date-formats 0.1.1 → 0.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: c60c4e7804fd0a6a4b11e70c35daccfb6db9acbb
-  data.tar.gz: b78c75c1941eeafd29333b7872fdd62facb93cd1
+  metadata.gz: 81aad612afdba8f85e731f97c1bc119499a8211d
+  data.tar.gz: e013e9bdac91e1938149d395b946c535b2c75714
 SHA512:
-  metadata.gz: d9f1630097d9f0d1ee5e64963b156093969791ba6626c16421dafe46b33b74966a45875a830f5c61a23d8c76afd471b9715d84fa9de52e68c6ef8aefc0b829c6
-  data.tar.gz: 436c8c910c926d09126f0bd4142d3e1b783f6f3c83d6af18463504470d477f2886be119ff0e03709c892eeafa021ed038b2c84c263b1cd1aafc406357333b1a1
+  metadata.gz: 718656e27eab527b211337e3a9fef08030017d2b31f7c59fa89b2292df61275a8b9f1ca247e5f9a5eacad03782d83890cd1a37f5d2d6c5ff541ce0ef67c579bc
+  data.tar.gz: c9470e38606e1b909230e8e62187cb1ba2f29f4bd4d99a48fc8125445d80b08e7ac37e37c7d08e463739d7bca767526ab535e0f0016d1f8abc546e16a55cb551

data/lib/date-formats.rb

@@ -11,40 +11,26 @@ require 'logutils'
 ###
 # our own code
 require 'date-formats/version' # let version always go first
-require 'date-formats/source'
 require 'date-formats/reader'
+require 'date-formats/source'
 
 
 module DateFormats
 
+#############
+# helpers for building format regex patterns
+MONTH_EN       = build_names( MONTH_NAMES[:en] )
+# e.g. Jan|Feb|March|Mar|April|Apr|May|June|Jun|...
+DAY_EN         = build_names( DAY_NAMES[:en] )
+# e.g.
 
-MONTH_EN_LINES    = Reader.parse_month( Source::MONTH_EN )
-MONTH_EN_TO_MM    = build_map( MONTH_EN_LINES )
-MONTH_EN          = build_re( MONTH_EN_LINES )
-
-WEEKDAY_EN_LINES  = Reader.parse_weekday( Source::WEEKDAY_EN )
-WEEKDAY_EN        = build_re( WEEKDAY_EN_LINES )
-
-
-
-MONTH_FR_LINES    = Reader.parse_month( Source::MONTH_FR )
-MONTH_FR_TO_MM    = build_map( MONTH_FR_LINES )
-MONTH_FR          = build_re( MONTH_FR_LINES )
-
-WEEKDAY_FR_LINES  = Reader.parse_weekday( Source::WEEKDAY_FR )
-WEEKDAY_FR        = build_re( WEEKDAY_FR_LINES )
-
-
-
-MONTH_ES_LINES    = Reader.parse_month( Source::MONTH_ES )
-MONTH_ES_TO_MM    = build_map( MONTH_ES_LINES )
-MONTH_ES          = build_re( MONTH_ES_LINES )
-
-
+MONTH_FR       = build_names( MONTH_NAMES[:fr] )
+DAY_FR         = build_names( DAY_NAMES[:fr] )
 
-MONTH_DE_LINES    = Reader.parse_month( Source::MONTH_DE )
-MONTH_DE_TO_MM    = build_map( MONTH_DE_LINES )
-MONTH_DE          = build_re( MONTH_DE_LINES )
+MONTH_ES       = build_names( MONTH_NAMES[:es] )
+MONTH_PT       = build_names( MONTH_NAMES[:pt] )
+MONTH_DE       = build_names( MONTH_NAMES[:de] )
+MONTH_IT       = build_names( MONTH_NAMES[:it] )
 
 end  # module DateFormats
 

data/lib/date-formats/date.rb

@@ -1,154 +1,248 @@
 # encoding: utf-8
 
+
 module DateFormats
 
 
-class DateFinderBase
+  def self.lang
+    @@lang ||= :en            ## defaults to english (:en)
+  end
+  def self.lang=( value )
+    @@lang = value.to_sym    ## note: make sure lang is always a symbol for now (NOT a string)
+  end
 
-private
-  def calc_year( month, day, start: )  ## note: start required param for now on!!!
 
-    logger.debug "   [calc_year] ????-#{month}-#{day} -- start: #{start}"
+  def self.parser( lang: )  ## find parser
+    lang = lang.to_sym  ## note: make sure lang is always a symbol for now (NOT a string)
 
-    if month >= start.month
-      # assume same year as start_at event (e.g. 2013 for 2013/14 season)
-      start.year
-    else
-      # assume year+1 as start_at event (e.g. 2014 for 2013/14 season)
-      start.year+1
-    end
+    ## note: cache all "built-in" lang versions (e.g. formats == nil)
+    @@parser ||= {}
+    parser = @@parser[ lang ] ||= DateParser.new( lang: lang )
   end
 
+  def self.parse( line,
+                  lang:    self.class.lang,
+                  start:   Date.new( Date.today.year, 1, 1 )  ## note: default to current YYYY.01.01. if no start provided
+                )
+    parser( lang: lang ).parse( line, start: start )
+  end
+
+  def self.find!( line,
+                  lang:  self.class.lang,
+                  start: Date.new( Date.today.year, 1, 1 ) ## note: default to current YYYY.01.01. if no start provided
+                )
+    parser( lang: lang ).find!( line, start: start )
+  end
 
-  def parse_date_time( match_data, start: )
 
-    # convert regex match_data captures to hash
-    # - note: cannont use match_data like a hash (e.g. raises exception if key/name not present/found)
-    h = {}
-    # - note: do NOT forget to turn name into symbol for lookup in new hash (name.to_sym)
-    match_data.names.each { |name| h[name.to_sym] = match_data[name] }  # or use match_data.names.zip( match_data.captures )  - more cryptic but "elegant"??
 
-    ## puts "[parse_date_time] match_data:"
-    ## pp h
-    logger.debug "   [parse_date_time] hash: >#{h.inspect}<"
+  class DateParser
 
-    h[ :month ] = MONTH_EN_TO_MM[ h[:month_en] ]  if h[:month_en]
-    h[ :month ] = MONTH_ES_TO_MM[ h[:month_es] ]  if h[:month_es]
-    h[ :month ] = MONTH_FR_TO_MM[ h[:month_fr] ]  if h[:month_fr]
-    h[ :month ] = MONTH_DE_TO_MM[ h[:month_de] ]  if h[:month_de]
-    h[ :month ] = MONTH_IT_TO_MM[ h[:month_it] ]  if h[:month_it]
-    h[ :month ] = MONTH_PT_TO_MM[ h[:month_pt] ]  if h[:month_pt]
+    include LogUtils::Logging
 
-    month   = h[:month]
-    day     = h[:day]
-    year    = h[:year]     || calc_year( month.to_i, day.to_i, start: start ).to_s
+    def initialize( lang:,
+                    formats: nil, month_names: nil, day_names: nil
+                  )
+      @lang = lang.to_sym   ## note: make sure lang is always a symbol for now (NOT a string)
 
-    hours   = h[:hours]    || '00'   # default to 00:00 for HH:MM (hours:minutes)
-    minutes = h[:minutes]  || '00'
+      if formats
+        @formats = formats
+      else
+        @formats = FORMATS[ @lang ]
+        if @formats
+          month_names = MONTH_NAMES[ @lang ]
+          day_names   = DAY_NAMES[ @lang ]
+        else
+          ## fallback to english if lang not available
+          ##  todo/fix: add/issue warning!!!!!
+          @formats    = FORMATS[ :en ]
+          month_names = MONTH_NAMES[ :en ]
+          day_names   = DAY_NAMES[ :en ]
+        end
+      end
 
-    value = '%d-%02d-%02d %02d:%02d' % [year.to_i, month.to_i, day.to_i, hours.to_i, minutes.to_i]
-    logger.debug "   date: >#{value}<"
+      ## convert month_names and day_names to map if present
+      @month_names = month_names ? build_map( month_names ) : nil
+      @day_names   = day_names   ? build_map( day_names )   : nil
+    end
 
-    DateTime.strptime( value, '%Y-%m-%d %H:%M' )
-  end
 
-end  # class DateFinderBase
+    def parse( line, start: )
+      date = nil
+      @formats.each do |format|
+        re = format[0]
+        m = re.match( line )
+        if m
+          date = parse_matchdata( m, start: start )
+          break
+        end
+        # no match; continue; try next regex pattern
+      end
 
+      ## todo/fix - raise ArgumentError - invalid date; no format match found
+      date  # note: nil if no match found
+    end
 
 
-class DateFinder < DateFinderBase
+    def find!( line, start: )
+      # fix: use more lookahead for all required trailing spaces!!!!!
+      # fix: use <name capturing group> for month,day,year etc.!!!
+
+      #
+      # fix: !!!!
+      #   date in [] will become [[DATE.DE4]] - when getting removed will keep ]!!!!
+      #   fix: change regex to \[[A-Z0-9.]\]  !!!!!!  plus add unit test too!!!
+      #
+
+      date = nil
+      @formats.each do |format|
+        re  = format[0]
+        tag = format[1]
+        m = re.match( line )
+        if m
+          date = parse_matchdata( m, start: start )
+          ## fix: use md[0] e.g. match for sub! instead of using regex again - why? why not???
+          ## fix: use md.begin(0), md.end(0)
+          line.sub!( m[0], tag )
+          ## todo/fix: make sure match data will not get changed (e.g. using sub! before parse_date_time)
+          break
+        end
+        # no match; continue; try next regex pattern
+      end
+      date  # note: nil if no match found
+    end
 
-  include LogUtils::Logging
+  private
+    def calc_year( month, day, start: )
 
+      logger.debug "   [calc_year] ????-#{month}-#{day} -- start: #{start}"
 
-  def self.lang()        @@lang ||= 'en';  end    ## defaults to english (en)
-  def self.lang=(value)  @@lang   = value; end
+      if month >= start.month
+        # assume same year as start_at event (e.g. 2013 for 2013/14 season)
+        start.year
+      else
+        # assume year+1 as start_at event (e.g. 2014 for 2013/14 season)
+        start.year+1
+      end
+    end
 
 
-  def initialize( lang: self.class.lang )
-    @lang    = lang.to_s
-    ## fallback to english if lang not available
-    ##  todo/fix: add/issue warning!!!!!
-    @formats = FORMATS[ @lang ] || FORMATS['en']
-  end
+    def parse_matchdata( m, start: )
+      # convert regex match_data captures to hash
+      # - note: cannont use match_data like a hash (e.g. raises exception if key/name not present/found)
+      h = {}
+      # - note: do NOT forget to turn name into symbol for lookup in new hash (name.to_sym)
+      m.names.each { |name| h[name.to_sym] = m[name] }  # or use match_data.names.zip( match_data.captures )  - more cryptic but "elegant"??
+
+      ## puts "[parse_date_time] match_data:"
+      ## pp h
+      logger.debug "   [parse_matchdata] hash: >#{h.inspect}<"
+
+      if h[:month_name]
+        ## todo/fix: issue error if no month names defined !!!
+        if @month_names
+         h[ :month ] = @month_names[ h[:month_name] ]
+        else
+          ## todo/fix: change to ArgumentError( "invalid date; ")
+          puts "** !!! ERROR !!! - no month names defined for lang #{@lang}; cannot match:"
+          pp m
+          exit 1
+        end
+      end
 
-  def find!( line, start_at: )     ## todo/fix: change start_at to start only!!!
-    # fix: use more lookahead for all required trailing spaces!!!!!
-    # fix: use <name capturing group> for month,day,year etc.!!!
-
-    #
-    # fix: !!!!
-    #   date in [] will become [[DATE.DE4]] - when getting removed will keep ]!!!!
-    #   fix: change regex to \[[A-Z0-9.]\]  !!!!!!  plus add unit test too!!!
-    #
-
-    m = nil
-    @formats.each do |format|
-      tag     = format[0]
-      pattern = format[1]
-      m=pattern.match( line )
-      if m
-        date = parse_date_time( m, start: start_at )
-        ## fix: use md[0] e.g. match for sub! instead of using regex again - why? why not???
-        ## fix: use md.begin(0), md.end(0)
-        line.sub!( m[0], tag )
-        ## todo/fix: make sure match data will not get changed (e.g. using sub! before parse_date_time)
-        return date
+      if h[:day_name]
+        if @day_names
+          ## note: use cwday  in ruby date to get days from 1-7 (Monday (1) to Sunday (7))
+          ##            wday  gives you 0-6 (Sunday (0), Monday (1) to Saturday (6))
+          h[ :cwday ] = @day_names[ h[:day_name ] ]
+        else
+          ## todo/fix: change to ArgumentError( "invalid date; ")
+          puts "** !!! ERROR !!! - no day names defined for lang #{@lang}; cannot match:"
+          pp m
+          exit 1
+        end
       end
-      # no match; continue; try next pattern
-    end
 
-    return nil  # no match found
-  end
+      month   = h[:month]
+      day     = h[:day]
+      year    = h[:year]     || calc_year( month.to_i, day.to_i, start: start ).to_s
+
+      if h[:hours] || h[:minutes]   ## check time (hours or minutes) is present (otherwise asume just Date and NOT DateTime)
+        hours   = h[:hours]    || '00'   # default to 00:00 for HH:MM (hours:minutes)
+        minutes = h[:minutes]  || '00'
 
-end # class DateFinder
-
-
-
-class RsssfDateFinder < DateFinderBase
-
-  include LogUtils::Logging
-
-  MONTH_EN = 'Jan|'+
-             'Feb|'+
-             'March|Mar|'+
-             'April|Apr|'+
-             'May|'+
-             'June|Jun|'+
-             'July|Jul|'+
-             'Aug|'+
-             'Sept|Sep|'+
-             'Oct|'+
-             'Nov|'+
-             'Dec'
-
-  ## e.g.
-  ##  [Jun 7]  or [Aug 12] etc.  - not MUST include brackets e.g. []
-  ##
-  ##  check add \b at the beginning and end - why?? why not?? working??
-  EN__MONTH_DD__DATE_RE = /\[
-                   (?<month_en>#{MONTH_EN})
-                      \s
-                   (?<day>\d{1,2})
-                     \]/x
-
-  def find!( line, start_at: )     ## todo/fix: change start_at to start only!!!
-    # fix: use more lookahead for all required trailing spaces!!!!!
-    # fix: use <name capturing group> for month,day,year etc.!!!
-
-    tag     = '[EN_MONTH_DD]'
-    pattern = EN__MONTH_DD__DATE_RE
-    m = pattern.match( line )
-    if m
-      date = parse_date_time( m, start: start_at )
-      ## fix: use md[0] e.g. match for sub! instead of using regex again - why? why not???
-      ## fix: use md.begin(0), md.end(0)
-      line.sub!( m[0], tag )
-      ## todo/fix: make sure match data will not get changed (e.g. using sub! before parse_date_time)
-      return date
+        value = '%d-%02d-%02d %02d:%02d' % [year.to_i, month.to_i, day.to_i, hours.to_i, minutes.to_i]
+        logger.debug "   datetime: >#{value}<"
+
+        date = DateTime.strptime( value, '%Y-%m-%d %H:%M' )
+      else
+        value = '%d-%02d-%02d' % [year.to_i, month.to_i, day.to_i]
+        logger.debug "   date: >#{value}<"
+
+        date = Date.strptime( value, '%Y-%m-%d' )
+      end
+
+      ## check/assert cwday if present!!!!
+      date
+    end  # method parse
+
+  ##################
+  # helpers
+  private
+    def build_map( lines )
+      ## build a lookup map that maps the word to the index (line no) plus 1 e.g.
+      ##   note: index is a string too
+      ##  {"January" => "1",  "Jan" => "1",
+      ##   "February" => "2", "Feb" => "2",
+      ##   "March" => "3",    "Mar" => "3",
+      ##   "April" => "4",    "Apr" => "4",
+      ##   "May" => "5",
+      ##   "June" => "6",     "Jun" => "6", ...
+      lines.each_with_index.reduce( {} ) do |h,(line,i)|
+        line.each { |name| h[ name ] = (i+1).to_s }  ## note: start mapping with 1 (and NOT zero-based, that is, 0)
+        h
+      end
     end
-    return nil  # no match found
-  end
-end  ## class RsssfDateFinder
+  end # class DateParser
+
+
+
+
+  class RsssfDateParser < DateParser
+
+    MONTH_NAMES = DateFormats.parse_month( <<TXT )
+Jan
+Feb
+March  Mar
+April  Apr
+May
+June   Jun
+July   Jul
+Aug
+Sept   Sep
+Oct
+Nov
+Dec
+TXT
+
+    MONTH_EN = DateFormats.build_names( MONTH_NAMES )    ## re helper e.g. Jan|Feb|March|Mar|...
+
+    ## e.g.
+    ##  [Jun 7]  or [Aug 12] etc.  - not MUST include brackets e.g. []
+    ##
+    ##  check add \b at the beginning and end - why?? why not?? working??
+    EN__MONTH_DD__DATE_RE = /\[
+                     (?<month_name>#{MONTH_EN})
+                        \s
+                     (?<day>\d{1,2})
+                       \]/x
+
+    def initialize
+      super( lang: 'en',
+             formats: [[EN__MONTH_DD__DATE_RE, '[EN_MONTH_DD]']],
+             month_names: MONTH_NAMES
+           )
+    end
+  end  ## class RsssfDateParser
 
 end # module DateFormats

data/lib/date-formats/formats.rb

@@ -18,7 +18,7 @@ DB__DATE_TIME_RE = /\b
                (?<minutes>\d{2})
                 \b/x
 
-# e.g. 2012-09-14  w/ implied hours (set to 12:00)
+# e.g. 2012-09-14    => YYYY-MM-DD
 #  note: allow 2012-9-3 e.g. no leading zero required
 # regex_db2
 DB__DATE_RE = /\b
@@ -61,7 +61,7 @@ DD_MM__DATE_TIME_RE = /\b
                       (?<minutes>\d{2})
                         \b/x
 
-# e.g. 14.09.2012  => DD.MM.YYYY w/ implied hours (set to 12:00)
+# e.g. 14.09.2012  => DD.MM.YYYY
 # regex_de3
 DD_MM_YYYY__DATE_RE = /\b
                   (?<day>\d{1,2})
@@ -71,7 +71,7 @@ DD_MM_YYYY__DATE_RE = /\b
                   (?<year>\d{4})
                     \b/x
 
-# e.g. 14.09.  => DD.MM. w/ implied year and implied hours (set to 12:00)
+# e.g. 14.09.  => DD.MM. w/ implied year
 #  note: allow end delimiter ] e.g. [Sa 12.01.] or end-of-string ($) too
 #  note: we use a lookahead for last part e.g. (?:\s+|$|[\]]) - do NOT cosume
 # regex_de4 (use lookahead assert)
@@ -89,7 +89,7 @@ DD_MM__DATE_RE = /\b
 EN__DD_MONTH_YYYY__DATE_TIME_RE = /\b
               (?<day>\d{1,2})
                  \s
-              (?<month_en>#{MONTH_EN})
+              (?<month_name>#{MONTH_EN})
                  \s
               (?<year>\d{4})
                  \s+
@@ -98,22 +98,16 @@ EN__DD_MONTH_YYYY__DATE_TIME_RE = /\b
               (?<minutes>\d{2})
                 \b/x
 
-###
-# fix: pass in lang (e.g. en or es)
-#  only process format for lang plus fallback to en?
-#   e.g.  EN__DD_MONTH and ES__DD_MONTH depend on order for match (first listed will match)
-
-# e.g. 12 May  => D|DD.MMM  w/ implied year and implied hours
+# e.g. 12 May  => D|DD.MMM  w/ implied year
 EN__DD_MONTH__DATE_RE = /\b
               (?<day>\d{1,2})
                  \s
-              (?<month_en>#{MONTH_EN})
+              (?<month_name>#{MONTH_EN})
                  \b/x
 
-
 # e.g.  Jun/12 2011 14:00
 EN__MONTH_DD_YYYY__DATE_TIME_RE = /\b
-                 (?<month_en>#{MONTH_EN})
+                 (?<month_name>#{MONTH_EN})
                    \/
                  (?<day>\d{1,2})
                    \s
@@ -126,7 +120,7 @@ EN__MONTH_DD_YYYY__DATE_TIME_RE = /\b
 
 # e.g.  Jun/12 14:00  w/ implied year H|HH:MM
 EN__MONTH_DD__DATE_TIME_RE = /\b
-                 (?<month_en>#{MONTH_EN})
+                 (?<month_name>#{MONTH_EN})
                    \/
                  (?<day>\d{1,2})
                    \s+
@@ -135,9 +129,9 @@ EN__MONTH_DD__DATE_TIME_RE = /\b
                  (?<minutes>\d{2})
                    \b/x
 
-# e.g. Jun/12 2013  w/ implied hours (set to 12:00)
+# e.g. Jun/12 2013
 EN__MONTH_DD_YYYY__DATE_RE = /\b
-              (?<month_en>#{MONTH_EN})
+              (?<month_name>#{MONTH_EN})
                  \/
               (?<day>\d{1,2})
                  \s
@@ -150,85 +144,66 @@ EN__MONTH_DD_YYYY__DATE_RE = /\b
 #  fix: remove  space again for now - and use simple en date reader or something!!!
 ##  was [\/ ]   changed back to \/
 EN__MONTH_DD__DATE_RE = /\b
-                 (?<month_en>#{MONTH_EN})
+                 (?<month_name>#{MONTH_EN})
                     \/
                  (?<day>\d{1,2})
                    \b/x
 
 
-# e.g.  12 Ene  w/ implied year and implied hours (set to 12:00)
+# e.g.  12 Ene  w/ implied year
 ES__DD_MONTH__DATE_RE = /\b
                  (?<day>\d{1,2})
                    \s
-                 (?<month_es>#{MONTH_ES})
+                 (?<month_name>#{MONTH_ES})
                    \b/x
 
 # e.g. Ven 8 Août  or [Ven 8 Août] or Ven 8. Août  or [Ven 8. Août]
 ### note: do NOT consume [] in regex (use lookahead assert)
-FR__WEEKDAY_DD_MONTH__DATE_RE = /\b
-     (?:#{WEEKDAY_FR})   # note: skip weekday for now; do NOT capture
+FR__DAY_DD_MONTH__DATE_RE = /\b
+     (?<day_name>#{DAY_FR})
        \s+
      (?<day>\d{1,2})
        \.?        # note: make dot optional
        \s+
-     (?<month_fr>#{MONTH_FR})
+     (?<month_name>#{MONTH_FR})
        (?=\s+|$|[\]])/x  ## note: allow end-of-string/line too
 
 
 
-#
-# map table - 1) tag, 2) regex - note: order matters; first come-first matched/served
-##  todo/fix: remove (move to attic)???  always use lang specific - why? why not?
-FORMATS_ALL = [
-  [ '[YYYY_MM_DD_hh_mm]',        DB__DATE_TIME_RE         ],
-  [ '[YYYY_MM_DD]',              DB__DATE_RE              ],
-  [ '[DD_MM_YYYY_hh_mm]',        DD_MM_YYYY__DATE_TIME_RE ],
-  [ '[DD_MM_hh_mm]',             DD_MM__DATE_TIME_RE ],
-  [ '[DD_MM_YYYY]',              DD_MM_YYYY__DATE_RE ],
-  [ '[DD_MM]',                   DD_MM__DATE_RE ],
-  [ '[FR_WEEKDAY_DD_MONTH]',     FR__WEEKDAY_DD_MONTH__DATE_RE ],
-  [ '[EN_DD_MONTH_YYYY_hh_mm]',  EN__DD_MONTH_YYYY__DATE_TIME_RE ],
-  [ '[EN_MONTH_DD_YYYY_hh_mm]',  EN__MONTH_DD_YYYY__DATE_TIME_RE ],
-  [ '[EN_MONTH_DD_hh_mm]',       EN__MONTH_DD__DATE_TIME_RE ],
-  [ '[EN_MONTH_DD_YYYY]',        EN__MONTH_DD_YYYY__DATE_RE ],
-  [ '[EN_MONTH_DD]',             EN__MONTH_DD__DATE_RE ],
-  [ '[EN_DD_MONTH]',             EN__DD_MONTH__DATE_RE ],
-  [ '[ES_DD_MONTH]',             ES__DD_MONTH__DATE_RE ]
-]
-
+#############################################
+# map tables - 1) regex,  2) tag - note: order matters; first come-first matched/served
 
 FORMATS_BASE = [    ### all numbers (no month names or weekday) - find a better name?
-  [ '[YYYY_MM_DD_hh_mm]',        DB__DATE_TIME_RE         ],
-  [ '[YYYY_MM_DD]',              DB__DATE_RE              ],
-  [ '[DD_MM_YYYY_hh_mm]',        DD_MM_YYYY__DATE_TIME_RE ],
-  [ '[DD_MM_hh_mm]',             DD_MM__DATE_TIME_RE ],
-  [ '[DD_MM_YYYY]',              DD_MM_YYYY__DATE_RE ],
-  [ '[DD_MM]',                   DD_MM__DATE_RE ],
+  [ DB__DATE_TIME_RE,         '[YYYY_MM_DD_hh_mm]' ],
+  [ DB__DATE_RE,              '[YYYY_MM_DD]'       ],
+  [ DD_MM_YYYY__DATE_TIME_RE, '[DD_MM_YYYY_hh_mm]' ],
+  [ DD_MM__DATE_TIME_RE,      '[DD_MM_hh_mm]'      ],
+  [ DD_MM_YYYY__DATE_RE,      '[DD_MM_YYYY]'       ],
+  [ DD_MM__DATE_RE,           '[DD_MM]'            ],
 ]
 
 FORMATS_EN = [
-  [ '[EN_DD_MONTH_YYYY_hh_mm]',  EN__DD_MONTH_YYYY__DATE_TIME_RE ],
-  [ '[EN_MONTH_DD_YYYY_hh_mm]',  EN__MONTH_DD_YYYY__DATE_TIME_RE ],
-  [ '[EN_MONTH_DD_hh_mm]',       EN__MONTH_DD__DATE_TIME_RE ],
-  [ '[EN_MONTH_DD_YYYY]',        EN__MONTH_DD_YYYY__DATE_RE ],
-  [ '[EN_MONTH_DD]',             EN__MONTH_DD__DATE_RE ],
-  [ '[EN_DD_MONTH]',             EN__DD_MONTH__DATE_RE ],
+  [ EN__DD_MONTH_YYYY__DATE_TIME_RE, '[EN_DD_MONTH_YYYY_hh_mm]' ],
+  [ EN__MONTH_DD_YYYY__DATE_TIME_RE, '[EN_MONTH_DD_YYYY_hh_mm]' ],
+  [ EN__MONTH_DD__DATE_TIME_RE,      '[EN_MONTH_DD_hh_mm]'      ],
+  [ EN__MONTH_DD_YYYY__DATE_RE,      '[EN_MONTH_DD_YYYY]'       ],
+  [ EN__MONTH_DD__DATE_RE,           '[EN_MONTH_DD]'            ],
+  [ EN__DD_MONTH__DATE_RE,           '[EN_DD_MONTH]'            ],
 ]
 
 FORMATS_FR = [
-  [ '[FR_WEEKDAY_DD_MONTH]',     FR__WEEKDAY_DD_MONTH__DATE_RE ],
+  [ FR__DAY_DD_MONTH__DATE_RE,       '[FR_DAY_DD_MONTH]' ],
 ]
 
 FORMATS_ES = [
-  [ '[ES_DD_MONTH]',             ES__DD_MONTH__DATE_RE ],
+  [ ES__DD_MONTH__DATE_RE,           '[ES_DD_MONTH]' ],
 ]
 
 
 FORMATS = {
-  'en' => FORMATS_BASE+FORMATS_EN,
-  'fr' => FORMATS_BASE+FORMATS_FR,
-  'es' => FORMATS_BASE+FORMATS_ES,
+  en: FORMATS_BASE+FORMATS_EN,
+  fr: FORMATS_BASE+FORMATS_FR,
+  es: FORMATS_BASE+FORMATS_ES,
 }
 
-
 end  # module DateFormats

data/lib/date-formats/reader.rb

@@ -34,45 +34,32 @@ class Reader   ## todo/check: rename to WordReader or something for easy (re)use
       end
       lines
     end # method parse
+end # class Reader
+
 
-    def self.parse_month( txt )
-        lines = parse( txt )
-        if lines.size != 12
-          puts "*** !!! ERROR !!! reading month names; got #{lines.size} lines - expected 12"
-          exit 1
-        end
-        lines
+
+def self.parse_month( txt )
+    lines = Reader.parse( txt )
+    if lines.size != 12
+      puts "*** !!! ERROR !!! reading month names; got #{lines.size} lines - expected 12"
+      exit 1
     end
+    lines
+end
 
-    def self.parse_weekday( txt )
-        lines = parse( txt )
-        if lines.size != 7
-          puts "*** !!! ERROR !!! reading weekday names; got #{lines.size} lines - expected 7"
-          exit 1
-        end
-        lines
+def self.parse_day( txt )
+    lines = Reader.parse( txt )
+    if lines.size != 7
+      puts "*** !!! ERROR !!! reading day names; got #{lines.size} lines - expected 7"
+      exit 1
     end
-end # class Reader
+    lines
+end
 
 
-def self.build_re( lines )
+def self.build_names( lines )
   ## join all words together into a single string e.g.
   ##   January|Jan|February|Feb|March|Mar|April|Apr|May|June|Jun|...
   lines.map { |line| line.join('|') }.join('|')
 end
-
-def self.build_map( lines )
-  ## build a lookup map that maps the word to the index (line no) plus 1 e.g.
-  ##   note: index is a string too
-  ##  {"January" => "1",  "Jan" => "1",
-  ##   "February" => "2", "Feb" => "2",
-  ##   "March" => "3",    "Mar" => "3",
-  ##   "April" => "4",    "Apr" => "4",
-  ##   "May" => "5",
-  ##   "June" => "6",     "Jun" => "6", ...
-  lines.each_with_index.reduce( {} ) do |h,(line,i)|
-    line.each { |name| h[ name ] = (i+1).to_s }  ## note: start mapping with 1 (and NOT zero-based, that is, 0)
-    h
-  end
-end
 end # module DateFormats

data/lib/date-formats/source.rb

@@ -1,11 +1,16 @@
 module DateFormats
-module Source
 
 # todo: make more generic for reuse
 ### fix:
 ##    use  date/en.txt or en.txt etc. --  why? why not?
 
-  MONTH_EN = <<TXT
+##  note: always sort lines with longest words, abbrevations first!!!!
+##  todo/fix: add/split into MONTH_NAMES and MONTH_ABBREVS (and DAY_NAMES and DAY_ABBREVS) - why? why not?
+  MONTH_NAMES = {}
+  DAY_NAMES   = {}
+
+
+  MONTH_NAMES[:en] = <<TXT
 January    Jan
 February   Feb
 March      Mar
@@ -20,11 +25,12 @@ November   Nov
 December   Dec
 TXT
 
-  WEEKDAY_EN = <<TXT
+
+  DAY_NAMES[:en] = <<TXT
 Monday      Mon
-Tuesday     Tu  Tue  Tues
+Tuesday     Tues  Tue  Tu
 Wednesday   Wed
-Thursday    Th  Thu  Thur  Thurs
+Thursday    Thurs  Thur  Thu  Th
 Friday      Fri
 Saturday    Sat
 Sunday      Sun
@@ -32,7 +38,7 @@ TXT
 
 
 
-  MONTH_FR = <<TXT
+  MONTH_NAMES[:fr] = <<TXT
 Janvier    Janv   Jan     ## check janv in use??
 Février    Févr   Fév     ## check fevr in use???
 Mars              Mar
@@ -47,7 +53,7 @@ Novembre   Nove   Nov     ##  check nove in use??
 Décembre   Déce   Déc     ## check dece in use??
 TXT
 
-  WEEKDAY_FR = <<TXT
+  DAY_NAMES[:fr] = <<TXT
 Lundi     Lun  L
 Mardi     Mar  Ma
 Mercredi  Mer  Me
@@ -59,7 +65,7 @@ TXT
 
 
 
-  MONTH_ES = <<TXT
+  MONTH_NAMES[:es] = <<TXT
 Enero      Ene
 Febrero    Feb
 Marzo      Mar
@@ -74,7 +80,7 @@ Noviembre  Nov
 Diciembre  Dic
 TXT
 
-  MONTH_DE = <<TXT
+  MONTH_NAMES[:de] = <<TXT
 Jänner  Januar    Jan  Jän    # note: in Austria - Jänner; in Deutschland Januar allow both ??
 Feber   Februar   Feb
 März              Mär
@@ -89,7 +95,7 @@ November          Nov
 Dezember          Dez
 TXT
 
-  MONTH_IT = <<TXT
+  MONTH_NAMES[:it] = <<TXT
 Gennaio
 Febbraio
 Marzo
@@ -104,7 +110,7 @@ Novembre
 Dicembre
 TXT
 
-  MONTH_PT = <<TXT
+  MONTH_NAMES[:pt] = <<TXT
 Janeiro
 Fevereiro
 Março
@@ -119,7 +125,7 @@ Novembro
 Dezembro
 TXT
 
-  MONTH_RO = <<TXT
+  MONTH_NAMES[:ro] = <<TXT
 Ianuarie
 Februarie
 Martie
@@ -134,5 +140,9 @@ Noiembrie
 Decembrie
 TXT
 
-end # module Source
+############################################
+## convert (unparsed) text to (parsed) lines with words
+MONTH_NAMES.each {|k,v| MONTH_NAMES[k] = parse_month(v) }
+DAY_NAMES.each   {|k,v| DAY_NAMES[k]   = parse_day(v) }
+
 end # module DateFormats

data/lib/date-formats/version.rb

@@ -3,8 +3,8 @@
 
 module DateFormats
   MAJOR = 0    ## todo: namespace inside version or something - why? why not??
-  MINOR = 1
-  PATCH = 1
+  MINOR = 2
+  PATCH = 0
   VERSION = [MAJOR,MINOR,PATCH].join('.')
 
   def self.version

data/test/test_date.rb

@@ -15,7 +15,7 @@ class TestDate < MiniTest::Test
       [ '21.01.2013 21.30', '2013-01-21 21:30', '[DD_MM_YYYY_hh_mm]' ],
       [ '26.01.2013',       '2013-01-26',       '[DD_MM_YYYY]'       ],
       [ '[26.01.2013]',     '2013-01-26',       '[[DD_MM_YYYY]]'     ],
-      [ '[21.1.]',          '2013-01-21 00:00', '[[DD_MM]]'          ]
+      [ '[21.1.]',          '2013-01-21',       '[[DD_MM]]'          ]
     ]
 
     assert_dates( data, start: Date.new( 2013, 1, 1 ) )
@@ -23,15 +23,15 @@ class TestDate < MiniTest::Test
 
   def test_date_fr
     data = [
-      [ '[Ven 08. Août]', '2014-08-08' ],
-      [ 'Ven 08. Août',   '2014-08-08' ],
-      [ 'Ven 8. Août',    '2014-08-08' ],
-      [ '[Sam 9. Août]',  '2014-08-09' ],
-      [ '[Dim 10. Août]', '2014-08-10' ],
-      [ '[Sam 31. Janv]', '2015-01-31' ],
-      [ '[Sam 7. Févr]',  '2015-02-07' ],
-      [ '[Sam 31. Jan]',  '2015-01-31' ],
-      [ '[Sam 7. Fév]',   '2015-02-07' ],
+      [ '[Ven 08. Août]', '2014-08-08', '[[FR_DAY_DD_MONTH]]' ],
+      [ 'Ven 08. Août',   '2014-08-08', '[FR_DAY_DD_MONTH]' ],
+      [ 'Ven 8. Août',    '2014-08-08', '[FR_DAY_DD_MONTH]' ],
+      [ '[Sam 9. Août]',  '2014-08-09', '[[FR_DAY_DD_MONTH]]' ],
+      [ '[Dim 10. Août]', '2014-08-10', '[[FR_DAY_DD_MONTH]]' ],
+      [ '[Sam 31. Janv]', '2015-01-31', '[[FR_DAY_DD_MONTH]]' ],
+      [ '[Sam 7. Févr]',  '2015-02-07', '[[FR_DAY_DD_MONTH]]' ],
+      [ '[Sam 31. Jan]',  '2015-01-31', '[[FR_DAY_DD_MONTH]]' ],
+      [ '[Sam 7. Fév]',   '2015-02-07', '[[FR_DAY_DD_MONTH]]' ],
     ]
 
     assert_dates( data, start: Date.new( 2014, 8, 1 ), lang: 'fr' )
@@ -39,21 +39,21 @@ class TestDate < MiniTest::Test
 
   def test_date_en
     data = [
-      [ 'Jun/12 2011 14:00', '2011-06-12 14:00' ],
-      [ 'Oct/12 2013 16:00', '2013-10-12 16:00' ],
+      [ 'Jun/12 2011 14:00', '2011-06-12 14:00', '[EN_MONTH_DD_YYYY_hh_mm]' ],
+      [ 'Oct/12 2013 16:00', '2013-10-12 16:00', '[EN_MONTH_DD_YYYY_hh_mm]' ],
 
-      [ 'Jan/26 2011',       '2011-01-26' ],
-      [ 'Jan/26 2011',       '2011-01-26 00:00' ],
+      [ 'Jan/26 2011',       '2011-01-26', '[EN_MONTH_DD_YYYY]' ],
+      [ 'Jan/26 2011',       '2011-01-26', '[EN_MONTH_DD_YYYY]' ],
 
-      [ 'Jan/26',            '2013-01-26' ],
-      [ 'Jan/26',            '2013-01-26 00:00' ],
-      [ '26 January',        '2013-01-26' ],
-      [ '26 January',        '2013-01-26 00:00' ],
+      [ 'Jan/26',            '2013-01-26', '[EN_MONTH_DD]' ],
+      [ 'Jan/26',            '2013-01-26', '[EN_MONTH_DD]' ],
+      [ '26 January',        '2013-01-26', '[EN_DD_MONTH]' ],
+      [ '26 January',        '2013-01-26', '[EN_DD_MONTH]' ],
 
-      [ 'Jun/13',            '2013-06-13' ],
-      [ 'Jun/13',            '2013-06-13 00:00' ],
-      [ '13 June',           '2013-06-13' ],
-      [ '13 June',           '2013-06-13 00:00' ]
+      [ 'Jun/13',            '2013-06-13', '[EN_MONTH_DD]' ],
+      [ 'Jun/13',            '2013-06-13', '[EN_MONTH_DD]' ],
+      [ '13 June',           '2013-06-13', '[EN_DD_MONTH]' ],
+      [ '13 June',           '2013-06-13', '[EN_DD_MONTH]' ]
     ]
 
     assert_dates( data, start: Date.new( 2013, 1, 1 ), lang: 'en' )
@@ -66,7 +66,11 @@ private
     data.each do |rec|
       line         = rec[0]
       str          = rec[1]
-      value = parse_date( line, start: start, lang: lang )
+
+      ## note: test / use parse and find!  -- parse MUST go first
+      values = []
+      values << DateFormats.parse( line, start: start, lang: lang )
+      values << DateFormats.find!( line, start: start, lang: lang )
 
       tagged_line  = rec[2]  ## optinal tagged line
       if tagged_line      ## note: line gets tagged inplace!!! (no new string)
@@ -74,14 +78,17 @@ private
         puts "#{line} == #{tagged_line}"
       end
 
-      if str.index( ':' )
-        assert_datetime( DateTime.strptime( str, '%Y-%m-%d %H:%M' ), value )
-      else
-        assert_date( DateTime.strptime( str, '%Y-%m-%d' ), value )
+      values.each do |value|
+        if str.index( ':' )
+          assert_datetime( DateTime.strptime( str, '%Y-%m-%d %H:%M' ), value )
+        else
+          assert_date( Date.strptime( str, '%Y-%m-%d' ), value )
+        end
       end
     end
   end
 
+
   ## todo: check if assert_datetime or assert_date exist already? what is the best practice to check dates ???
   def assert_date( exp, value )
     assert_equal exp.year,  value.year
@@ -98,12 +105,4 @@ private
     assert_date( exp, value )
     assert_time( exp, value )
   end
-
-
-  def parse_date( line, start:, lang: )
-     # e.g. lets you pass in opts[:start_at] ???
-     finder = DateFormats::DateFinder.new( lang: lang )
-     finder.find!( line, start_at: start )
-  end
-
 end # class TestDate

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: date-formats
 version: !ruby/object:Gem::Version
-  version: 0.1.1
+  version: 0.2.0
 platform: ruby
 authors:
 - Gerald Bauer
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2019-11-08 00:00:00.000000000 Z
+date: 2019-11-09 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: logutils