food_ingredient_parser 1.1.2 → 1.1.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
-SHA1:
-  metadata.gz: 64ad7a10a1480b520602113bbcdfc10ba1daf8b5
-  data.tar.gz: 4068a9edbe1dca908228f38d2795ad63a5cbcf76
+SHA256:
+  metadata.gz: 59825ee90990b2c4f52c9e59fae2e34e5b4558bc63a57fc59946db7f71335351
+  data.tar.gz: b8201945554a11fddbac8eb6676c4cbff8c1f5d3523780ec54125e3172fd05ac
 SHA512:
-  metadata.gz: 73ce876757b08e1d2cf0b5126e8d024b3728260134c3c4f3fe49fee14793da77ecc48d286165dc0c86e8363f2eddf6081355ac26d38f524371b367f4aa3cee23
-  data.tar.gz: befa97dc0fd4605cd2019a2cf7a39aa15d5dfa15acf118f10e9b28e104bbc9bd5925f28286e780acb142eb73a935dba26657f1c6fe9cb766e7f4f88310d5ce55
+  metadata.gz: 3b97c863f9da5b26162883a3627809857fa9277e60e76f2c805312ceea34ad40317bcdf6ca2f1c56b2eafc4a5c637bbe0b5ac3c3c24ef8f874dd008bb5cc3bd7
+  data.tar.gz: bfcc88aea38c3db84670e84dd1166cbcdf8ce60edff297a2ed7324616ea53a4625ab137423f2f6701981a27a51b398ab652b207abe435328e4fa67974bad3f01

data/README.md

@@ -185,9 +185,23 @@ So if you only use the stable interface (`parse`, `to_h` and `to_html`), you can
 to e.g. `~> 1.0`. If you depend on more, lock your version against e.g. `~> 1.0.0` and test when you
 upgrade to `1.1`.
 
+## Languages
+
+While most of the parsing is language-independent, some parts need knowledge about certain words
+(like abbreviations and amount specifiers). The gem was developed with ingredient lists in Dutch (nl),
+plus a bit of English and German. Support for other languages is already good, but lacks in certain
+areas: improvements are welcome (starting with a corpus in [data/](data/)).
+
+Many ingredient lists from the USA are structured a bit differently than those from Europe, they
+parse less well (that that's a matter of tine-tuning).
+
 ## Test data
 
-[`data/ingredient-samples-nl`](data/ingredient-samples-nl) contains about 150k
+[`data/ingredient-samples-qm-nl`](data/ingredient-samples-qm-nl) contains about 150k
 real-world ingredient lists found on the Dutch market. Each line contains one ingredient
 list (newlines are encoded as `\n`, empty lines and those starting with `#` are ignored).
 The strict parser currently parses 80%, while the loose parser returns something for all of them.
+
+## License
+
+This software is distributed under the [MIT license](LICENSE). Data may have a [different license](data/README.md).

data/bin/food_ingredient_parser

@@ -41,8 +41,10 @@ def parse_single(s, parsed=nil, parser:, verbosity: 1, print: nil, escape: false
   if parsed
     puts(parsed.inspect) if verbosity > 1
     pp(parsed.to_h, color: color) if verbosity > 0
+    return true
   else
     puts "(no result: #{parser.parser.failure_reason})" if verbosity > 0
+    return false
   end
 end
 
@@ -63,6 +65,7 @@ def parse_file(path, parser:, verbosity: 1, print: nil, escape: false, color: fa
   pct_parsed   = 100.0 * count_parsed / (count_parsed + count_noresult)
   pct_noresult = 100.0 * count_noresult / (count_parsed + count_noresult)
   puts "parsed #{colorize(color && "1;32", count_parsed)} (#{pct_parsed.round(1)}%), no result #{colorize(color && "1;31", count_noresult)} (#{pct_noresult.round(1)}%)"
+  return count_noresult
 end
 
 verbosity = 1
@@ -108,8 +111,10 @@ if strings.any? || files.any?
     STDERR.puts("Please specify one of the known parsers: #{PARSERS.keys.join(", ")}.")
     exit(1)
   end
-  strings.each {|s| parse_single(s, parser: parser, verbosity: verbosity, print: print, escape: escape, color: color) }
-  files.each   {|f| parse_file(f, parser: parser, verbosity: verbosity, print: print, escape: escape, color: color) }
+  success = true
+  strings.each {|s| success &= parse_single(s, parser: parser, verbosity: verbosity, print: print, escape: escape, color: color) }
+  files.each   {|f| success &= parse_file(f, parser: parser, verbosity: verbosity, print: print, escape: escape, color: color) == 0 }
+  success or exit(1)
 else
   STDERR.puts("Please specify one or more --file or --string arguments (see --help).")
 end

data/lib/food_ingredient_parser/cleaner.rb

@@ -2,13 +2,14 @@ module FoodIngredientParser
   module Cleaner
 
     def self.clean(s)
-      s.gsub!("\u00ad", "")             # strip soft hyphen
-      s.gsub!("\u0092", "'")            # windows-1252 apostrophe - https://stackoverflow.com/a/15564279/2866660
-      s.gsub!("aÄs", "aïs")             # encoding issue for maïs
-      s.gsub!("ï", "ï")                # encoding issue
-      s.gsub!("ë", "ë")                # encoding issue
-      s.gsub!(/\A\s*"(.*)"\s*\z/, '\1') # enclosing double quotation marks
-      s.gsub!(/\A\s*'(.*)'\s*\z/, '\1') # enclosing single quotation marks
+      s.gsub!(/(_x005f_|_)x000d_/i, "\n") # fix sometimes encoding for newline
+      s.gsub!("\u00ad", "")               # strip soft hyphen
+      s.gsub!("\u0092", "'")              # windows-1252 apostrophe - https://stackoverflow.com/a/15564279/2866660
+      s.gsub!("‚", ",")                   # normalize unicode comma
+      s.gsub!("aÄs", "aïs")               # encoding issue for maïs
+      s.gsub!("ï", "ï")                  # encoding issue
+      s.gsub!("ë", "ë")                  # encoding issue
+      s.gsub!(/\A\s*(["']+)(.*)\1\s*\z/, '\2') # enclosing quotation marks
       s
     end
 

data/lib/food_ingredient_parser/loose/scanner.rb

@@ -4,16 +4,31 @@ module FoodIngredientParser::Loose
   class Scanner
 
     SEP_CHARS  = "|;,.".freeze
-    MARK_CHARS = "¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº⁽⁾†‡•°#^*".freeze
+    MARK_CHARS = "¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº⁽⁾†‡⁺•°▪◊#^˄*~".freeze
     PREFIX_RE  = /\A\s*(ingredients|contains|ingred[iï][eë]nt(en)?(declaratie)?|bevat|dit zit er\s?in|samenstelling|zutaten)\b\s*[:;.]?\s*/i.freeze
-    NOTE_RE    = /\A\b(dit product kan\b|deze verpakking kan\b|kan sporen\b.*?\bbevatten\b|voor allergenen\b|allergenen\b|E\s*=|gemaakt in\b|geproduceerd in\b|bevat mogelijk\b|kijk voor meer\b|allergie-info|in de fabriek\b|in dit bedrijf\b|voor [0-9,.]+ (g\.?|gr\.?|ram|ml).*\bis [0-9,.]+ (g\.?|gr\.?|ram|ml).*\bgebruikt\b)/i.freeze
+    NOTE_RE    = /\A\b(dit product kan\b|deze verpakking kan\b|kan sporen\b.*?\bbevatten\b|voor allergenen\b|allergenen\b|allergie[- ]informatie(\s*:|\b)|E\s*=|gemaakt in\b|geproduceerd in\b|bevat mogelijk\b|kijk voor meer\b|allergie-info|in de fabriek\b|in dit bedrijf\b|voor [0-9,.]+ (g\.?|gr\.?|ram|ml).*\bis [0-9,.]+ (g\.?|gr\.?|ram|ml).*\bgebruikt\b)/i.freeze
     # Keep in sync with +abbrev+ in the +Common+ grammar, plus relevant ones from the +Amount+ grammar.
-    ABBREV_RE  = Regexp.union(/\A(N°|°C|(ijzer|chroom|koper)\s*\(I+\)\s*[[:alnum:]]+)\b/i, *%w[
-      a.o.p b.g.a b.o.b c.a c.i d.e d.m.v d.o.c d.o.p d.s e.a e.g e.u f.i.l f.o.s i.a
-      i.d i.e i.g.m.e i.g.p i.m.v i.o i.v.m l.s.l n.a n.b n.o n.v.t o.a o.b.v p.d.o
-      p.g.i q.s s.l s.s t.o.v u.h.t v.g v.s w.a w.o w.v vit denat
-      min max ca
-    ].map {|s| /\A#{Regexp.escape(s)}\b\.?/i}).freeze
+    ABBREV_RE  = Regexp.union(
+      /\A(
+        N°\b |
+        °C\b |
+        (ijzer|chroom|koper)\s*\(I+\)\s*[[:alnum:]]+\b |
+        L\(\+\)[ -][[:alnum:]]+\b |
+        L\.\s+rhamnosus\b | L\.\s+acidophilus\b | L\.\s+casei\b | B\.\s+lactis | A\.\s+oryzae |
+        S\.\s+thermophilus\b | L\.\sbulgaricus\b |
+        T\.\s*aestivum\b(\s+vitt\.)? |
+        nucifera\s+L\. |
+        type\s+"\d+" |
+        E-e?\d{3}[a-z]?\s*\(i+\) |
+        www\.[-_\/:%.A-Za-z0-9]+
+      )/xi,
+      *%w[
+        a.o.p b.g.a b.o.b c.a c.i d.e d.m.v d.o.c d.o.p d.s e.a e.g e.u f.i.l f.o.s i.a
+        i.d i.e i.g.m.e i.g.p i.m.v i.o i.v.m l.s.l n.a n.b n.o n.v.t o.a o.b.v p.d.o
+        p.g.i q.s s.l s.s t.o.v u.h.t v.g v.s w.a w.o w.v vit denat alc vol conc subsp
+        min max ca
+      ].map {|s| /\A#{Regexp.escape(s)}\b\.?/i}
+    ).freeze
 
     def initialize(s, index: 0)
       @s = s                           # input string

data/lib/food_ingredient_parser/loose/transform/amount_from_name.treetop

@@ -5,8 +5,9 @@ module FoodIngredientParser::Loose::Transform
 
     rule amount_from_name
       # just amount, amount in front or at the end
-      ws* amount:amount ws+ name:(.*) /
-      ws* amount:amount ws* /
+      ws* amount:amount                ws+ name:(.*) /
+      ws* amount:amount_simple_percent ws* name:(.*) /
+      ws* amount:amount                ws* /
       ws* name:( !amount word ( ws+ !amount word )* )+ ws* amount:amount ws*
     end
   end

data/lib/food_ingredient_parser/loose/transform/split_e_numbers.rb

@@ -7,8 +7,9 @@ module FoodIngredientParser::Loose
       #
       # @note mark and amount is lost, this is not expected on e-numbers
 
-      SPLIT_RE = /\s*-\s*/.freeze
-      MATCH_RE = /\A\s*(e[0-9]{3}[a-z]?)(?:#{SPLIT_RE}(e[0-9]{3}[a-z]?))+\s*\z/i.freeze
+      SPLIT_RE  = /\s*-\s*/.freeze
+      SINGLE_RE = /E-?\d{3}[a-z]?(?:\s*\(i+\))?/i.freeze
+      MATCH_RE  = /\A\s*(#{SINGLE_RE})(?:#{SPLIT_RE}(#{SINGLE_RE}))+\s*\z/i.freeze
 
       def self.transform!(node)
         new(node).transform!

data/lib/food_ingredient_parser/strict/grammar/amount.treetop

@@ -9,6 +9,10 @@ module FoodIngredientParser::Strict::Grammar
       amount:amount_simple                 <AmountNode>
     end
 
+    rule amount_simple_percent
+      amount:(amount_simple_number ws* percent) <AmountNode>
+    end
+
     rule amount_simple
       ( (
         'of which'i / 'at least'i / 'minimal'i / 'maximal'i / 'less than'i / 'more than'i /
@@ -17,21 +21,30 @@ module FoodIngredientParser::Strict::Grammar
       ) ws* )?
       amount_simple_quantity
       ( ws+ (
-        'minimaal'i / 'minimum'i / 'van het uitlekgewicht'i / 'van het geheel'i /
+        'of a'i / 'of'i / 'or less of'i / 'or more of'i /
+        'van een'i / 'minimaal'i / 'minimum'i / 'van het uitlekgewicht'i / 'van het geheel'i /
         'min.'i / 'min'i / 'max.'i / 'max'i
       ) )?
     end
 
     rule amount_simple_quantity
-      amount_simple_number ( ws* amount_simple_unit? ws* dash ws* amount_simple_number )? ws* amount_simple_unit
+      amount_simple_number ( ws* amount_simple_unit? ws* dash ws* amount_simple_number )? ( ws* amount_simple_unit )?
     end
 
     rule amount_simple_number
-      ( [±∓~∼∽≂≃≈≲≤<>≥≳] ws* )? number
+      ( amount_simple_comparator ws* )? number
+    end
+
+    rule amount_simple_comparator
+      '=' ws* [<>] /
+      [<>] ws* ( '=' / 'of gelijk aan'i !char / 'or equal to'i !char ) /
+      [±∓~∼∽≂≃≈≲≤<>≥≳] / '+/-' / '-/+'
     end
 
     rule amount_simple_unit
-      ( percent / ( ( 'procent' / 'percent' / 'gram'i / 'ml'i / 'mg'i / 'gr'i / 'g'i ) !char ) )
+      ( percent / ( ( 'procent' / 'percent' / 'gram'i / 'ml'i / 'mg'i / 'gr'i / 'g'i / 'ppm'i ) !char ) )
+      ( ws 'vol'i ( !char / '.' ) )?
+      ( ws* '℮' )?
     end
   end
 end

data/lib/food_ingredient_parser/strict/grammar/common.treetop

@@ -10,17 +10,22 @@ module FoodIngredientParser::Strict::Grammar
     end
 
     rule char
-      [[:alnum:]] /
+      !mark [[:alnum:]] /
       fraction /
-      [-/\`'´’+=_{}&] /
-      [®™] /
-      [¿?] /                         # weird characters turning up in names (e.g. encoding issues)
+      [-/\`'"´‘’+=_{}&] /
+      [®©™♣] /
+      [¿?¯] /                        # weird characters turning up in names (e.g. encoding issues)
       [₁₂₃₄₅₆₇₈₉]                    # can occur with vitamins
     end
 
     rule mark
       # mark referencing a footnote
-      [¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº] '⁾'? / '⁽' [¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº] '⁾' / [†‡•°#^] / '*'+ / '(' ws* ( [†‡•°#^] / '*'+ ) ws* ')'
+      [¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº] '⁾'? /
+      '⁽' [¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº] '⁾' /
+      [˄^] digit /
+      [†‡⁺•°▪◊#˄^~˛] /
+      '*'+ /
+      '(' ws* ( [†‡⁺•°▪◊#˄^~˛] / '*'+ ) ws* ')'
     end
 
     rule digit
@@ -28,7 +33,8 @@ module FoodIngredientParser::Strict::Grammar
     end
 
     rule fraction
-      [½⅓⅔¼¾⅕⅖⅗⅘⅙⅚⅐⅛⅜⅝⅞⅑⅒]
+      [½⅓⅔¼¾⅕⅖⅗⅘⅙⅚⅐⅛⅜⅝⅞⅑⅒] /
+      digit+ '/' digit+
     end
 
     rule percent
@@ -48,7 +54,26 @@ module FoodIngredientParser::Strict::Grammar
     end
 
     rule and
-      ( 'and' / 'en' / 'und' / '&' ) !char
+      ( 'and' / 'en' / 'und' ) !char / '&'
+    end
+
+    rule e_number
+      ( 'E'i '-'? [0-9] [0-9] [0-9] [[:alpha:]]? )
+      ![[:alnum:]] / ( ws* '(' 'i'i+ ')' ) # e.g. "E450 (iii)"
+    end
+
+    rule chem_systematic_name
+      ( chem_systematic_name_num dash ) ( chem_systematic_name_word dash chem_systematic_name_num dash ws? )* chem_systematic_name_word /
+      ( chem_systematic_name_word dash chem_systematic_name_num dash ws? )+ chem_systematic_name_word
+    end
+
+    rule chem_systematic_name_word
+      [A-Za-z]+ ( dash [A-Za-z]+ dash [A-Za-z]+ )*
+    end
+
+    rule chem_systematic_name_num
+      digit+ [RH] /
+      digit+ ( ',' digit+ )* '\''?
     end
 
     rule abbrev
@@ -105,8 +130,14 @@ module FoodIngredientParser::Strict::Grammar
         'w.o'i /
         'w.v'i /
         # not auto-generated additions
-        'vit'i /
-        'denat'i
+        'nr.'i /
+        'vit'i /   # vitamin
+        'denat'i / # denaturated
+        'alc'i /   # alcohol
+        'vol'i /   # volume
+        'conc'i /  # concentration
+        'subsp'i / # subspecies
+        'www.'i [-_\/:%.A-Za-z0-9]+
       )
       '.'? ![[:alpha:]]
     end
@@ -116,7 +147,15 @@ module FoodIngredientParser::Strict::Grammar
       (
         'N°'i /
         '°C'i /
-        ( 'ijzer'i / 'chroom'i / 'koper'i ) ws* '(' 'I'i+ ')' ws* [[:alnum:]]+
+        ( 'ijzer'i / 'chroom'i / 'koper'i ) ws* '(' 'I'i+ ')' ws* [[:alnum:]]+ /
+        'L(+)' ('-' / ws) [[:alnum:]]+ /
+        'L.' ws+ 'rhamnosus'i / 'L.' ws+ 'acidophilus'i / 'L.' ws+ 'casei' / 'B.'i ws+ 'lactis'i / 'A.'i ws+ 'oryzae'i /
+        'S.' ws+ 'thermophilus'i / 'L.' ws+ 'bulgaricus'i /
+        'T.' ws* 'aestivum'i (ws+ 'vitt.'i)? /
+        'nucifera' ws+ 'L.'i /
+        'type'i ws+ '"' [0-9]+ '"' /
+        e_number /
+        chem_systematic_name
       ) ![[:alpha:]]
     end
   end

data/lib/food_ingredient_parser/strict/grammar/ingredient_nested.treetop

@@ -21,7 +21,7 @@ module FoodIngredientParser::Strict::Grammar
 
     rule ingredient_nested_contains
       'contains'i /
-      'bevat'i
+      'bevat'i / 'bevat o.a.'i / 'o.a.'i / 'met'i
     end
 
   end

data/lib/food_ingredient_parser/strict/grammar/ingredient_simple.treetop

@@ -9,13 +9,14 @@ module FoodIngredientParser::Strict::Grammar
     end
 
     rule ingredient_simple_with_amount
-      pre:( '{' ws* )? amount:amount ws+ ing:ingredient_simple                   <IngredientNode> /
+      pre:( '{' ws* )? amount:amount                ws+ ing:ingredient_simple    <IngredientNode> /
+      pre:( '{' ws* )? amount:amount_simple_percent ws* ing:ingredient_simple    <IngredientNode> /
       ing:ingredient_simple ws* amount:amount post:( ws* '}' )? (ws? mark:mark)? <IngredientNode> /
       ing:ingredient_simple                                                      <IngredientNode>
     end
 
     rule ingredient_simple_e_number
-      name:( [Ee] [0-9] [0-9] [0-9] [a-zA-Z]? ) ![a-zA-Z0-9] <IngredientNode>
+      name:e_number <IngredientNode>
     end
 
   end

data/lib/food_ingredient_parser/strict/grammar/list_coloned.treetop

@@ -5,8 +5,12 @@ module FoodIngredientParser::Strict::Grammar
     include Ingredient
 
     rule list_coloned
+      contains:( ( ws* list_coloned_ingredient ws* '.,')+ ws* list_coloned_ingredient ) <ListNode> /
+      contains:( ( ws* list_coloned_ingredient ws* '.,')+                             ) <ListNode> /
       contains:( ( ws* list_coloned_ingredient ws* '.' )+ ws* list_coloned_ingredient ) <ListNode> /
       contains:( ( ws* list_coloned_ingredient ws* '.' )+                             ) <ListNode> /
+      contains:( ( ws* list_coloned_ingredient ws* ';,')+ ws* list_coloned_ingredient ) <ListNode> /
+      contains:( ( ws* list_coloned_ingredient ws* ';,')+                             ) <ListNode> /
       contains:( ( ws* list_coloned_ingredient ws* ';' )+ ws* list_coloned_ingredient ) <ListNode> /
       contains:( ( ws* list_coloned_ingredient ws* ';' )+                             ) <ListNode> /
       contains:(   ws* list_coloned_ingredient )                                        <ListNode>

data/lib/food_ingredient_parser/strict/grammar/root.treetop

@@ -40,7 +40,7 @@ module FoodIngredientParser::Strict::Grammar
     end
 
     rule root_mark_sentences_in_list
-      ( ( ws* [,.;] / ws ) ws* root_mark_sentence_in_list )+
+      ( ( ws* [,.;] / ws )+ root_mark_sentence_in_list )+
     end
 
     rule root_mark_sentence_in_list

data/lib/food_ingredient_parser/version.rb

@@ -1,4 +1,4 @@
 module FoodIngredientParser
-  VERSION      = '1.1.2'
-  VERSION_DATE = '2018-09-28'
+  VERSION      = '1.1.7'
+  VERSION_DATE = '2020-12-28'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: food_ingredient_parser
 version: !ruby/object:Gem::Version
-  version: 1.1.2
+  version: 1.1.7
 platform: ruby
 authors:
 - wvengen
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-09-28 00:00:00.000000000 Z
+date: 2020-12-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: treetop
@@ -87,8 +87,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.6.13
+rubygems_version: 3.0.3
 signing_key: 
 specification_version: 4
 summary: Parser for ingredient lists found on food products.