Class: SportDb::AutoConfParser

Inherits:
Object
  • Object
show all
Includes:
Logging, ParserHelper
Defined in:
lib/sportdb/formats/match/match_parser_auto_conf.rb

Overview

todo/check: rename/change to MatchAutoConfParser - why? why not?

Class Method Summary collapse

Instance Method Summary collapse

Methods included from ParserHelper

#is_goals?, #is_group?, #is_group_def?, #is_knockout_round?, #is_round?, #is_round_def?, #read_lines

Constructor Details

#initialize(lines, start) ⇒ AutoConfParser

Returns a new instance of AutoConfParser.



20
21
22
23
24
25
26
 
# File 'lib/sportdb/formats/match/match_parser_auto_conf.rb', line 20

def initialize( lines, start )
  # for convenience split string into lines
  ##    note: removes/strips empty lines
  ## todo/check: change to text instead of array of lines - why? why not?
  @lines        = lines.is_a?( String ) ? read_lines( lines ) : lines
  @start        = start
end

Class Method Details

.parse(lines, start:) ⇒ Object 



8
9
10
11
12
13
 
# File 'lib/sportdb/formats/match/match_parser_auto_conf.rb', line 8

def self.parse( lines, start: )
  ##  todo/fix: add support for txt and lines
  ##    check if lines_or_txt is an array or just a string
  parser = new( lines, start )
  parser.parse
end

Instance Method Details

#find_date!(line, start:) ⇒ Object 



192
193
194
195
196
197
198
199
200
 
# File 'lib/sportdb/formats/match/match_parser_auto_conf.rb', line 192

def find_date!( line, start: )
  ## NB: lets us pass in start_at/end_at date (for event)
  #   for auto-complete year

  # extract date from line
  # and return it
  # NB: side effect - removes date from line string
  DateFormats.find!( line, start: start )
end

#find_score!(line) ⇒ Object 



185
186
187
188
189
190
 
# File 'lib/sportdb/formats/match/match_parser_auto_conf.rb', line 185

def find_score!( line )
  # note: always call after find_dates !!!
  #  scores match date-like patterns!!  e.g. 10-11  or 10:00 etc.
  #   -- note: score might have two digits too
  ScoreFormats.find!( line )
end

#parse ⇒ Object 



28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
 
# File 'lib/sportdb/formats/match/match_parser_auto_conf.rb', line 28

def parse
  ## try to  find all teams in match schedule
  @last_round   = nil
  @last_group   = nil

  ## definitions/defs
  @round_defs = Hash.new(0)
  @group_defs = Hash.new(0)

  ## usage/refs
  @rounds       = {}           ## track usage counter and match (two teams) counter
  @groups       = {}           ##  -"-
  @teams        = Hash.new(0)   ## keep track of usage counter

  @warns        = []    ## track list of warnings (unmatched lines)  too - why? why not?


  @lines.each do |line|
    if is_goals?( line )
      logger.debug "skipping matched goals line: >#{line}<"
    elsif is_round_def?( line )
      ## todo/fix:  add round definition (w begin n end date)
      ## todo: do not patch rounds with definition (already assume begin/end date is good)
      ##  -- how to deal with matches that get rescheduled/postponed?
      logger.debug "skipping matched round def line: >#{line}<"
      @round_defs[ line ] += 1
    elsif is_round?( line )
      logger.debug "skipping matched round line: >#{line}<"

      round = @rounds[ line ] ||= {count: 0, match_count: 0}   ## usage counter, match counter
      round[:count] +=1
      @last_round = round
    elsif is_group_def?( line ) ## NB: group goes after round (round may contain group marker too)
      ### todo: add pipe (|) marker (required)
      logger.debug "skipping matched group def line: >#{line}<"
      @group_defs[ line ] += 1
    elsif is_group?( line )
      ##  -- lets you set group  e.g. Group A etc.
      logger.debug "skipping matched group line: >#{line}<"

      group = @groups[ line ] ||= {count: 0, match_count: 0}
      group[:count] +=1
      @last_group = group
      ## todo/fix:  parse group line!!!
    elsif try_parse_game( line )
      # do nothing here
    else
      logger.warn "skipping line (no match found): >#{line}<"
      @warns << line
    end
  end # lines.each

  [@teams, @rounds, @groups, @round_defs, @group_defs, @warns]
end

#parse_game(line) ⇒ Object 



90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
 
# File 'lib/sportdb/formats/match/match_parser_auto_conf.rb', line 90

def parse_game( line )
  logger.debug "parsing game (fixture) line: >#{line}<"

  ## remove all protected text runs e.g. []
  ##   fix: add [ to end-of-line too
  ##  todo/fix: move remove protected text runs AFTER find date!! - why? why not?

  line = line.gsub( /\[
                      [^\]]+?
                     \]/x, '' ).strip
  return true if line.empty?    ## note: return true (for valid line with no match/teams)


  ## split by geo (@) - remove for now
  values = line.split( '@' )
  line = values[0]


  ## try find date
  date = find_date!( line, start: @start )
  if date   ## if found remove tagged run too; note using singular sub (NOT global gsub)
    line = line.sub( /\[
                        [^\]]+?
                       \]/x, '' ).strip

  else
    ##  check for leading hours only e.g.  20.30 or 20:30 or 20h30 or 20H30 or 09h00
    ##   todo/fix: make language dependent (or move to find_date/hour etc.) - why? why not?
    line = line.sub(  %r{^           ## MUST be anchored to beginning of line
                          [012]?[0-9]
                          [.:hH]
                          [0-9][0-9]
                         (?=[ ])    ## must be followed by space for now (add end of line too - why? why not?)
                        }x, '' ).strip
  end

  return true if line.empty?    ## note: return true (for valid line with no match/teams)


  score = find_score!( line )

  logger.debug "  line: >#{line}<"

  line = line.sub( /\[
                      [^\]]+?
                     \]/x, '$$' )  # note: replace first score tag with $$
  line = line.gsub( /\[
                  [^\]]+?
                 \]/x, '' )    # note: replace/remove all other score tags with nothing

   ##  clean-up  remove all text run inside () or empty () too
   line = line.gsub( /\(
                   [^)]*?
                  \)/x, '' )


   ## check for more match separators e.g. - or vs for now
   line = line.sub( / \s+
                        (   -
                          | v
                          | vs\.?    # note: allow optional dot eg. vs.
                        )
                      \s+
                     /ix, '$$' )

   values = line.split( '$$' )
   values = values.map { |value| value.strip }        ## strip spaces
   values = values.select { |value| !value.empty? }   ## remove empty strings

   return true    if values.size == 0  ## note: return true (for valid line with no match/teams)

   if values.size == 1
     puts "(auto config) try matching teams separated by spaces (2+):"
     pp values

     values = values[0].split( /[ ]{2,}/ )
     pp values
   end

   return false   if values.size != 2

   puts "(auto config) try matching teams:"
   pp values

   @teams[ values[0] ] += 1    ## update usage counters
   @teams[ values[1] ] += 1

   @last_round[ :match_count ] += 1    if @last_round
   @last_group[ :match_count ] += 1    if @last_group

   true
end

#try_parse_game(line) ⇒ Object 



84
85
86
87
88
 
# File 'lib/sportdb/formats/match/match_parser_auto_conf.rb', line 84

def try_parse_game( line )
  # note: clone line; for possible test do NOT modify in place for now
  # note: returns true if parsed, false if no match
  parse_game( line.dup )
end