prestodb/presto · error · IllegalArgumentException

Invalid mapping for column '%s'

Error message

Invalid mapping for column '%s'

What it means

In RawRowEncoder's ColumnMapping constructor, a present mapping string that does not match MAPPING_PATTERN (an optional start-end integer range like '<start>-<end>' or a dataFormat:fieldType spec) causes an IllegalArgumentException 'Invalid mapping for column <name>'.

Source

Thrown at presto-kafka/src/main/java/com/facebook/presto/kafka/encoder/raw/RawRowEncoder.java:142

        private final String name;
        private final Type type;
        private final FieldType fieldType;
        private final int start;
        private final int end;

        public ColumnMapping(EncoderColumnHandle columnHandle)
        {
            this.name = columnHandle.getName();
            this.type = columnHandle.getType();

            this.fieldType = parseFieldType(columnHandle.getDataFormat(), this.name);
            checkFieldType(this.name, this.type, this.fieldType);

            Optional<String> mapping = Optional.ofNullable(columnHandle.getMapping());
            if (mapping.isPresent()) {
                Matcher mappingMatcher = MAPPING_PATTERN.matcher(mapping.get());
                if (!mappingMatcher.matches()) {
                    throw new IllegalArgumentException(format("Invalid mapping for column '%s'", this.name));
                }

                if (mappingMatcher.group(2) != null) {
                    this.start = parseOffset(mappingMatcher.group(1), "start", this.name);
                    this.end = parseOffset(mappingMatcher.group(2), "end", this.name);
                }
                else {
                    this.start = parseOffset(mappingMatcher.group(1), "start", this.name);
                    this.end = this.start + this.fieldType.getSize();
                }
            }
            else {
                throw new IllegalArgumentException(format("No mapping defined for column '%s'", this.name));
            }
        }

        private static int parseOffset(String group, String offsetName, String columnName)
        {

View on GitHub (pinned to 55bb57d202)

Solutions

1. Correct the mapping string to the expected format: either '<start>-<end>' byte offsets or '<dataFormat>:<fieldType>', e.g. '0-8'.
2. Remove the mapping property entirely if the intent was to omit it (which produces the clearer 'No mapping defined' error).
3. Validate the topic definition JSON so the mapping value reaches the encoder intact.

Example fix

// before
"mapping": "bytes:LONG"
// after
"mapping": "0-7"

// mapping must look like '<start>-<end>' or '<dataFormat>:<fieldType>'
java.util.regex.Pattern p = java.util.regex.Pattern.compile("...MAPPING_PATTERN...");
if (mapping != null && !p.matcher(mapping).matches()) {
    throw new IllegalArgumentException("Bad mapping: " + mapping);
}

try {
    createKafkaTable(defn);
} catch (IllegalArgumentException e) {
    if (e.getMessage().startsWith("Invalid mapping")) {
        // surface the offending column and fix its mapping syntax
    }
    throw e;
}

When it happens

Trigger: A column's mapping property is set but its text does not match the accepted pattern, e.g. 'foo', '5-', '-7', 'abc:def:ghi', or non-numeric offsets like 'a-b'.

Common situations: Typo in the mapping syntax in the topic data table definition; using a named/keyword mapping (like 'key' or 'message') unsupported by this encoder; missing quotes causing parts of the mapping to be dropped from the JSON.

Related errors

• Mapping length '%s' is not equal to expected length '%s' for
• No mapping defined for column '%s'
• Unable to parse '%s' offset for column '%s'
• unexpected internal column '%s'
• Column '%s' does not support 'null' value

AI-assisted analysis of prestodb/presto@55bb57d202 (2026-09-04). Data as JSON: /api/errors/822c3e542d803300. Report an issue: GitHub.