sportdb-formats 2.0.1 → 2.1.0

data/lib/sportdb/formats/csv/match_status_parser.rb

@@ -1,90 +0,0 @@
-#####################
-# helpers for parsing & finding match status e.g.
-#    - cancelled / canceled
-#    - awarded
-#    - abandoned
-#    - replay
-#    etc.
-
-
-module SportDb
-
-
-  ### todo/fix: move Status inside Match struct - why? why not?
-
-  class Status
-# note: use a class as an "enum"-like namespace for now - why? why not?
-#   move class into Match e.g. Match::Status  - why? why not?
-    CANCELLED = 'CANCELLED'   # canceled (US spelling), cancelled (UK spelling) - what to use?
-    AWARDED   = 'AWARDED'
-    POSTPONED = 'POSTPONED'
-    ABANDONED = 'ABANDONED'
-    REPLAY    = 'REPLAY'
-  end # class Status
-
-
-
-  class StatusParser
-
-    def self.parse( str )
-      ## note: returns nil if no match found
-      ## note: english usage - cancelled (in UK), canceled (in US)
-      if str =~ /^(cancelled|
-                   canceled|
-                   can\.
-                  )/xi
-        Status::CANCELLED
-      elsif str =~ /^(awarded|
-                       awd\.
-                      )/xi
-        Status::AWARDED
-      elsif str =~ /^(postponed
-                      )/xi
-        Status::POSTPONED
-      elsif str =~ /^(abandoned|
-                       abd\.
-                      )/xi
-        Status::ABANDONED
-      elsif str =~ /^(replay
-                      )/xi
-        Status::REPLAY
-      else
-        # no match
-        nil
-      end
-    end
-
-
-    RUN_RE = /\[
-                 (?<text>[^\]]+)
-               \]
-             /x
-    def self.find!( line )
-      ## for now check all "protected" text run blocks e.g. []
-      ##  puts "line: >#{line}<"
-
-      status = nil
-
-      str = line
-      while m = str.match( RUN_RE )
-        str = m.post_match  ## keep on processing rest of line/str (a.k.a. post match string)
-
-        ## check for status match
-        match_str = m[0]  ## keep a copy of the match string (for later sub)
-        text = m[:text].strip
-        ## puts "  text: >#{text}<"
-
-        status = parse( text )
-
-        if status
-           line.sub!( match_str, "[STATUS.#{status}]" )
-           break
-        end
-      end  # while match
-
-      status
-    end # method find!
-  end # class StatusParser
-
-end # module SportDb
-

data/lib/sportdb/formats/datafile.rb

@@ -1,59 +0,0 @@
-# encoding: utf-8
-
-
-module Datafile      # note: keep Datafile in its own top-level module/namespace for now - why? why not?
-
-  def self.read( path )   ## todo/check: use as a shortcut helper - why? why not?
-     ## note: always assume utf-8 for now!!!
-     File.open( path, 'r:utf-8') {|f| f.read }
-  end
-
-
-  ########################
-  ## todo/fix: turn into Datafile::Bundle.new  and Bundle#write/save -why? why not?
-  class Bundle
-    def initialize( path )
-      @path = path
-      @buf  = String.new('')
-    end
-
-    def <<(value)
-       if value.is_a?( Array )  ## assume array of datafiles (file paths)
-         datafiles = value
-         datafiles.each do |datafile|
-           text = Datafile.read( datafile )
-           ## todo/fix/check:  move  sub __END__ to Datafile.read and turn it always on  -  why? why not?
-           text = text.sub( /__END__.*/m, '' )    ## note: add/allow support for __END__; use m-multiline flag
-           @buf << text
-           @buf << "\n\n"
-          end
-        else ## assume string (e.g. header, comments, etc.)
-          text = value
-          @buf << text
-          @buf << "\n\n"
-        end
-    end
-    alias_method :write, :<<
-
-    ## todo/fix/check: write only on close? or write on every write and use close for close?
-    def close
-      File.open( @path, 'w:utf-8' ) do |f|
-        f.write @buf
-      end
-    end
-  end  # class Bundle
-
-
-  def self.write_bundle( path, datafiles:, header: nil )
-    bundle = Bundle.new( path )
-    bundle.write( header )   if header
-    datafiles.each do |datafile|
-      text = read( datafile )
-      ## todo/fix/check:  move  sub __END__ to Datafile.read and turn it always on  -  why? why not?
-      text = text.sub( /__END__.*/m, '' )    ## note: add/allow support for __END__; use m-multiline flag
-      bundle.write( text )
-    end
-    bundle.close
-  end
-
-end # module Datafile

data/lib/sportdb/formats/event/event_reader.rb

@@ -1,119 +0,0 @@
-
-module SportDb
-module Import
-  class EventInfoReader
-    def catalog() Import.catalog; end
-
-
-    def self.read( path )
-      txt = File.open( path, 'r:utf-8') {|f| f.read }
-      new( txt ).parse
-    end
-
-    def self.parse( txt )
-      new( txt ).parse
-    end
-
-    def initialize( txt )
-      @txt = txt
-    end
-
-    def parse
-      recs = []
-
-      parse_csv( @txt ).each do |row|
-        league_col = row['League']
-        season_col = row['Season'] || row['Year']
-        dates_col  = row['Dates']
-
-        season = Season.parse( season_col )
-        league = catalog.leagues.find!( league_col )
-
-
-        dates = []
-        if dates_col.nil? || dates_col.empty?
-           ## do nothing; no dates - keep dates array empty
-        else
-          ## squish spaces
-          dates_col = dates_col.gsub( /[ ]{2,}/, ' ' )  ## squish/fold spaces
-
-          puts "#{league.name} (#{league.key}) | #{season.key} | #{dates_col}"
-
-          ### todo/check:  check what parts "Aug 15" return ???
-          ###                       short form for "Aug 15 -" - works?
-
-          ## todo/fix!!! - check EventInfo.include?
-          ##    now allow dates with only start_date too!! (WITHOUT end_date)
-          parts = dates_col.split( /[ ]*[–-][ ]*/ )
-          if parts.size == 1
-            pp parts
-            dates << DateFormats.parse( parts[0], start: Date.new( season.start_year, 1, 1 ), lang: 'en' )
-            pp dates
-          elsif parts.size == 2
-            pp parts
-            dates << DateFormats.parse( parts[0], start: Date.new( season.start_year, 1, 1 ), lang: 'en' )
-            dates << DateFormats.parse( parts[1], start: Date.new( season.end_year ? season.end_year : season.start_year, 1, 1 ), lang: 'en' )
-            pp dates
-
-            ## assert/check if period is less than 365 days for now
-            diff = dates[1].to_date.jd - dates[0].to_date.jd
-            puts "#{diff}d"
-            if diff > 365
-              puts "!! ERROR - date range / period assertion failed; expected diff < 365 days"
-              exit 1
-            end
-          else
-            puts "!! ERRROR - expected data range / period - one or two dates; got #{parts.size}:"
-            pp dates_col
-            pp parts
-            exit 1
-          end
-        end
-
-
-        teams_col    = row['Clubs'] || row['Teams']
-        goals_col    = row['Goals']
-
-        ## note: remove (and allow) all non-digits e.g. 370 goals, 20 clubs, etc.
-        teams_col    = teams_col.gsub( /[^0-9]/, '' )    if teams_col
-        goals_col    = goals_col.gsub( /[^0-9]/, '' )    if goals_col
-
-        teams  =   (teams_col.nil? || teams_col.empty?) ? nil : teams_col.to_i
-        goals  =   (goals_col.nil? || goals_col.empty?) ? nil : goals_col.to_i
-
-        matches_col  = row['Matches']
-        ## note: support additions in matches (played) e.g.
-        #   132 + 63 Play-off-Spiele
-        matches_col   = matches_col.gsub( /[^0-9+]/, '' )  if matches_col
-
-        matches =  if matches_col.nil? || matches_col.empty?
-                     nil
-                   else
-                     if matches_col.index( '+' )  ### check for calculations
-                        ## note: for now only supports additions
-                        matches_col.split( '+' ).reduce( 0 ) do |sum,str|
-                          sum + str.to_i
-                        end
-                     else            ## assume single (integer) number
-                        matches_col.to_i
-                     end
-                  end
-
-        rec = EventInfo.new( league:     league,
-                             season:     season,
-                             start_date: dates[0],
-                             end_date:   dates[1],
-                             teams:      teams,
-                             matches:    matches,
-                             goals:      goals
-                           )
-        recs << rec
-      end  # each row
-      recs
-    end  # method parse
-  end  # class EventInfoReader
-
-
-end ## module Import
-end ## module SportDb
-

data/lib/sportdb/formats/ground/ground_reader.rb

@@ -1,289 +0,0 @@
-###
-#  todo -   based on ClubReader
-#               share GeoReader or BaseReader or such for both
-#                  plus maybe for PlayerReader too!!!
-#
-#  fix/todo/cleanup  - move alt_names_auto from reader to indexer!!!!
-#                          indexer now handles unaccent (variants) etc.
-
-module SportDb
-module Import
-
-
-class GroundReader
-
-  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 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,} )
-                      (?: [ ]|$)                 # space or end of line
-                       }x
-
-
-def parse
-  recs = []
-  last_rec  = nil
-  headings = []   ## headings stack
-
-  OutlineReader.parse( @txt ).each do |node|
-    if [:h1,:h2,:h3,:h4,:h5,:h6].include?( node[0] )
-      heading_level  = node[0][1].to_i
-      heading        = node[1]
-
-      puts "heading #{heading_level} >#{heading}<"
-
-      ## 1) first pop headings if present
-      while headings.size+1 > heading_level
-        headings.pop
-      end
-
-      ## 2) add missing (hierarchy) level if
-      while headings.size+1 < heading_level
-        ##  todo/fix: issue warning about "skipping" hierarchy level
-        puts "!!! warn [ground reader] - skipping hierarchy level in headings "
-        headings.push( nil )
-      end
-
-      if heading =~ /^\?+$/    ## 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 [ground 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 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
-        ## squish line here - why? why not?
-        last_rec.address = _squish( line )
-      else
-        values = line.split( ',' )
-
-        rec = Ground.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 =~/^[0-9_]+$/   # e.g 1904
-            ##  skip capacity for now
-          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 !!! [ground 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 GroundReader
-
-
-end ## module Import
-end ## module SportDb