prestodb/presto · error · PrestoException

INVALID_FUNCTION_ARGUMENT

INVALID_FUNCTION_ARGUMENT

Error message

Invalid JSON value: %s

What it means

is_json_scalar throws this when the JSON parser yields no token at all, i.e. the input is empty or not parseable as any JSON value. It is a validity check: the function only answers true/false for well-formed JSON, and anything unparseable raises INVALID_FUNCTION_ARGUMENT with the (truncated) input echoed back.

Source

Thrown at presto-main-base/src/main/java/com/facebook/presto/operator/scalar/JsonFunctions.java:107

        return JsonPath.build(padSpaces(pattern, charLength.intValue()).toStringUtf8());
    }

    @ScalarFunction("is_json_scalar")
    @LiteralParameters("x")
    @SqlType(StandardTypes.BOOLEAN)
    public static boolean varcharIsJsonScalar(@SqlType("varchar(x)") Slice json)
    {
        return isJsonScalar(json);
    }

    @ScalarFunction
    @SqlType(StandardTypes.BOOLEAN)
    public static boolean isJsonScalar(@SqlType(StandardTypes.JSON) Slice json)
    {
        try (JsonParser parser = createJsonParser(JSON_FACTORY, json)) {
            JsonToken nextToken = parser.nextToken();
            if (nextToken == null) {
                throw new PrestoException(INVALID_FUNCTION_ARGUMENT, "Invalid JSON value: " + truncateIfNecessaryForErrorMessage(json));
            }

            if (nextToken == START_ARRAY || nextToken == START_OBJECT) {
                parser.skipChildren();
                if (parser.nextToken() != null) {
                    // extra trailing token after json array/object
                    throw new PrestoException(INVALID_FUNCTION_ARGUMENT, "Invalid JSON value: " + truncateIfNecessaryForErrorMessage(json));
                }
                return false;
            }

            if (parser.nextToken() != null) {
                // extra trailing token after json scalar
                throw new PrestoException(INVALID_FUNCTION_ARGUMENT, "Invalid JSON value: " + truncateIfNecessaryForErrorMessage(json));
            }
            return true;
        }
        catch (IOException e) {

View on GitHub (pinned to 55bb57d202)

Solutions

1. Filter out empty/blank strings before calling is_json_scalar.
2. Fix the producing pipeline to emit valid JSON or NULL.
3. Wrap the check with try() in Presto to return NULL for invalid input: try(is_json_scalar(x)).

Example fix

-- before
SELECT is_json_scalar(status) FROM t; -- status can be ''
-- after
SELECT if(length(trim(status)) > 0, is_json_scalar(status), false) FROM t;

SELECT is_json_scalar(x) FROM t WHERE length(trim(x)) > 0;

try { is_json_scalar(x) } catch (PrestoException e) { return false; } // or SQL: try(is_json_scalar(x))

When it happens

Trigger: varchar passed to is_json_scalar that is empty, whitespace-only, or syntactically invalid JSON so nextToken() returns null or the IOException catch path fires.

Common situations: Empty-string columns fed from upstream pipelines; columns containing quoted-but-invalid fragments; missing data encoded as '' instead of NULL.

Understand the failure class

• Parsing and encoding errors: unexpected token, malformed input — why parsers reject input and how to find the real culprit.

Related errors

• INVALID_FUNCTION_ARGUMENT
• Timeout is negative
• LANCE_ERROR
• INVALID_CAST_ARGUMENT
• INVALID_FUNCTION_ARGUMENT

AI-assisted analysis of prestodb/presto@55bb57d202 (2026-09-04). Data as JSON: /api/errors/686389d1feec40fb. Report an issue: GitHub.