I am stuck at a very weird problem. I mean I do not how to threat this.
Basically, I got a webpage where I list all stops for a defined bus route (example, route_id = 141). Sometime, a route, can start from point A and go to point B so theoretically from one terminus to another terminus. BUT, sometime, the same route can begin its trip somewhere between A and B and go at point A or B.
You can see a live working example at this address : http://stm.tranzit.ca/bus/stops/85/Hochelaga/
My super draw :
Regular trip
Start(A) ----------------------------------- End(B)
Sometime
Start --------------------- End(B)
Start(A) ---------------------- End
etc.. Its always going to be between point A and B. 90% of the time it’s from A to B but the other 10% it can be anywhere between there until another point between there (sometime is from middle to A or B).
Ok. Before I was sorting my result using stop_sequence. I know the order of the stop sequence, so stop 1, stop 2, etc and how many stop by trip (a trip is from A to B, or A to middle, etc).
Now, I found a way to sort my result by trip and distinct them so now I know all the trip I got (A-B, B-A, A-middle, middle-B, etc..).
So if let’s say I got A-B, B-A, A-Middle and B-Middle, I have 4 possibles trips for a route. I then fetch all stop linked to it.
Example, this is a list for trip A to B for a defined route. Please note the value at stop_sequence 35.
1 #53014 Station Frontenac (Frontenac / Ontario)
... 2, 3. etc..
34 #53293 Honoré-Beaugrand / Roux
35 #54257 Station Honoré-Beaugrand / Sherbrooke
36 #53290 Saint-Donat / Sherbrooke
37 #53265 Saint-Donat / De Forbin-Janson
38 #54676 Saint-Donat / De Grosbois
39 #54674 Saint-Donat / Roi-René
40 #54672 Saint-Donat / les Reaux
41 #54668 Saint-Donat / Chénier
42 #54661 Joseph-Renaud / Yves-Prévost
43 #54646 Joseph-Renaud / Châtillon
44 #54629 Joseph-Renaud / Wilfrid-Pelletier
45 #54605 Joseph-Renaud / Châteauneuf
46 #54609 Châteauneuf / Vaujours
47 #54610 Châteauneuf / Rabelais
48 #54612 Châteauneuf / de la Loire
49 #54621 Châteauneuf / Roi-René
50 #54623 Châteauneuf / des Ormeaux
51 #54639 Châteauneuf / Rondeau
52 #54724 Rondeau / Georges
53 #54735 Rondeau / De La Vérendrye
54 #54738 Rhéaume / Chaumont
55 #54740 Chaumont / Guy
56 #54741 Chaumont / Azilda
57 #54742 Yves-Prévost / Azilda
58 #54659 des Ormeaux / Chaumont
Here’s a trip between middle and B.
1 #54257 Station Honoré-Beaugrand / Sherbrooke
2 #53290 Saint-Donat / Sherbrooke
3 #53265 Saint-Donat / De Forbin-Janson
4 #54676 Saint-Donat / De Grosbois
5 #54674 Saint-Donat / Roi-René
6 #54672 Saint-Donat / les Reaux
7 #54668 Saint-Donat / Chénier
8 #54661 Joseph-Renaud / Yves-Prévost
9 #54646 Joseph-Renaud / Châtillon
10 #54629 Joseph-Renaud / Wilfrid-Pelletier
11 #54605 Joseph-Renaud / Châteauneuf
12 #54609 Châteauneuf / Vaujours
13 #54610 Châteauneuf / Rabelais
14 #54612 Châteauneuf / de la Loire
15 #54621 Châteauneuf / Roi-René
16 #54623 Châteauneuf / des Ormeaux
17 #54639 Châteauneuf / Rondeau
18 #54724 Rondeau / Georges
19 #54735 Rondeau / De La Vérendrye
20 #54738 Rhéaume / Chaumont
21 #54740 Chaumont / Guy
22 #54741 Chaumont / Azilda
23 #54742 Yves-Prévost / Azilda
24 #54659 des Ormeaux / Chaumont
Now, as you can see, stop_sequence 1 here is the same as 35. Second stop is the same as stop 36 above. As you can see also, the count from stop 35 to 58 is 24. The same as the second trip.
Now, what I want to know, is how I can merge them using PHP ? I would like to take let’s say second group here and merge it inside the first one at the first occurence that match, so stop_sequence 1 from second group will become stop_sequence 35 because it match and it is the same.
Using this way, I can keep the stop order, not by using stop_sequence at the end of ordering, because as you saw, the second trip has stop_sequence 1 equal to 24 on the first trip. It mean that if I order by stop_sequence using MySQL, it will order sequentially all stop_sequence so the resulst will be stop_sequence 1 from first trip, sequence 1 from second trip, etc and my order will be wrong of all stops will be cleary wrong.
I want to do this in PHP but I do not know how to match result in an array and avoid X result after so I can match perfectly.
If you want to know what is my SQL query, here it is :
SELECT t.trip_id, t.trip_headsign, st.stop_sequence, s.stop_id, s.stop_code, s.stop_name
FROM trips AS t
LEFT JOIN stop_times AS st ON st.trip_id = t.trip_id
LEFT JOIN stops AS s ON s.stop_id = st.stop_id
WHERE t.route_id = 141
AND t.trip_id IN (
SELECT trip_id
FROM (
SELECT trip_id
FROM (
SELECT COUNT(*) AS count, trips.trip_id, trips.trip_headsign
FROM trips
LEFT JOIN stop_times ON trips.trip_id = stop_times.trip_id
WHERE route_id = 141
AND trips.trip_id LIKE (SELECT CONCAT(service_id, "%") FROM calendar_dates WHERE date = "20120628")
GROUP BY trips.trip_id
) a
GROUP BY count, trip_headsign
) a
)
GROUP BY t.trip_id, st.stop_id
ORDER BY t.trip_id ASC, st.stop_sequence ASC
trip_id define if it’s from A-B, B-A, A-middle, etc.
stop_sequence is the order of the stop for a defined trip
stop_id and stop_code share the same thing – unique id (one for user and one for the internal system)
trip_headsign is like 141-W or 141-N (as for west or north).!
EDIT:
With my query, I can get all stop for each trip (green and red here).
I want to merge this in one list by keeping the order.
example schema
Route can be from stop 1 (red) to stop 8. So trip 1 can be stop 1, stop 2, etc until stop 8.
Route can be from stop 1 (green) to stop 8. So trip can be stop 1, stop 7 and stop 8.
My query give me both trip stop’s list.
Now, I want to merge them as follow :
List:
– Stop 1 (red)
– Stop 2
– Stop 3
….
– Stop 6
– Stop 1 (green)
– Stop 7 (once even if he is in trip red and green)
– Stop 8 (once even if he is in trip red and green)
To summarize what has been said in the comments:
Optionally, you can start on either side of the shorter trip and remove any stops that don’t appear in the main trip.