food_ingredient_parser 1.1.2 → 1.1.7
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +5 -5
- data/README.md +15 -1
- data/bin/food_ingredient_parser +7 -2
- data/lib/food_ingredient_parser/cleaner.rb +8 -7
- data/lib/food_ingredient_parser/loose/scanner.rb +23 -8
- data/lib/food_ingredient_parser/loose/transform/amount_from_name.treetop +3 -2
- data/lib/food_ingredient_parser/loose/transform/split_e_numbers.rb +3 -2
- data/lib/food_ingredient_parser/strict/grammar/amount.treetop +17 -4
- data/lib/food_ingredient_parser/strict/grammar/common.treetop +49 -10
- data/lib/food_ingredient_parser/strict/grammar/ingredient_nested.treetop +1 -1
- data/lib/food_ingredient_parser/strict/grammar/ingredient_simple.treetop +3 -2
- data/lib/food_ingredient_parser/strict/grammar/list_coloned.treetop +4 -0
- data/lib/food_ingredient_parser/strict/grammar/root.treetop +1 -1
- data/lib/food_ingredient_parser/version.rb +2 -2
- metadata +3 -4
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
|
-
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
2
|
+
SHA256:
|
3
|
+
metadata.gz: 59825ee90990b2c4f52c9e59fae2e34e5b4558bc63a57fc59946db7f71335351
|
4
|
+
data.tar.gz: b8201945554a11fddbac8eb6676c4cbff8c1f5d3523780ec54125e3172fd05ac
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 3b97c863f9da5b26162883a3627809857fa9277e60e76f2c805312ceea34ad40317bcdf6ca2f1c56b2eafc4a5c637bbe0b5ac3c3c24ef8f874dd008bb5cc3bd7
|
7
|
+
data.tar.gz: bfcc88aea38c3db84670e84dd1166cbcdf8ce60edff297a2ed7324616ea53a4625ab137423f2f6701981a27a51b398ab652b207abe435328e4fa67974bad3f01
|
data/README.md
CHANGED
@@ -185,9 +185,23 @@ So if you only use the stable interface (`parse`, `to_h` and `to_html`), you can
|
|
185
185
|
to e.g. `~> 1.0`. If you depend on more, lock your version against e.g. `~> 1.0.0` and test when you
|
186
186
|
upgrade to `1.1`.
|
187
187
|
|
188
|
+
## Languages
|
189
|
+
|
190
|
+
While most of the parsing is language-independent, some parts need knowledge about certain words
|
191
|
+
(like abbreviations and amount specifiers). The gem was developed with ingredient lists in Dutch (nl),
|
192
|
+
plus a bit of English and German. Support for other languages is already good, but lacks in certain
|
193
|
+
areas: improvements are welcome (starting with a corpus in [data/](data/)).
|
194
|
+
|
195
|
+
Many ingredient lists from the USA are structured a bit differently than those from Europe, they
|
196
|
+
parse less well (that that's a matter of tine-tuning).
|
197
|
+
|
188
198
|
## Test data
|
189
199
|
|
190
|
-
[`data/ingredient-samples-nl`](data/ingredient-samples-nl) contains about 150k
|
200
|
+
[`data/ingredient-samples-qm-nl`](data/ingredient-samples-qm-nl) contains about 150k
|
191
201
|
real-world ingredient lists found on the Dutch market. Each line contains one ingredient
|
192
202
|
list (newlines are encoded as `\n`, empty lines and those starting with `#` are ignored).
|
193
203
|
The strict parser currently parses 80%, while the loose parser returns something for all of them.
|
204
|
+
|
205
|
+
## License
|
206
|
+
|
207
|
+
This software is distributed under the [MIT license](LICENSE). Data may have a [different license](data/README.md).
|
data/bin/food_ingredient_parser
CHANGED
@@ -41,8 +41,10 @@ def parse_single(s, parsed=nil, parser:, verbosity: 1, print: nil, escape: false
|
|
41
41
|
if parsed
|
42
42
|
puts(parsed.inspect) if verbosity > 1
|
43
43
|
pp(parsed.to_h, color: color) if verbosity > 0
|
44
|
+
return true
|
44
45
|
else
|
45
46
|
puts "(no result: #{parser.parser.failure_reason})" if verbosity > 0
|
47
|
+
return false
|
46
48
|
end
|
47
49
|
end
|
48
50
|
|
@@ -63,6 +65,7 @@ def parse_file(path, parser:, verbosity: 1, print: nil, escape: false, color: fa
|
|
63
65
|
pct_parsed = 100.0 * count_parsed / (count_parsed + count_noresult)
|
64
66
|
pct_noresult = 100.0 * count_noresult / (count_parsed + count_noresult)
|
65
67
|
puts "parsed #{colorize(color && "1;32", count_parsed)} (#{pct_parsed.round(1)}%), no result #{colorize(color && "1;31", count_noresult)} (#{pct_noresult.round(1)}%)"
|
68
|
+
return count_noresult
|
66
69
|
end
|
67
70
|
|
68
71
|
verbosity = 1
|
@@ -108,8 +111,10 @@ if strings.any? || files.any?
|
|
108
111
|
STDERR.puts("Please specify one of the known parsers: #{PARSERS.keys.join(", ")}.")
|
109
112
|
exit(1)
|
110
113
|
end
|
111
|
-
|
112
|
-
|
114
|
+
success = true
|
115
|
+
strings.each {|s| success &= parse_single(s, parser: parser, verbosity: verbosity, print: print, escape: escape, color: color) }
|
116
|
+
files.each {|f| success &= parse_file(f, parser: parser, verbosity: verbosity, print: print, escape: escape, color: color) == 0 }
|
117
|
+
success or exit(1)
|
113
118
|
else
|
114
119
|
STDERR.puts("Please specify one or more --file or --string arguments (see --help).")
|
115
120
|
end
|
@@ -2,13 +2,14 @@ module FoodIngredientParser
|
|
2
2
|
module Cleaner
|
3
3
|
|
4
4
|
def self.clean(s)
|
5
|
-
s.gsub!(
|
6
|
-
s.gsub!("\
|
7
|
-
s.gsub!("
|
8
|
-
s.gsub!("
|
9
|
-
s.gsub!("
|
10
|
-
s.gsub!(
|
11
|
-
s.gsub!(
|
5
|
+
s.gsub!(/(_x005f_|_)x000d_/i, "\n") # fix sometimes encoding for newline
|
6
|
+
s.gsub!("\u00ad", "") # strip soft hyphen
|
7
|
+
s.gsub!("\u0092", "'") # windows-1252 apostrophe - https://stackoverflow.com/a/15564279/2866660
|
8
|
+
s.gsub!("‚", ",") # normalize unicode comma
|
9
|
+
s.gsub!("aÄs", "aïs") # encoding issue for maïs
|
10
|
+
s.gsub!("ï", "ï") # encoding issue
|
11
|
+
s.gsub!("ë", "ë") # encoding issue
|
12
|
+
s.gsub!(/\A\s*(["']+)(.*)\1\s*\z/, '\2') # enclosing quotation marks
|
12
13
|
s
|
13
14
|
end
|
14
15
|
|
@@ -4,16 +4,31 @@ module FoodIngredientParser::Loose
|
|
4
4
|
class Scanner
|
5
5
|
|
6
6
|
SEP_CHARS = "|;,.".freeze
|
7
|
-
MARK_CHARS = "
|
7
|
+
MARK_CHARS = "¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº⁽⁾†‡⁺•°▪◊#^˄*~".freeze
|
8
8
|
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
|
9
|
-
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
|
9
|
+
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
|
10
10
|
# Keep in sync with +abbrev+ in the +Common+ grammar, plus relevant ones from the +Amount+ grammar.
|
11
|
-
ABBREV_RE = Regexp.union(
|
12
|
-
|
13
|
-
|
14
|
-
|
15
|
-
|
16
|
-
|
11
|
+
ABBREV_RE = Regexp.union(
|
12
|
+
/\A(
|
13
|
+
N°\b |
|
14
|
+
°C\b |
|
15
|
+
(ijzer|chroom|koper)\s*\(I+\)\s*[[:alnum:]]+\b |
|
16
|
+
L\(\+\)[ -][[:alnum:]]+\b |
|
17
|
+
L\.\s+rhamnosus\b | L\.\s+acidophilus\b | L\.\s+casei\b | B\.\s+lactis | A\.\s+oryzae |
|
18
|
+
S\.\s+thermophilus\b | L\.\sbulgaricus\b |
|
19
|
+
T\.\s*aestivum\b(\s+vitt\.)? |
|
20
|
+
nucifera\s+L\. |
|
21
|
+
type\s+"\d+" |
|
22
|
+
E-e?\d{3}[a-z]?\s*\(i+\) |
|
23
|
+
www\.[-_\/:%.A-Za-z0-9]+
|
24
|
+
)/xi,
|
25
|
+
*%w[
|
26
|
+
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
|
27
|
+
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
|
28
|
+
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
|
29
|
+
min max ca
|
30
|
+
].map {|s| /\A#{Regexp.escape(s)}\b\.?/i}
|
31
|
+
).freeze
|
17
32
|
|
18
33
|
def initialize(s, index: 0)
|
19
34
|
@s = s # input string
|
@@ -5,8 +5,9 @@ module FoodIngredientParser::Loose::Transform
|
|
5
5
|
|
6
6
|
rule amount_from_name
|
7
7
|
# just amount, amount in front or at the end
|
8
|
-
ws* amount:amount
|
9
|
-
ws* amount:
|
8
|
+
ws* amount:amount ws+ name:(.*) /
|
9
|
+
ws* amount:amount_simple_percent ws* name:(.*) /
|
10
|
+
ws* amount:amount ws* /
|
10
11
|
ws* name:( !amount word ( ws+ !amount word )* )+ ws* amount:amount ws*
|
11
12
|
end
|
12
13
|
end
|
@@ -7,8 +7,9 @@ module FoodIngredientParser::Loose
|
|
7
7
|
#
|
8
8
|
# @note mark and amount is lost, this is not expected on e-numbers
|
9
9
|
|
10
|
-
SPLIT_RE
|
11
|
-
|
10
|
+
SPLIT_RE = /\s*-\s*/.freeze
|
11
|
+
SINGLE_RE = /E-?\d{3}[a-z]?(?:\s*\(i+\))?/i.freeze
|
12
|
+
MATCH_RE = /\A\s*(#{SINGLE_RE})(?:#{SPLIT_RE}(#{SINGLE_RE}))+\s*\z/i.freeze
|
12
13
|
|
13
14
|
def self.transform!(node)
|
14
15
|
new(node).transform!
|
@@ -9,6 +9,10 @@ module FoodIngredientParser::Strict::Grammar
|
|
9
9
|
amount:amount_simple <AmountNode>
|
10
10
|
end
|
11
11
|
|
12
|
+
rule amount_simple_percent
|
13
|
+
amount:(amount_simple_number ws* percent) <AmountNode>
|
14
|
+
end
|
15
|
+
|
12
16
|
rule amount_simple
|
13
17
|
( (
|
14
18
|
'of which'i / 'at least'i / 'minimal'i / 'maximal'i / 'less than'i / 'more than'i /
|
@@ -17,21 +21,30 @@ module FoodIngredientParser::Strict::Grammar
|
|
17
21
|
) ws* )?
|
18
22
|
amount_simple_quantity
|
19
23
|
( ws+ (
|
20
|
-
'
|
24
|
+
'of a'i / 'of'i / 'or less of'i / 'or more of'i /
|
25
|
+
'van een'i / 'minimaal'i / 'minimum'i / 'van het uitlekgewicht'i / 'van het geheel'i /
|
21
26
|
'min.'i / 'min'i / 'max.'i / 'max'i
|
22
27
|
) )?
|
23
28
|
end
|
24
29
|
|
25
30
|
rule amount_simple_quantity
|
26
|
-
amount_simple_number ( ws* amount_simple_unit? ws* dash ws* amount_simple_number )? ws* amount_simple_unit
|
31
|
+
amount_simple_number ( ws* amount_simple_unit? ws* dash ws* amount_simple_number )? ( ws* amount_simple_unit )?
|
27
32
|
end
|
28
33
|
|
29
34
|
rule amount_simple_number
|
30
|
-
(
|
35
|
+
( amount_simple_comparator ws* )? number
|
36
|
+
end
|
37
|
+
|
38
|
+
rule amount_simple_comparator
|
39
|
+
'=' ws* [<>] /
|
40
|
+
[<>] ws* ( '=' / 'of gelijk aan'i !char / 'or equal to'i !char ) /
|
41
|
+
[±∓~∼∽≂≃≈≲≤<>≥≳] / '+/-' / '-/+'
|
31
42
|
end
|
32
43
|
|
33
44
|
rule amount_simple_unit
|
34
|
-
( percent / ( ( 'procent' / 'percent' / 'gram'i / 'ml'i / 'mg'i / 'gr'i / 'g'i ) !char ) )
|
45
|
+
( percent / ( ( 'procent' / 'percent' / 'gram'i / 'ml'i / 'mg'i / 'gr'i / 'g'i / 'ppm'i ) !char ) )
|
46
|
+
( ws 'vol'i ( !char / '.' ) )?
|
47
|
+
( ws* '℮' )?
|
35
48
|
end
|
36
49
|
end
|
37
50
|
end
|
@@ -10,17 +10,22 @@ module FoodIngredientParser::Strict::Grammar
|
|
10
10
|
end
|
11
11
|
|
12
12
|
rule char
|
13
|
-
[[:alnum:]] /
|
13
|
+
!mark [[:alnum:]] /
|
14
14
|
fraction /
|
15
|
-
[-/\`'
|
16
|
-
[
|
17
|
-
[
|
15
|
+
[-/\`'"´‘’+=_{}&] /
|
16
|
+
[®©™♣] /
|
17
|
+
[¿?¯] / # weird characters turning up in names (e.g. encoding issues)
|
18
18
|
[₁₂₃₄₅₆₇₈₉] # can occur with vitamins
|
19
19
|
end
|
20
20
|
|
21
21
|
rule mark
|
22
22
|
# mark referencing a footnote
|
23
|
-
[¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº] '⁾'? /
|
23
|
+
[¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº] '⁾'? /
|
24
|
+
'⁽' [¹²³⁴⁵ᵃᵇᶜᵈᵉᶠᵍªº] '⁾' /
|
25
|
+
[˄^] digit /
|
26
|
+
[†‡⁺•°▪◊#˄^~˛] /
|
27
|
+
'*'+ /
|
28
|
+
'(' ws* ( [†‡⁺•°▪◊#˄^~˛] / '*'+ ) ws* ')'
|
24
29
|
end
|
25
30
|
|
26
31
|
rule digit
|
@@ -28,7 +33,8 @@ module FoodIngredientParser::Strict::Grammar
|
|
28
33
|
end
|
29
34
|
|
30
35
|
rule fraction
|
31
|
-
[½⅓⅔¼¾⅕⅖⅗⅘⅙⅚⅐⅛⅜⅝⅞⅑⅒]
|
36
|
+
[½⅓⅔¼¾⅕⅖⅗⅘⅙⅚⅐⅛⅜⅝⅞⅑⅒] /
|
37
|
+
digit+ '/' digit+
|
32
38
|
end
|
33
39
|
|
34
40
|
rule percent
|
@@ -48,7 +54,26 @@ module FoodIngredientParser::Strict::Grammar
|
|
48
54
|
end
|
49
55
|
|
50
56
|
rule and
|
51
|
-
( 'and' / 'en' / 'und' / '&'
|
57
|
+
( 'and' / 'en' / 'und' ) !char / '&'
|
58
|
+
end
|
59
|
+
|
60
|
+
rule e_number
|
61
|
+
( 'E'i '-'? [0-9] [0-9] [0-9] [[:alpha:]]? )
|
62
|
+
![[:alnum:]] / ( ws* '(' 'i'i+ ')' ) # e.g. "E450 (iii)"
|
63
|
+
end
|
64
|
+
|
65
|
+
rule chem_systematic_name
|
66
|
+
( chem_systematic_name_num dash ) ( chem_systematic_name_word dash chem_systematic_name_num dash ws? )* chem_systematic_name_word /
|
67
|
+
( chem_systematic_name_word dash chem_systematic_name_num dash ws? )+ chem_systematic_name_word
|
68
|
+
end
|
69
|
+
|
70
|
+
rule chem_systematic_name_word
|
71
|
+
[A-Za-z]+ ( dash [A-Za-z]+ dash [A-Za-z]+ )*
|
72
|
+
end
|
73
|
+
|
74
|
+
rule chem_systematic_name_num
|
75
|
+
digit+ [RH] /
|
76
|
+
digit+ ( ',' digit+ )* '\''?
|
52
77
|
end
|
53
78
|
|
54
79
|
rule abbrev
|
@@ -105,8 +130,14 @@ module FoodIngredientParser::Strict::Grammar
|
|
105
130
|
'w.o'i /
|
106
131
|
'w.v'i /
|
107
132
|
# not auto-generated additions
|
108
|
-
'
|
109
|
-
'
|
133
|
+
'nr.'i /
|
134
|
+
'vit'i / # vitamin
|
135
|
+
'denat'i / # denaturated
|
136
|
+
'alc'i / # alcohol
|
137
|
+
'vol'i / # volume
|
138
|
+
'conc'i / # concentration
|
139
|
+
'subsp'i / # subspecies
|
140
|
+
'www.'i [-_\/:%.A-Za-z0-9]+
|
110
141
|
)
|
111
142
|
'.'? ![[:alpha:]]
|
112
143
|
end
|
@@ -116,7 +147,15 @@ module FoodIngredientParser::Strict::Grammar
|
|
116
147
|
(
|
117
148
|
'N°'i /
|
118
149
|
'°C'i /
|
119
|
-
( 'ijzer'i / 'chroom'i / 'koper'i ) ws* '(' 'I'i+ ')' ws* [[:alnum:]]+
|
150
|
+
( 'ijzer'i / 'chroom'i / 'koper'i ) ws* '(' 'I'i+ ')' ws* [[:alnum:]]+ /
|
151
|
+
'L(+)' ('-' / ws) [[:alnum:]]+ /
|
152
|
+
'L.' ws+ 'rhamnosus'i / 'L.' ws+ 'acidophilus'i / 'L.' ws+ 'casei' / 'B.'i ws+ 'lactis'i / 'A.'i ws+ 'oryzae'i /
|
153
|
+
'S.' ws+ 'thermophilus'i / 'L.' ws+ 'bulgaricus'i /
|
154
|
+
'T.' ws* 'aestivum'i (ws+ 'vitt.'i)? /
|
155
|
+
'nucifera' ws+ 'L.'i /
|
156
|
+
'type'i ws+ '"' [0-9]+ '"' /
|
157
|
+
e_number /
|
158
|
+
chem_systematic_name
|
120
159
|
) ![[:alpha:]]
|
121
160
|
end
|
122
161
|
end
|
@@ -9,13 +9,14 @@ module FoodIngredientParser::Strict::Grammar
|
|
9
9
|
end
|
10
10
|
|
11
11
|
rule ingredient_simple_with_amount
|
12
|
-
pre:( '{' ws* )? amount:amount
|
12
|
+
pre:( '{' ws* )? amount:amount ws+ ing:ingredient_simple <IngredientNode> /
|
13
|
+
pre:( '{' ws* )? amount:amount_simple_percent ws* ing:ingredient_simple <IngredientNode> /
|
13
14
|
ing:ingredient_simple ws* amount:amount post:( ws* '}' )? (ws? mark:mark)? <IngredientNode> /
|
14
15
|
ing:ingredient_simple <IngredientNode>
|
15
16
|
end
|
16
17
|
|
17
18
|
rule ingredient_simple_e_number
|
18
|
-
name:
|
19
|
+
name:e_number <IngredientNode>
|
19
20
|
end
|
20
21
|
|
21
22
|
end
|
@@ -5,8 +5,12 @@ module FoodIngredientParser::Strict::Grammar
|
|
5
5
|
include Ingredient
|
6
6
|
|
7
7
|
rule list_coloned
|
8
|
+
contains:( ( ws* list_coloned_ingredient ws* '.,')+ ws* list_coloned_ingredient ) <ListNode> /
|
9
|
+
contains:( ( ws* list_coloned_ingredient ws* '.,')+ ) <ListNode> /
|
8
10
|
contains:( ( ws* list_coloned_ingredient ws* '.' )+ ws* list_coloned_ingredient ) <ListNode> /
|
9
11
|
contains:( ( ws* list_coloned_ingredient ws* '.' )+ ) <ListNode> /
|
12
|
+
contains:( ( ws* list_coloned_ingredient ws* ';,')+ ws* list_coloned_ingredient ) <ListNode> /
|
13
|
+
contains:( ( ws* list_coloned_ingredient ws* ';,')+ ) <ListNode> /
|
10
14
|
contains:( ( ws* list_coloned_ingredient ws* ';' )+ ws* list_coloned_ingredient ) <ListNode> /
|
11
15
|
contains:( ( ws* list_coloned_ingredient ws* ';' )+ ) <ListNode> /
|
12
16
|
contains:( ws* list_coloned_ingredient ) <ListNode>
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: food_ingredient_parser
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.1.
|
4
|
+
version: 1.1.7
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- wvengen
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date:
|
11
|
+
date: 2020-12-28 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: treetop
|
@@ -87,8 +87,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
87
87
|
- !ruby/object:Gem::Version
|
88
88
|
version: '0'
|
89
89
|
requirements: []
|
90
|
-
|
91
|
-
rubygems_version: 2.6.13
|
90
|
+
rubygems_version: 3.0.3
|
92
91
|
signing_key:
|
93
92
|
specification_version: 4
|
94
93
|
summary: Parser for ingredient lists found on food products.
|