Disable HTML escaping for JSON-LD, rewrite htmlEscape

lszeremeta · Apr 9, 2021 · 80573f1 · 80573f1
1 parent 12de21d
commit 80573f1
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 12 deletions.
diff --git a/src/main/java/pl/edu/uwb/ii/sdfeater/Molecule.java b/src/main/java/pl/edu/uwb/ii/sdfeater/Molecule.java
@@ -24,7 +24,7 @@
  */
 package pl.edu.uwb.ii.sdfeater;
 
-import com.google.gson.Gson;
+import com.google.gson.GsonBuilder;
 import org.apache.jena.rdf.model.Property;
 import org.apache.jena.rdf.model.Resource;
 import org.apache.jena.rdf.model.ResourceFactory;
@@ -406,7 +406,7 @@ StringBuilder constructJSONLDMolecule(SDFEater.Subject subject) {
 
         output_str.append("    {\n");
         if (subject == SDFEater.Subject.iri) {
-            output_str.append("      \"@id\" : \"" + subjectBase + createID() + "\",\n");
+            output_str.append("      \"@id\" : " + printValueAsNumberOrStringInJSONLD(subjectBase + createID()) + ",\n");
         } else if (subject == SDFEater.Subject.uuid) {
             output_str.append("      \"@id\" : \"urn:uuid:" + uuid + "\",\n");
         } else if (subject == SDFEater.Subject.bnode) {
@@ -504,7 +504,7 @@ void printRDFaMolecule(SDFEater.Subject subject) {
         if (output_str.length() > 0) {
             if (subject == SDFEater.Subject.iri) {
                 String mID = createID();
-                System.out.print("    <div typeof='schema:MolecularEntity' about='" + htmlEscape(subjectBase, true) + mID + "'");
+                System.out.print("    <div typeof='schema:MolecularEntity' about='" + htmlEscape(subjectBase + mID, true) + "'");
 
                 if (subjectBase.contains("#")) {
                     System.out.print(" id='" + subjectBase.substring(subjectBase.lastIndexOf('#') + 1) + mID + "'");
@@ -571,7 +571,7 @@ void printMicrodataMolecule(SDFEater.Subject subject) {
         if (output_str.length() > 0) {
             if (subject == SDFEater.Subject.iri) {
                 String mID = createID();
-                System.out.print("    <div itemscope itemtype='http://schema.org/MolecularEntity' itemid='" + htmlEscape(subjectBase, true) + mID + "'");
+                System.out.print("    <div itemscope itemtype='http://schema.org/MolecularEntity' itemid='" + htmlEscape(subjectBase + mID, true) + "'");
 
                 if (subjectBase.contains("#")) {
                     System.out.print(" id='" + subjectBase.substring(subjectBase.lastIndexOf('#') + 1) + mID + "'");
@@ -630,7 +630,7 @@ private String printValueAsNumberOrStringInJSONLD(String value) {
         if (isNumber(value)) {
             return value;
         } else {
-            return new Gson().toJson(value);
+            return new GsonBuilder().disableHtmlEscaping().create().toJson(value);
         }
     }
 
@@ -644,15 +644,14 @@ private String printValueAsNumberOrStringInJSONLD(String value) {
      * @return Escaped HTML value
      */
     private String htmlEscape(String value, boolean quote) {
-        value = value.replace("&", "&amp;");
-        value = value.replace("<", "&lt;");
-        value = value.replace(">", "&gt;");
+        String escapedValue = value.replace("&", "&amp;").replace("<", "&lt;").replace(">", "&gt;");
+
         if (quote) {
-            value = value.replace("\"", "&quot;");
-            value = value.replace("'", "&#x27;");
+            escapedValue = escapedValue.replace("\"", "&quot;");
+            escapedValue = escapedValue.replace("'", "&#x27;");
         }
 
-        return value;
+        return escapedValue;
     }
 
     /**

diff --git a/src/test/java/pl/edu/uwb/ii/sdfeater/OutputTest.java b/src/test/java/pl/edu/uwb/ii/sdfeater/OutputTest.java
@@ -20,7 +20,7 @@
 class OutputTest {
     private final ByteArrayOutputStream outputStreamCaptor = new ByteArrayOutputStream();
     private final String[] testMoleculeDataFields = {"alternateName", "description", "identifier", "inChI", "inChIKey", "iupacName", "molecularFormula", "molecularWeight", "smiles"};
-    private final String[] testMoleculeData = {"(+)-2-fenchanone", "A fenchone that has", "4695-62-9", "InChI", "1S/C10H16O/c1-9(2)7-4-5-10(3,6-7)8(9)11/h7H,4-6H2,1-3H3/t7-,10+/m1/s1", "LHXDLQBQYFFVNW-XCBNKYQSSA-N", "(1S,4R)-fenchan-2-one", "C10H16O", "152.23340", "CC1(C)[C@@H]2CC[C@@](C)(C2)C1", "(-)-Epicatechin", "A catechin with", "490-46-0", "1S/C15H14O6/c16-8-4-11(18)9-6-13(20)15(21-14(9)5-8)7-1-2-10(17)12(19)3-7/h1-5,13,15-20H,6H2/t13-,15-/m1/s1", "PFTAWBLQPZVEMU-UKRRQHHQSA-N", "(2R,3R)-2-(3,4-dihydroxyphenyl)-3,4-dihydro-2H-chromene-3,5,7-triol", "C15H14O6", "290.26810", "[H][C@@]1(Oc2cc(O)cc(O)c2C[C@H]1O)c1ccc(O)c(O)c1"};
+    private final String[] testMoleculeData = {"(+)-2-fenchanone", "A fenchone that has", "4695-62-9", "InChI=1S/C10H16O/c1-9(2)7-4-5-10(3,6-7)8(9)11/h7H,4-6H2,1-3H3/t7-,10+/m1/s1", "LHXDLQBQYFFVNW-XCBNKYQSSA-N", "(1S,4R)-fenchan-2-one", "C10H16O", "152.23340", "CC1(C)[C@@H]2CC[C@@](C)(C2)C1=O", "(-)-Epicatechin", "A catechin with", "490-46-0", "InChI=1S/C15H14O6/c16-8-4-11(18)9-6-13(20)15(21-14(9)5-8)7-1-2-10(17)12(19)3-7/h1-5,13,15-20H,6H2/t13-,15-/m1/s1", "PFTAWBLQPZVEMU-UKRRQHHQSA-N", "(2R,3R)-2-(3,4-dihydroxyphenyl)-3,4-dihydro-2H-chromene-3,5,7-triol", "C15H14O6", "290.26810", "[H][C@@]1(Oc2cc(O)cc(O)c2C[C@H]1O)c1ccc(O)c(O)c1"};
     private final File file = new File(Paths.get("src", "test", "resources", "chebi_test.sdf").toFile().getAbsolutePath());
     private Molecule molecule;