followed by word (incl. specials .°' for now)
+                       [\p{L}\d.°']+
+                    ##  todo/fix - possible in regex here
+                    ##     only end in alphanum a-z0-9 (not dot or & ???)
+
+                       ## or
+              ##   parenthesis enclosed closed text
+              ##   e.g. Dublin (Dalymount Park)
+              ##        Bucuresti (23 August)
+              ##        Paris (Parc des Princes)
+              ##        Ost-Berlin (Walter-Ulbricht)
+              ##        Athinai (OAKA - Maroussi)
+              ##
+              ##   or   Valencia (Spain) or Solna
+                   |   \(
+                        [^()\[\],;:›<>]+    ## todo - add more special chars
+                                            ##   maybe list only allowed ones??
+                                            ##   make pattern more strict - why? why not?
+                      \)
+                 )
+               )*
+        )
+            #{GEO_RIGHT_BOUNDARY_}
+   )
+}ix
+
+
+
+
+
+
+##  note - add "hacky" check for comma that is followed by a prop(erty)
+##
+##  make sure to NOT match
+##      props  e.g.  att: 18000
+##   July 10 @ Paris, Parc des Princes, att: 18000
+##   July 10 @ Paris, Parc des Princes,   att: 18000
+##
+
+
+GEO_END_RE = %r{
+   (?<geo_end>
+        ,
+    )
+    ## POSITIVE lookahead for props
+    ##   todo/fix - use generic [a-z]+ - why? why not?
+    (?=
+        [ ]*  ## optional spaces
+         (?:     attendance|att
+              |  referee?s|refs?
+          )
+         :
+    )
+}ix
+
+
+
+
+GEO_RE = Regexp.union(
+                    SPACES_RE,
+                    GEO_END_RE,
+                    GEO_TEXT_RE,
+                    /  (?<sym> [,›>\[▪] ) /x,
+                    ANY_RE,
+                      )
+
+
+
+###
+## helper for testing regex match for geo names
+
+def self._parse_geo( str )
+    ## note - strip - leading/trailing spaces
+    m = GEO_TEXT_RE.match( str.strip )
+    if m && m.pre_match == '' && m.post_match == ''
+      m
+    elsif  m
+        ## note - match BUT not anchored to start and end-of-string!!!
+        ##  report, error somehow??
+      nil
+    else
+      nil  ## no match - return nil
+    end
+end
+
+
+end # class Lexer
+end # module Fbtxt

data/lib/fbtxt/lexer/token-goals--helpers.rb

@@ -0,0 +1,114 @@
+module Fbtxt
+class Lexer
+
+
+
+def self._build_goal_minute( m )
+    minute = {}
+
+    minute[:m]     =  m[:value].to_i(10)   ## always required
+
+    ## stoppage/injury time (offset)
+    minute[:offset] = m[:value2].to_i(10)   if m[:value2]
+
+    minute[:og]  = true       if m[:og]
+    minute[:pen] = true       if m[:pen]
+    minute[:freekick] = true  if m[:fk]
+    minute[:header] = true    if m[:hdr]
+
+    minute[:secs] = m[:secs].to_i(10)   if m[:secs]
+
+    minute
+end
+
+def self._build_goal_minute_na( m )
+    minute = {}
+
+    minute[:m]     =  '?'   ##  or use nil or 999 or -1 or ???
+
+    minute[:og]  = true       if m[:og]
+    minute[:pen] = true       if m[:pen]
+    minute[:freekick] = true  if m[:fk]
+    minute[:header] = true    if m[:hdr]
+
+    minute
+end
+
+
+
+def self._build_minute( m )
+    minute = {}
+    minute[:m]      = m[:value].to_i(10)   ## always required
+
+    ## stoppage/injury time (offset)
+    minute[:offset] = m[:value2].to_i(10)   if m[:value2]
+
+    minute
+end
+
+
+def self._build_goal_count( m )
+    count = {}
+    count[:count] = m[:value].to_i(10)        if m[:value]
+    count[:og]    = m[:og_value] ? m[:og_value].to_i(10) : 1      if m[:og]   ## check flag
+    count[:pen]   = m[:pen_value] ? m[:pen_value].to_i(10) : 1    if m[:pen]  ## check flag
+    count
+end
+
+def self._build_goal_type( m )
+    goal = {}
+    goal[:og]       = true  if m[:og]
+    goal[:pen]      = true  if m[:pen]
+    goal[:freekick] = true  if m[:fk]
+    goal[:header]   = true  if m[:hdr]
+    goal
+end
+
+
+def _build_goal_minute( m ) self.class._build_goal_minute( m ); end
+def _build_goal_minute_na( m ) self.class._build_goal_minute_na( m ); end
+def _build_minute( m ) self.class._build_minute( m ); end
+def _build_goal_count( m ) self.class._build_goal_count( m ); end
+def _build_goal_type( m ) self.class._build_goal_type( m ); end
+
+
+
+
+
+###
+# parse helpers
+
+def self._parse_goal_minute( str )
+    ## note - strip - leading/trailing spaces
+    m = GOAL_MINUTE_RE.match( str.strip )
+    if m && m.pre_match == '' && m.post_match == ''
+      _build_goal_minute( m )
+    elsif  m
+        ## note - match BUT not anchored to start and end-of-string!!!
+        ##  report, error somehow??
+      nil
+    else
+      nil  ## no match - return nil
+    end
+end
+
+def self._parse_goal_count( str )
+    ## note - strip - leading/trailing spaces
+    m = GOAL_COUNT_RE.match( str.strip )
+    if m && m.pre_match == '' && m.post_match == ''
+      _build_goal_count( m )
+    elsif  m
+        ## note - match BUT not anchored to start and end-of-string!!!
+        ##  report, error somehow??
+      nil
+    else
+      nil  ## no match - return nil
+    end
+end
+
+
+
+
+
+end   # class Lexer
+end   # module Fbtxt