improve connection detection heuristic
This commit is contained in:
parent
35cc18dfae
commit
a9850d073e
1 changed files with 21 additions and 1 deletions
|
@ -13,8 +13,28 @@ sub mark_if_missed_connection {
|
|||
my $wait_time
|
||||
= ( $next_journey->{rt_departure}->epoch - $journey->{rt_arrival}->epoch )
|
||||
/ 60;
|
||||
|
||||
|
||||
# Assumption: $next_journey is a missed connection (i.e., if $journey had
|
||||
# been on time it would have been an earlier train)
|
||||
# * the wait time between arrival and departure is less than 70 minutes
|
||||
# (up to 60 minutes to wait for the next train in an hourly connection
|
||||
# + up to 10 minutes transfer time between platforms)
|
||||
# * the delay between scheduled arrival at the interchange station and
|
||||
# real departure of $next_journey (which is hopefully the same as the
|
||||
# total delay at the destination of $next_journey) is more than 120
|
||||
# minutes (-> 50% fare reduction) or it is more than 60 minutes and the
|
||||
# single-journey delay is less than 60 minutes (-> 25% fare reduction)
|
||||
# This ensures that $next_journey is only referenced if the missed
|
||||
# connection makes a difference from a passenger rights point of view --
|
||||
# if $journey itself is already 60 .. 119 minutes delayed and the
|
||||
# delay with the connection to $next_journey is also 60 .. 119 minutes,
|
||||
# including it is not worth the effort. Similarly, if $journey is already
|
||||
# ≥120 minutes delayed, looking for connections and more delay is
|
||||
# pointless.
|
||||
|
||||
if (
|
||||
$wait_time < 120
|
||||
$wait_time < 70
|
||||
and ( $possible_delay >= 120
|
||||
or ( $journey->{delay} < 60 and $possible_delay >= 60 ) )
|
||||
)
|
||||
|
|
Loading…
Reference in a new issue