UXValidationRule

UXValidationRule is one entry in a UXValidator’s list.

#use <UXKit>            // or #import "UXValidator.xc"

Overview

class UXValidationRule : Object {
    i32      type;        // UXV_REQUIRED / UXV_REGEX / UXV_MINLEN / UXV_MAXLEN / UXV_INTRANGE
    UXRegex* rx;          // UXV_REGEX only
    i32      a;           // length, or range low
    i32      b;           // range high
    u8*      message;     // what to show when this rule fails
}

The require* methods on the validator build these; you rarely construct one directly.

One shape for five rules

a and b mean different things per type. This avoids five separate classes:

type	`a`	`b`	`rx`
`UXV_REQUIRED`	—	—	—
`UXV_REGEX`	—	—	the compiled pattern
`UXV_MINLEN`	minimum length	—	—
`UXV_MAXLEN`	maximum length	—	—
`UXV_INTRANGE`	low	high	—

Both length rules use a as their single operand, so UXV_MINLEN and UXV_MAXLEN differ only in type. A rule table read from a file needs only the type, two integers and a string.

The message travels with the rule

u8* message

The rule knows what to say when it fails, so the error label and the enabled state come from one place. See the validator.

The string is kept, not copied, so it must outlive the rule. A literal is the usual case; a string built at run time needs UXStr.dup.

message is read only when the rule fails. A rule with no message validates correctly but reports nothing useful.

The regex is pre-compiled

UXRegex* rx

requireMatch compiles the pattern when the rule is added, so validating on every keystroke costs a match and not a compile.

Fields

type

i32 type

One of the five constants.

rx

UXRegex* rx        // UXV_REGEX only; null otherwise

a / b

i32 a; i32 b

Operands, per the table above.

message

u8* message

Conforms to

Inherits Object