Add stripped unit tests for decoder

That is, test cases where there's no newline or other whitespace at
the beginning or end of input. This was implemented by adding a
--strip option to split-testfile to strip the input file after writing
it.

The actual test JSON texts are the same as testdata/invalid and
testdata/valid. The expected output of the invalid cases had to be
adjusted a bit: because there's no newline at the end, some of the
line numbers needed to be changed.

test/testdata/valid-strip

@@ -0,0 +1,68 @@
+==== empty-string ====
+[""]
+==== short-string ====
+["a"]
+==== simple-ascii-string ====
+["abcdefghijklmnopqrstuvwxyz1234567890 "]
+==== utf-8-string ====
+["€þıœəßð some utf-8 ĸʒ×ŋµåäö𝄞"]
+==== string-escapes ====
+["\"\\\/\b\f\n\r\t"]
+==== one-byte-utf-8 ====
+["\u002c one-byte UTF-8"]
+==== two-byte-utf-8 ====
+["\u0123 two-byte UTF-8"]
+==== three-byte-utf-8 ====
+["\u0821 three-byte UTF-8"]
+==== utf-surrogate-four-byte-encoding ====
+["\uD834\uDD1E surrogate, four-byte UTF-8"]
+==== escaped-utf-control-char ====
+["\u0012 escaped control character"]
+==== simple-int-0 ====
+[0]
+==== simple-int-1 ====
+[1]
+==== simple-int-123 ====
+[123]
+==== negative-zero ====
+[-0]
+==== negative-one ====
+[-1]
+==== negative-int ====
+[-123]
+==== simple-real ====
+[123.456789]
+==== real-exponent ====
+[123e45]
+==== real-capital-e ====
+[1E22]
+==== real-positive-exponent ====
+[1e+2]
+==== real-negative-exponent ====
+[1e-2]
+==== real-capital-e-positive-exponent ====
+[1E+2]
+==== real-capital-e-negative-exponent ====
+[1E-2]
+==== real-fraction-exponent ====
+[123.456e78]
+==== true ====
+[true]
+==== false ====
+[false]
+==== null ====
+[null]
+==== empty-array ====
+[]
+==== empty-object-in-array ====
+[{}]
+==== complex-array ====
+[1,2,3,4,
+"a", "b", "c",
+{"foo": "bar", "core": "dump"},
+true, false, true, true, null, false
+]
+==== empty-object ====
+{}
+==== simple-object ====
+{"a":[]}