[CSV-283] Remove Whitespace Check Determines Delimiter Twice

• Type: Improvement
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 1.9.0
• External issue URL: https://github.com/apache/commons-csv/pull/167
• Assignee: Unassigned
• Reporter: David Mollitor
• Created: 13/Jul/21 21:05
• Updated: 14/Jul/21 00:19
• Resolved: 14/Jul/21 00:18

Description

Lexer.java

    /**
     * Tests if the given char is a whitespace character.
     *
     * @return true if the given char is a whitespace character.
     * @throws IOException If an I/O error occurs.
     */
    boolean isWhitespace(final int ch) throws IOException {
        return !isDelimiter(ch) && Character.isWhitespace((char) ch);
    }

                    while (true) {
                        c = reader.read();
                        if (isDelimiter(c)) {
                            token.type = TOKEN;
                            return token;
                        }
                        if (isEndOfFile(c)) {
                            ...
                        }
                        if (readEndOfLine(c)) {
                            ...
                        }
                        if (!isWhitespace(c)) {
                            // error invalid char between token and next delimiter
                            throw new IOException("(line " + getCurrentLineNumber() +
                                    ") invalid char between encapsulated token and delimiter");
                        }
                    }

So the first check is for the delimiter, and it returns quick if it finds it. After that point, it's known that this is NOT a delimiter, so no need to re-check it in isWhiteSpace. The delimiter check can be somewhat expensive given it may involve a look-ahead IO read.

Remove the redundancy.