Cleanup after UI migration, updated docs, fixed retransmission bug

[szczygiel-m]

No description provided.

[pitagoras3]

question: If only one timestamp (either start or end) is null, then both properties should have null value?

[szczygiel-m]

Actually it doesn't matter, because the validation will evoke 400 from an endpoint with message that timestamp cannot be null.

[pitagoras3]

suggestion: Catching Exception can be too wide in this case - DateTimeParseException could be a better fit.

[pitagoras3]

suggestion: Remove code repetition by extracting DateTimeFormatter creation to private final value.

[pitagoras3]

suggestion: I've refactored this method to:

private void initializeTimestamps(String startTimestamp, String endTimestamp) {
    if (startTimestamp == null || endTimestamp == null) {
        this.startTimestamp = null;
        this.endTimestamp = null;
        return;
    }

    for (DateTimeFormatter formatter : formatters) {
        try {
            this.startTimestamp = formatter.parse(startTimestamp, Instant::from);
            this.endTimestamp = formatter.parse(endTimestamp, Instant::from);
            break;
        } catch (DateTimeParseException ignored) {}
    }
}

where formatters is:

private final List<DateTimeFormatter> formatters = List.of(
        DateTimeFormatter.ISO_INSTANT,
        DateTimeFormatter.ofPattern("yyyy-MM-dd'T'HH:mm'Z'").withZone(ZoneId.of("UTC"))
);

What do you think about it?

[moscicky]

Aren't lines 45 and 46 redundant?

[szczygiel-m]

Thats right, we could just assign one of them and the validation will fail, but I decided to do it for both to be more consistent and clear.