feat: Improvements to GS1 imports to prepare integration of Agena3000 (…

…#6566) * update label logos #5239 * equadis message with nested products * add more GS1 tests files * try to use Perl XML::XML2JSON instead of nodejs xml2json * more comments and small fix * 1 GS1 message can contain several products #6537 * 1 GS1 message can contain several products #6537 * update tests * added carrying bag (GS1 PUG) * added carrying bag (GS1 PUG) * update tests * more comments and doc * rename script * more comments and doc * move/rename equadis-xml2json.js script * fix lint issues * predeclare recursive functions * import -> require * Update lib/ProductOpener/GS1.pm Co-authored-by: Alex Garel <[email protected]> * Update lib/ProductOpener/GS1.pm Co-authored-by: Alex Garel <[email protected]> * Update scripts/convert_gs1_xml_to_json_in_dir.pl Co-authored-by: Alex Garel <[email protected]> * add missing ; * fix lint issue * fix lint issue Co-authored-by: off <[email protected]> Co-authored-by: Alex Garel <[email protected]>
openfoodfacts · Apr 8, 2022 · ce4eb51 · ce4eb51
1 parent d875e06
commit ce4eb51
Show file tree

Hide file tree

Showing 39 changed files with 9,228 additions and 1,104 deletions.
diff --git a/cpanfile b/cpanfile
@@ -62,6 +62,8 @@ requires 'Devel::Size'; # deps: libdevel-size-perl
 requires 'JSON::Create';
 requires 'JSON::Parse';
 requires 'Data::DeepAccess';
+requires 'XML::XML2JSON';
+
 
 # Mojolicious/Minion
 requires 'Mojolicious::Lite';

diff --git a/lib/ProductOpener/GS1.pm b/lib/ProductOpener/GS1.pm
@@ -71,6 +71,7 @@ use ProductOpener::Tags qw/:all/;
 
 use JSON::PP;
 use boolean;
+use Data::DeepAccess qw(deep_get);
 
 
 =head1 GS1 MAPS
@@ -248,6 +249,7 @@ my %unknown_entries_in_gs1_maps = ();
 		"EN" => "Enveloppe",
 		"JR" => "Bocal",
 		"PO" => "Poche",
+		"PUG" => "Sac de transport",
 		"TU" => "Tube",
 		"WRP" => "Film",
 	},		
@@ -267,6 +269,7 @@ my %unknown_entries_in_gs1_maps = ();
 		"EN" => "en:envelope",
 		"JR" => "en:jar",
 		"PO" => "en:bag",
+		"PUG" => "en:carrying-bag",
 		"TU" => "en:tube",
 		"WRP" => "en:film",
 	},	
@@ -862,6 +865,13 @@ sub gs1_to_off ($$$) {
 			if ($source_field eq "nutrientHeader") {
 
 				$log->debug("gs1_to_off - special handling for nutrientHeader array") if $log->is_debug();
+
+				# If there is only one nutrition facts table, nutrientHeader might not be an array
+				# depending on how the XML was converted to JSON
+				# In that case, create an array
+				if (ref($json_ref->{$source_field}) eq 'HASH') {
+					$json_ref->{$source_field} = [$json_ref->{$source_field}];
+				}
 
 				# Some products like ice cream may have nutrients per 100g + nutrients per 100ml
 				# in that case, the last values (e.g. for 100g) will override previous values (e.g. for 100ml)
@@ -1200,6 +1210,23 @@ sub gs1_to_off ($$$) {
 
 				# The source structure may be a hash or an array of hashes
 				# e.g. Equadis: allergenRelatedInformation is a hash, CodeOnline: it is an array
+
+				# CodeOnline:
+
+				# allergenRelatedInformation: [
+				# 	{
+				# 		allergen: [
+				# 			{
+				# 				allergenTypeCode: "AC",
+				# 				levelOfContainmentCode: "FREE_FROM"
+				# 			},
+				# 			{
+				# 				allergenTypeCode: "AE",
+				# 				levelOfContainmentCode: "CONTAINS"
+				# 			},				
+
+				$log->debug("gs1_to_off - source_target is a hash",
+					{ source_field => $source_field,  source_target => $source_target, json_ref => $json_ref }) if $log->is_debug();
 
 				if (ref($json_ref->{$source_field}) eq "HASH") {
 
@@ -1208,7 +1235,19 @@ sub gs1_to_off ($$$) {
 				elsif (ref($json_ref->{$source_field}) eq "ARRAY") {
 					foreach my $json_array_entry_ref (@{$json_ref->{$source_field}}) {
 
-						gs1_to_off($source_target, $json_array_entry_ref, $results_ref);
+						# We should have an array of hashes, but in some CodeOnline files we have an array with an empty array..
+
+						# allergenRelatedInformation: [
+						# 	[ ]
+						# ]
+
+						if (ref($json_array_entry_ref) eq "HASH") {
+							gs1_to_off($source_target, $json_array_entry_ref, $results_ref);
+						}
+						else {
+							$log->debug("gs1_to_off - expected a hash but got an array",
+								{ source_field => $source_field,  source_target => $source_target, json_ref => $json_ref, json_array_entry_ref => $json_array_entry_ref }) if $log->is_debug();
+						}
 					}
 				}
 			}
@@ -1217,31 +1256,110 @@ sub gs1_to_off ($$$) {
 }
 
 
-=head2 convert_gs1_json_to_off_csv_fields ($json)
+=head2 convert_single_text_property_to_direct_value ($json )
+
+There are different ways to convert a XML document to a JSON data structure.
+
+Historically, we used nodejs xml2json module to convert the GS1 XML to JSON.
+
+Then we added support for CodeOnline JSON exports which used slightly different conversions.
+
+In order to remove the dependency on nodejs, we are now supporting Perl's XML:XML2JSON module that results in different structures.
+
+This function is a recursive function to make the output of Perl XML::XML2JSON similar to nodejs xml2json, as the GS1 module expects this format.
+
+Difference:
+
+XML2JSON creates a hash for simple text values. Text values of tags are converted to $t properties.
+e.g. <gtin>03449862093657</gtin>
 
-Thus function converts the data for one product in the GS1 format converted to JSON.
+becomes:
+
+gtin: {
+	 $t: "03449865355608"
+},
+
+This function converts those hashes with one single $t scalar values to a direct value.
+
+gtin: "03449865355608"
+
+=head3 Arguments
+
+=head4 $json_ref Reference to a decoded JSON structure
+
+=cut
+
+# pre-declare the function as it is recursive
+sub convert_single_text_property_to_direct_value($);
+
+sub convert_single_text_property_to_direct_value($) {
+
+	my $json_ref = shift;
+
+    my $type = ref $json_ref or return;
+
+    if ($type eq 'HASH') {
+		foreach my $key (keys %$json_ref) {
+			if (ref $json_ref->{$key}) {
+				# Hash with a single $t value?
+				if ((ref $json_ref->{$key} eq 'HASH') and ((scalar keys %{$json_ref->{$key}}) == 1) and (defined $json_ref->{$key}{'$t'})) {
+					$json_ref->{$key} = $json_ref->{$key}{'$t'};
+				}
+				else {
+					convert_single_text_property_to_direct_value($json_ref->{$key});
+				}
+			}
+		}
+    }
+    elsif ($type eq 'ARRAY') {
+
+        foreach my $elem (@$json_ref) {
+            if (ref $elem) {
+                convert_single_text_property_to_direct_value($elem);
+            }
+        }
+    }
+}
+
+
+=head2 convert_gs1_json_message_to_off_products_csv_fields ($json, $products_ref)
+
+Thus function converts the data for one or more products in the GS1 format converted to JSON.
 GS1 format is in XML, it needs to be transformed to JSON with xml2json first.
 In some cases, the conversion to JSON has already be done by a third party (e.g. the CodeOnline database from GS1 France).
 
-=head3 Arguments
+Note: This function is recursive if there are child products.
 
-=head4 json text
+One GS1 message can include 1 or more products, typically products that contain other products
+(e.g. a pallet of cartons of products).
+
+=head3 Arguments
 
-=head3 Return value
+=head4 $json_ref Reference to a decoded JSON structure
 
-=head4 Reference to a hash of fields
+=head4 $product_ref - Reference to an array of product data
 
-The function returns a reference to a hash.
+Each product data will be added as one element (a hash ref) of the product data array.
 
-Each key is the name of the OFF csv field, and it is associated with the corresponding value for the product.
+For each product, the key of the hash is the name of the OFF csv field, and it is associated with the corresponding value for the product.
 
 =cut
 
-sub convert_gs1_json_to_off_csv($) {
+# pre-declare the function as it is recursive
+sub convert_gs1_json_message_to_off_products_csv($$);
 
-	my $json = shift;
+sub convert_gs1_json_message_to_off_products_csv($$) {
+
+	my $json_ref = shift;
+	my $products_ref = shift;
 
-	my $json_ref = decode_json($json);
+	# Depending on how the original XML was converted to JSON,
+	# text values of XML tags can be assigned directly as the value of the corresponding key
+	# or they can be stored inside a hash with the $t key
+	# e.g.
+	# levelOfContainmentCode: {
+	#	$t: "MAY_CONTAIN"
+	# },
 
 	# The JSON can contain only the product information "tradeItem" level
 	# or the tradeItem can be encapsulated in a message
@@ -1259,15 +1377,25 @@ sub convert_gs1_json_to_off_csv($) {
 		documentCommand
 		catalogue_item_notification:catalogueItemNotification
 		catalogueItem
-		tradeItem)) {
+		)) {
 		if (defined $json_ref->{$field}) {
 			$json_ref = $json_ref->{$field};
 			$log->debug("convert_gs1_json_to_off_csv - remove encapsulating field", { field => $field }) if $log->is_debug();
 		}
 	}
+
+	# A product can contain a child product
+	my $child_product_json_ref = deep_get($json_ref, qw(catalogueItemChildItemLink catalogueItem));
+	if (defined $child_product_json_ref) {
+		$log->debug("convert_gs1_json_to_off_csv - found a child item", {  }) if $log->is_debug();
+		convert_gs1_json_message_to_off_products_csv($child_product_json_ref, $products_ref)
+	}
+
+	if (defined $json_ref->{tradeItem}) {
+		$json_ref = $json_ref->{tradeItem};
+	}
 
 	if (not defined $json_ref->{gtin}) {
-
 		$log->debug("convert_gs1_json_to_off_csv - no gtin - skipping", { json_ref => $json_ref }) if $log->is_debug();
 		return {};
 	}
@@ -1291,14 +1419,14 @@ sub convert_gs1_json_to_off_csv($) {
 		delete $results_ref->{languages};
 	}
 
-	return $results_ref;
+	push @$products_ref, $results_ref;
 }
 
 
 =head2 read_gs1_json_file ($json_file, $products_ref)
 
-Read a GS1 file on json format, convert it to the OFF format, return the
-result, and store the result in the $products_ref array (if not undef)
+Read a GS1 message file in json format, convert the included products in the OFF format,
+and store the resulting products in the $products_ref array
 
 =head3 Arguments
 
@@ -1318,14 +1446,15 @@ sub read_gs1_json_file($$) {
 	open (my $in, "<", $json_file) or die("Cannot open json file $json_file : $!\n");
 	my $json = join (q{}, (<$in>));
 	close($in);
+
+	my $json_ref = decode_json($json);
+
+	# Convert JSON structures created from the XML::XML2JSON module
+	# to the format generated by the nodejs xml2json module
+	# which is the expected format of the ProductOpener::GS1 module
+	convert_single_text_property_to_direct_value($json_ref);
 
-	my $results_ref = convert_gs1_json_to_off_csv($json);
-
-	if ((defined $products_ref) and (defined $results_ref->{code})) {
-		push @$products_ref, $results_ref;
-	}
-
-	return $results_ref;
+	convert_gs1_json_message_to_off_products_csv($json_ref, $products_ref);
 }
 
 

diff --git a/scripts/convert_gs1_json_to_off_csv.pl b/scripts/convert_gs1_json_to_off_csv.pl
@@ -72,7 +72,7 @@
 
 	next if $file !~ /\.json$/;
 
-	my $product_ref = read_gs1_json_file("$input_dir/$file", $products_ref);
+	read_gs1_json_file("$input_dir/$file", $products_ref);
 }
 
 write_off_csv_file($output, $products_ref);

diff --git a/scripts/equadis-import/equadis-xml2json.js → scripts/convert_gs1_xml_to_json_in_dir.js b/scripts/equadis-import/equadis-xml2json.js → scripts/convert_gs1_xml_to_json_in_dir.js
@@ -1,12 +1,14 @@
 // This script is used to convert GDSN data from Equadis in XML format
 // to a corresponding JSON structure
 
-const xml2json = require('xml2json')
-const fs = require("fs")
+const process = require('process');
+const xml2json = require('xml2json');
+const fs = require("fs");
 
-const directoryPath = "/srv2/off-pro/equadis-data-tmp/"
+const myArgs = process.argv.slice(2);
+const directoryPath = myArgs[0];
 
-const filter = /\.xml$/
+const filter = /\.xml$/;
 
 // force arrays for some fields even if there is only one value supplied
 const options = {
@@ -15,18 +17,18 @@ const options = {
 
 fs.readdir(directoryPath, function(err, files) {
   if (err) {
-    console.log("Error getting directory information.")
+    console.log("Error getting directory information.");
   } else {
     files.forEach(function(file) {
 
      if (filter.test(file)) {
 
-       let content = fs.readFileSync(directoryPath+file, 'utf8');
-       let json = xml2json.toJson(content, options);
+       /* eslint-disable no-sync */
+       const content = fs.readFileSync(directoryPath+file, 'utf8');
+       const json = xml2json.toJson(content, options);
        fs.writeFileSync(directoryPath+file.replace('.xml','.json'), json);
      }
 
-    })
+    });
   }
-})
-
+});