sportdb-formats 1.1.6 → 1.2.1

data/lib/sportdb/formats/team/club_reader.rb

@@ -1,350 +1,318 @@
-# encoding: utf-8
-
-
-module SportDb
-module Import
-
-
-class ClubReader
-
-  def catalog() Import.catalog; end
-
-
-
-def self.read( path )   ## use - rename to read_file or from_file etc. - why? why not?
-  txt = File.open( path, 'r:utf-8' ) { |f| f.read }
-  parse( txt )
-end
-
-def self.parse( txt )
-  new( txt ).parse
-end
-
-def initialize( txt )
-  @txt = txt
-end
-
-##  pattern for b (child) team / club marker e.g.
-##    (ii) or ii) or ii.) or (ii.) or (II)
-##    (b)  or b)  or b.)  or (b.)  or (B)
-##    (2)  or 2)  or 2.)  or (2.)
-B_TEAM_MARKER_RE = %r{^  \(?     # optional opening bracket
-                          (?: ii|b|2 )
-                          \.?     # optional dot - keep and allow dot - why? why not?
-                          \)      # required closing bracket
-                     }xi   ## note: add case-insenstive (e.g. II/ii or B/b)
-
-## pattern for checking for address line e.g.
-##    use just one style / syntax - why? why not?
-##  Fischhofgasse 12 ~ 1100 Wien or
-##  Fischhofgasse 12 // 1100 Wien   or Fischhofgasse 12 /// 1100 Wien
-##  Fischhofgasse 12 ++ 1100 Wien   or Fischhofgasse 12 +++ 1100 Wien
-ADDR_MARKER_RE =  %r{ (?: ^|[ ] )                # space or beginning of line
-                        (?: ~ | /{2,} | \+{2,} )
-                      (?: [ ]|$)                 # space or end of line
-                       }x
-
-
-def add_alt_names( rec, names )   ## helper for adding alternat names
-
-  ## strip and  squish (white)spaces
-  #   e.g. New York FC      (2011-)  => New York FC (2011-)
-  names = names.map { |name| name.gsub( '$', '' ).strip
-                                 .gsub( /[ \t]+/, ' ' ) }
-  rec.alt_names += names
-  rec.add_variants( names ) # auto-add (possible) auto-generated variant names
-
-  ## check for duplicates
-  if rec.duplicates?
-    duplicates = rec.duplicates
-    puts "*** !!! WARN !!! - #{duplicates.size} duplicate alt name mapping(s):"
-    pp duplicates
-    pp rec
-    ##
-    ##  todo/fix:  make it only an error with exit 1
-    ##               if (not normalized) names are the same (not unique/uniq)
-    ##                  e.g. don't exit on  A.F.C. == AFC etc.
-    ## exit 1
-  end
-end
-
-
-def parse
-  recs = []
-  last_rec  = nil
-  headings = []   ## headings stack
-
-  OutlineReader.parse( @txt ).each do |node|
-    if [:h1,:h2,:h3,:h4,:h5,:h6].include?( node[0] )
-      heading_level  = node[0][1].to_i
-      heading        = node[1]
-
-      puts "heading #{heading_level} >#{heading}<"
-
-      ## 1) first pop headings if present
-      while headings.size+1 > heading_level
-        headings.pop
-      end
-
-      ## 2) add missing (hierarchy) level if
-      while headings.size+1 < heading_level
-        ##  todo/fix: issue warning about "skipping" hierarchy level
-        puts "!!! warn [team reader] - skipping hierarchy level in headings "
-        headings.push( nil )
-      end
-
-      if heading =~ /^\?+$/    ## note: use ? or ?? or ?? to reset level to nil
-        ## keep level empty
-      else
-        ## note: if level is 1 assume country for now
-        if heading_level == 1
-            ## assume country in heading; allow all "formats" supported by parse e.g.
-            ##   Österreich • Austria (at)
-            ##   Österreich • Austria
-            ##   Austria
-            ##   Deutschland (de) • Germany
-            country = catalog.countries.parse( heading )
-            ## check country code - MUST exist for now!!!!
-            if country.nil?
-              puts "!!! error [club reader] - unknown country >#{heading}< - sorry - add country to config to fix"
-              exit 1
-            end
-
-            headings.push( country.key )
-        else
-         ## quick hack:
-         ##   remove known fill/dummy words incl:
-         ##     Provincia San Juan  =>  San Juan   (see argentina, for example)
-         ##
-         ##   use geo tree long term with alternative names - why? why not?
-          words = ['Provincia']
-          words.each { |word| heading = heading.gsub( word, '' ) }
-          heading = heading.strip
-
-          headings.push( heading )
-        end
-
-        ## assert that hierarchy level is ok
-        if headings.size != heading_level
-          puts "!!! error - headings hierarchy/stack out of order - #{heading.size}<=>#{heading_level}"
-          exit 1
-        end
-      end
-
-      pp headings
-
-    elsif node[0] == :p   ## paragraph with (text) lines
-      lines = node[1]
-      lines.each do |line|
-      if line.start_with?( '|' )
-        ## assume continuation with line of alternative names
-        ##  note: skip leading pipe
-        values = line[1..-1].split( '|' )   # team names - allow/use pipe(|)
-
-        add_alt_names( last_rec, values )   ## note: use alt_names helper for (re)use
-
-      ## check for b (child) team / club marker e.g.
-      ##    (ii) or ii) or ii.) or (ii.)
-      ##    (b)  or b)  or b.)  or (b.)
-      ##    (2)  or 2)  or 2.)  or (2.)
-      elsif line =~ B_TEAM_MARKER_RE
-         line = line.sub( B_TEAM_MARKER_RE, '' ).strip   ## remove (leading) b team marker
-
-         ## todo/fix: move into "regular" club branch - (re)use, that is, use the same code
-         #                                                for both a and b team / club
-         rec = Club.new
-         value = line    ## note: assume / allow just canonical name for now
-         ## strip and  squish (white)spaces
-         #   e.g. New York FC      (2011-)  => New York FC (2011-)
-         value = value.gsub( '$', '' ).strip
-                      .gsub( /[ \t]+/, ' ' )
-
-         rec.name = value            # canoncial name (global unique "beautiful/long" name)
-         rec.add_variants( value )   # auto-add (possible) auto-generated variant names
-
-         ### link a and b team / clubs
-         ##   assume last_rec is the a team
-         ##   todo/fix: check last_rec required NOT null
-         rec.a      = last_rec
-         last_rec.b = rec
-
-         last_rec = rec
-         recs << rec
-
-       ## check for address line e.g.
-       ##    use just one style / syntax - why? why not?
-       ##  Fischhofgasse 12 ~ 1100 Wien or
-       ##  Fischhofgasse 12 // 1100 Wien or Fischhofgasse 12 /// 1100 Wien
-       ##  Fischhofgasse 12 ++ 1100 Wien or Fischhofgasse 12 +++ 1100 Wien
-       elsif line =~ ADDR_MARKER_RE
-         # note skip for now!!!
-         # todo/fix: add support for address line!!!
-         puts "  skipping address line for now >#{line}<"
-      else
-        values = line.split( ',' )
-
-        rec = Club.new
-
-        col  = values.shift    ## get first item
-        ## note: allow optional alt names for convenience with required canoncial name
-        names = col.split( '|' )   # team names - allow/use pipe(|)
-        value     = names[0]         ## canonical name
-        alt_names = names[1..-1]     ## optional (inline) alt names
-
-        ## strip and  squish (white)spaces
-        #   e.g. New York FC      (2011-)  => New York FC (2011-)
-        value = value.gsub( '$', '' ).strip
-                     .gsub( /[ \t]+/, ' ' )
-        rec.name = value            # canoncial name (global unique "beautiful/long" name)
-        rec.add_variants( value )   # auto-add (possible) auto-generated variant names
-
-        ## note: add optional (inline) alternate names if present
-        add_alt_names( rec, alt_names )   if alt_names.size > 0
-
-        ## note:
-        ##   check/todo!!!!!!!!!!!!!!!!!-
-        ##  strip year if to present e.g. (2011-)
-        ##
-        ##  do NOT strip for defunct / historic clubs e.g.
-        ##    (1899-1910)
-        ## or (-1914) or (-2011) etc.
-
-        ###
-        ##  todo: move year out of canonical team name - why? why not?
-
-        ## check if canonical name include (2011-) or similar in name
-        ##   if yes, remove (2011-) and add to (alt) names
-        ##   e.g. New York FC (2011) => New York FC
-        if rec.name =~ /\(.+?\)/   ## note: use non-greedy (?) match
-          name = rec.name.gsub( /\(.+?\)/, '' ).strip
-
-          if rec.name =~ /\(([0-9]{4})-\)/            ## e.g. (2014-)
-            rec.year     = $1.to_i
-          elsif  rec.name =~ /\(-([0-9]{4})\)/            ## e.g. (-2014)
-            rec.year_end = $1.to_i
-          elsif  rec.name =~ /\(([0-9]{4})-([0-9]{4})\)/  ## e.g. (2011-2014)
-            rec.year     = $1.to_i
-            rec.year_end = $2.to_i
-          else
-            ## todo/check: warn about unknown year format
-          end
-        end
-
-        ##  todo/check - check for unknown format values
-        ##    e.g. too many values, duplicate years, etc.
-        ##         check for overwritting, etc.
-        while values.size > 0
-          value = values.shift
-          ##  strip and squish (white)spaces
-          #   e.g. León     › Guanajuato     => León › Guanajuato
-          value = value.strip.gsub( /[ \t]+/, ' ' )
-          if value =~/^\d{4}$/   # e.g 1904
-            ## todo/check: issue warning if year is already set!!!!!!!
-            if rec.year
-              puts "!!! error - year already set to #{rec.year} - CANNOT overwrite with #{value}:"
-              pp rec
-              exit 1
-            end
-            rec.year  = value.to_i
-          elsif value.start_with?( '@' )   # e.g. @ Anfield
-            ## cut-off leading @ and spaces
-            rec.ground  = value[1..-1].strip
-          else
-            ## assume city / geo tree
-            ## split into geo tree
-            geos = split_geo( value )
-            city = geos[0]
-            ## check for "embedded" district e.g. London (Fulham) or Hamburg (St. Pauli) etc.
-            if city =~ /\((.+?)\)/   ## note: use non-greedy (?) match
-              rec.district  = $1.strip
-              city          = city.gsub( /\(.+?\)/, '' ).strip
-            end
-            rec.city = city
-
-            if geos.size > 1
-               ## cut-off city and keep the rest (of geo tree)
-               rec.geos = geos[1..-1]
-            end
-          end
-        end  ## while values
-
-
-        ###############
-        ## use headings text for geo tree
-
-        ## 1) add country if present
-        if headings.size > 0 && headings[0]
-          country = catalog.countries.find( headings[0] )
-          rec.country = country
-        else
-          ## make it an error - why? why not?
-          puts "!!! error - country missing in headings hierarchy - sorry - add to quicklist"
-          exit 1
-        end
-
-        ## 2) check geo tree with headings hierarchy
-        if headings.size > 1 && headings[1]
-           geos = split_geo( headings[1] )
-           if rec.geos
-             if rec.geos[0] != geos[0]
-               puts "!!! error - geo tree - headings mismatch >#{rec.geos[0]}< <=> >#{geos[0]}<"
-               exit 1
-             end
-             if rec.geos[1] && rec.geos[1] != geos[1]   ## check optional 2nd level too
-               puts "!!! error - geo tree - headings mismatch >#{rec.geos[1]}< <=> >#{geos[1]}<"
-               exit 1
-             end
-           else
-             ## add missing region (state/province) from headings hierarchy
-             rec.geos = geos
-           end
-        end
-
-        last_rec = rec
-
-
-      ### todo/fix:
-      ##  auto-add alt name with dots stripped - why? why not?
-      ##    e.g.  D.C. United    => DC United
-      ##    e.g.  Liverpool F.C. => Liverpool FC
-      ##    e.g.  St. Albin       => St Albin etc.
-      ##    e.g.  1. FC Köln     => 1 FC Köln  -- make special case for 1. - why? why not?
-
-      ##
-      ## todo/fix:  unify mapping entries
-      ##   always lowercase !!!!  (case insensitive)
-      ##   always strip (2011-) !!!
-      ##   always strip dots (e.g. St., F.C, etc.)
-
-        recs << rec
-      end
-      end  # each line (in paragraph)
-    else
-      puts "** !!! ERROR !!! [club reader] - unknown line type:"
-      pp node
-      exit 1
-    end
-  end
-
-  recs
-end  # method read
-
-#######################
-###  helpers
-def split_geo( str )
-  ## assume city / geo tree
-  ##  strip and squish (white)spaces
-  #   e.g. León     › Guanajuato     => León › Guanajuato
-  str = str.strip.gsub( /[ \t]+/, ' ' )
-
-  ## split into geo tree
-  geos = str.split( /[<>‹›]/ )   ## note: allow > < or › ‹
-  geos = geos.map { |geo| geo.strip }   ## remove all whitespaces
-  geos
-end
-
-end  # class ClubReader
-
-
-end ## module Import
-end ## module SportDb
+
+module SportDb
+module Import
+
+
+class ClubReader
+
+  def world() Import.world; end
+
+
+
+def self.read( path )   ## use - rename to read_file or from_file etc. - why? why not?
+  txt = File.open( path, 'r:utf-8' ) { |f| f.read }
+  parse( txt )
+end
+
+def self.parse( txt )
+  new( txt ).parse
+end
+
+def initialize( txt )
+  @txt = txt
+end
+
+##  pattern for b (child) team / club marker e.g.
+##    (ii) or ii) or ii.) or (ii.) or (II)
+##    (b)  or b)  or b.)  or (b.)  or (B)
+##    (2)  or 2)  or 2.)  or (2.)
+B_TEAM_MARKER_RE = %r{^  \(?     # optional opening bracket
+                          (?: ii|b|2 )
+                          \.?     # optional dot - keep and allow dot - why? why not?
+                          \)      # required closing bracket
+                     }xi   ## note: add case-insenstive (e.g. II/ii or B/b)
+
+## pattern for checking for address line e.g.
+##    use just one style / syntax - why? why not?
+##  Fischhofgasse 12 ~ 1100 Wien or
+##  Fischhofgasse 12 // 1100 Wien   or Fischhofgasse 12 /// 1100 Wien
+##  Fischhofgasse 12 ++ 1100 Wien   or Fischhofgasse 12 +++ 1100 Wien
+ADDR_MARKER_RE =  %r{ (?: ^|[ ] )                # space or beginning of line
+                        (?: ~ | /{2,} | \+{2,} )
+                      (?: [ ]|$)                 # space or end of line
+                       }x
+
+
+
+def parse
+  recs = []
+  last_rec  = nil
+  headings = []   ## headings stack
+
+  OutlineReader.parse( @txt ).each do |node|
+    if [:h1,:h2,:h3,:h4,:h5,:h6].include?( node[0] )
+      heading_level  = node[0][1].to_i
+      heading        = node[1]
+
+      puts "heading #{heading_level} >#{heading}<"
+
+      ## 1) first pop headings if present
+      while headings.size+1 > heading_level
+        headings.pop
+      end
+
+      ## 2) add missing (hierarchy) level if
+      while headings.size+1 < heading_level
+        ##  todo/fix: issue warning about "skipping" hierarchy level
+        puts "!!! warn [team reader] - skipping hierarchy level in headings "
+        headings.push( nil )
+      end
+
+      if heading =~ /^\?+$/    ## note: use ? or ?? or ?? to reset level to nil
+        ## keep level empty
+      else
+        ## note: if level is 1 assume country for now
+        if heading_level == 1
+            ## assume country in heading; allow all "formats" supported by parse e.g.
+            ##   Österreich • Austria (at)
+            ##   Österreich • Austria
+            ##   Austria
+            ##   Deutschland (de) • Germany
+            country = world.countries.parse( heading )
+            ## check country code - MUST exist for now!!!!
+            if country.nil?
+              puts "!!! error [club reader] - unknown country >#{heading}< - sorry - add country to config to fix"
+              exit 1
+            end
+
+            headings.push( country.key )
+        else
+         ## quick hack:
+         ##   remove known fill/dummy words incl:
+         ##     Provincia San Juan  =>  San Juan   (see argentina, for example)
+         ##
+         ##   use geo tree long term with alternative names - why? why not?
+          words = ['Provincia']
+          words.each { |word| heading = heading.gsub( word, '' ) }
+          heading = heading.strip
+
+          headings.push( heading )
+        end
+
+        ## assert that hierarchy level is ok
+        if headings.size != heading_level
+          puts "!!! error - headings hierarchy/stack out of order - #{heading.size}<=>#{heading_level}"
+          exit 1
+        end
+      end
+
+      pp headings
+
+    elsif node[0] == :p   ## paragraph with (text) lines
+      lines = node[1]
+      lines.each do |line|
+      if line.start_with?( '|' )
+        ## assume continuation with line of alternative names
+        ##  note: skip leading pipe
+        values = line[1..-1].split( '|' )   # team names - allow/use pipe(|)
+        values = values.map {|value| _norm(value) }  ## squish/strip etc.
+
+        last_rec.alt_names += values 
+
+      ## check for b (child) team / club marker e.g.
+      ##    (ii) or ii) or ii.) or (ii.)
+      ##    (b)  or b)  or b.)  or (b.)
+      ##    (2)  or 2)  or 2.)  or (2.)
+      elsif line =~ B_TEAM_MARKER_RE
+         line = line.sub( B_TEAM_MARKER_RE, '' ).strip   ## remove (leading) b team marker
+
+         ## todo/fix: move into "regular" club branch - (re)use, that is, use the same code
+         #                                                for both a and b team / club
+         rec = Club.new
+         ## note: assume / allow just canonical name for now
+         ## strip and  squish (white)spaces
+         #   e.g. New York FC      (2011-)  => New York FC (2011-)
+         rec.name = _norm( line )            # canoncial name (global unique "beautiful/long" name)
+
+         ### link a and b team / clubs
+         ##   assume last_rec is the a team
+         ##   todo/fix: check last_rec required NOT null
+         rec.a      = last_rec
+         last_rec.b = rec
+
+         last_rec = rec
+         recs << rec
+
+       ## check for address line e.g.
+       ##    use just one style / syntax - why? why not?
+       ##  Fischhofgasse 12 ~ 1100 Wien or
+       ##  Fischhofgasse 12 // 1100 Wien or Fischhofgasse 12 /// 1100 Wien
+       ##  Fischhofgasse 12 ++ 1100 Wien or Fischhofgasse 12 +++ 1100 Wien
+      elsif line =~ ADDR_MARKER_RE
+         last_rec.address = _squish( line )
+      else
+        values = line.split( ',' )
+
+        rec = Club.new
+
+        col  = values.shift    ## get first item
+        ## note: allow optional alt names for convenience with required canoncial name
+        names = col.split( '|' )   # team names - allow/use pipe(|)
+        names = names.map {|name| _norm(name) }  ## squish/strip etc.
+
+        value     = names[0]         ## canonical name
+        alt_names = names[1..-1]     ## optional (inline) alt names
+
+        rec.name = value            # canoncial name (global unique "beautiful/long" name)
+        ## note: add optional (inline) alternate names if present
+        rec.alt_names += alt_names    if alt_names.size > 0
+
+        ## note:
+        ##   check/todo!!!!!!!!!!!!!!!!!-
+        ##  strip year if to present e.g. (2011-)
+        ##
+        ##  do NOT strip for defunct / historic clubs e.g.
+        ##    (1899-1910)
+        ## or (-1914) or (-2011) etc.
+
+        ###
+        ##  todo: move year out of canonical team name - why? why not?
+
+        ## check if canonical name include (2011-) or similar in name
+        ##   if yes, remove (2011-) and add to (alt) names
+        ##   e.g. New York FC (2011) => New York FC
+        if rec.name =~ /\(.+?\)/   ## note: use non-greedy (?) match
+          name = rec.name.gsub( /\(.+?\)/, '' ).strip
+
+          if rec.name =~ /\(([0-9]{4})-\)/            ## e.g. (2014-)
+            rec.year     = $1.to_i
+          elsif  rec.name =~ /\(-([0-9]{4})\)/            ## e.g. (-2014)
+            rec.year_end = $1.to_i
+          elsif  rec.name =~ /\(([0-9]{4})-([0-9]{4})\)/  ## e.g. (2011-2014)
+            rec.year     = $1.to_i
+            rec.year_end = $2.to_i
+          else
+            ## todo/check: warn about unknown year format
+          end
+        end
+
+        ##  todo/check - check for unknown format values
+        ##    e.g. too many values, duplicate years, etc.
+        ##         check for overwritting, etc.
+
+        ##  strip and squish (white)spaces
+        #   e.g. León     › Guanajuato     => León › Guanajuato
+        values = values.map {|value| _squish(value) }
+
+        while values.size > 0
+          value = values.shift
+          if value =~/^\d{4}$/   # e.g 1904
+            ## todo/check: issue warning if year is already set!!!!!!!
+            if rec.year
+              puts "!!! error - year already set to #{rec.year} - CANNOT overwrite with #{value}:"
+              pp rec
+              exit 1
+            end
+            rec.year  = value.to_i
+          elsif value.start_with?( '@' )   # e.g. @ Anfield
+            ## cut-off leading @ and spaces
+            rec.ground  = value[1..-1].strip
+          else
+            ## assume city / geo tree
+            ## split into geo tree
+            geos = split_geo( value )
+            city = geos[0]
+            ## check for "embedded" district e.g. London (Fulham) or Hamburg (St. Pauli) etc.
+            if city =~ /\((.+?)\)/   ## note: use non-greedy (?) match
+              rec.district  = $1.strip
+              city          = city.gsub( /\(.+?\)/, '' ).strip
+            end
+            rec.city = city
+
+            if geos.size > 1
+               ## cut-off city and keep the rest (of geo tree)
+               rec.geos = geos[1..-1]
+            end
+          end
+        end  ## while values
+
+
+        ###############
+        ## use headings text for geo tree
+
+        ## 1) add country if present
+        if headings.size > 0 && headings[0]
+          country = world.countries.find( headings[0] )
+          rec.country = country
+        else
+          ## make it an error - why? why not?
+          puts "!!! error - country missing in headings hierarchy - sorry - add to quicklist"
+          exit 1
+        end
+
+        ## 2) check geo tree with headings hierarchy
+        if headings.size > 1 && headings[1]
+           geos = split_geo( headings[1] )
+           if rec.geos
+             if rec.geos[0] != geos[0]
+               puts "!!! error - geo tree - headings mismatch >#{rec.geos[0]}< <=> >#{geos[0]}<"
+               exit 1
+             end
+             if rec.geos[1] && rec.geos[1] != geos[1]   ## check optional 2nd level too
+               puts "!!! error - geo tree - headings mismatch >#{rec.geos[1]}< <=> >#{geos[1]}<"
+               exit 1
+             end
+           else
+             ## add missing region (state/province) from headings hierarchy
+             rec.geos = geos
+           end
+        end
+
+        last_rec = rec
+
+        recs << rec
+      end
+      end  # each line (in paragraph)
+    else
+      puts "** !!! ERROR !!! [club reader] - unknown line type:"
+      pp node
+      exit 1
+    end
+  end
+
+  recs
+end  # method read
+
+#######################
+###  helpers
+def split_geo( str )
+  ## assume city / geo tree
+  ##  strip and squish (white)spaces
+  #   e.g. León     › Guanajuato     => León › Guanajuato
+  str = _squish( str ) 
+  
+  ## split into geo tree
+  geos = str.split( /[<>‹›]/ )   ## note: allow > < or › ‹
+  geos = geos.map { |geo| geo.strip }   ## remove all whitespaces
+  geos
+end
+
+
+## norm(alize) helper  - squish (spaces) 
+##                      and remove dollars ($$$)
+##                      and remove leading and trailing spaces
+def _norm( str )
+  ## only extra clean-up of dollars for now ($$$)
+  _squish( str.gsub( '$', '' ) )
+end
+
+def _squish( str )
+  str.gsub( /[ \t\u00a0]+/, ' ' ).strip
+end
+
+
+end  # class ClubReader
+
+
+end ## module Import
+end ## module SportDb