feat(web): simplify contextual parsing

* More relaxted parsing in the browser for stable variable/node
  identifier classification.
* String support in the browser.
* contains, refers, and opposite keywords are no longer contextual,
  which simplifies parsing in the browser.

1 files changed, 2 insertions, 1 deletions

diff --git a/language-web/src/main/js/editor/problemLanguageSupport.ts b/language-web/src/main/js/editor/problemLanguageSupport.ts
index 2bf7c7a4..c9e61b31 100644
--- a/language-web/src/main/js/editor/problemLanguageSupport.ts
+++ b/language-web/src/main/js/editor/problemLanguageSupport.ts
@@ -26,7 +26,7 @@ const parserWithMetadata = (parser as LRParser).configure({
       LineComment: t.lineComment,
       BlockComment: t.blockComment,
       'problem class enum pred unique scope': t.definitionKeyword,
-      'abstract refers contains opposite error default': t.modifier,
+      'abstract extends refers contains opposite error default': t.modifier,
       'true false unknown error': t.keyword,
       NotOp: t.keyword,
       UnknownOp: t.keyword,
@@ -34,6 +34,7 @@ const parserWithMetadata = (parser as LRParser).configure({
       StarArgument: t.keyword,
       'IntMult StarMult Real': t.number,
       StarMult: t.number,
+      String: t.string,
       'RelationName/QualifiedName': t.typeName,
       'UniqueNodeName/QualifiedName': t.atom,
       'VariableName/QualifiedName': t.variableName,