Let Lexer.prototype.getNumber treat more cases of a single minus sign as zero (bug 1953099)

This patch extends the approach of PR 14543, by also treating e.g. minus signs followed by '(' or '<' as zero.
Inside of a /Contents stream those characters will generally mean the start of one or more glyphs.

test/pdfs/.gitignore

@@ -16,6 +16,7 @@
 !bug1727053.pdf
 !issue18408_reduced.pdf
 !bug1907000_reduced.pdf
+!bug1953099.pdf
 !issue11913.pdf
 !issue2391-1.pdf
 !issue2391-2.pdf

test/test_manifest.json

@@ -10414,6 +10414,14 @@
     "link": true,
     "type": "eq"
   },
+  {
+    "id": "bug1953099",
+    "file": "pdfs/bug1953099.pdf",
+    "md5": "15295cfa4999ccc08442423fca96c28f",
+    "rounds": 1,
+    "link": false,
+    "type": "eq"
+  },
   {
     "id": "bug1899804_print",
     "file": "pdfs/bug1899804.pdf",

test/unit/parser_spec.js

@@ -152,7 +152,17 @@ describe("parser", function () {
       });
 
       it("should treat a single decimal point, or minus/plus sign, as zero", function () {
-        const validNums = [".", "-", "+", "-.", "+.", "-\r\n.", "+\r\n."];
+        const validNums = [
+          ".",
+          "-",
+          "+",
+          "-.",
+          "+.",
+          "-\r\n.",
+          "+\r\n.",
+          "-(",
+          "-<",
+        ];
         for (const number of validNums) {
           const validInput = new StringStream(number);
           const validLexer = new Lexer(validInput);