Lesson 06

Error codes that name failures

Return named error codes when validation can fail for multiple reasons, so callers can branch without guessing.

errors
validation
contracts

Good Code

src/review_validation.c

enum ReviewStatus {
    REVIEW_OK = 0,
    REVIEW_EMPTY_TITLE = -1,
    REVIEW_SCORE_OUT_OF_RANGE = -2,
};

enum ReviewStatus validate_review(const char *title, int score)
{
    // Named result codes tell callers which input failed.
    if (title == 0 || title[0] == '\0') {
        return REVIEW_EMPTY_TITLE;
    }
    if (score < 0 || score > 100) {
        return REVIEW_SCORE_OUT_OF_RANGE;
    }

    return REVIEW_OK;
}

Bad Code

review_validation.c

int validate_review(const char *title, int score)
{
    // A single -1 erases which validation failed.
    if (title == 0 || title[0] == '\0') {
        return -1;
    }
    if (score < 0 || score > 100) {
        return -1;
    }

    return 0;
}

Review Notes

What to review

Good Code

The good version gives each failure a named enum value. Callers can log a useful reason, map it to a response, or assert it in tests.

Bad Code

The bad version returns -1 for every validation failure. Callers must repeat validation logic or show a generic error to the user.

Takeaways

A C error code is useful when it preserves the reason a caller needs for logging, retrying, or showing a message.