Hacking PNR <a href="https://media.ccc.de/v/33c3-7964-where_in_the_world_is_carmen_sandiego" rel="nofollow">https://media.ccc.de/v/33c3-7964-where_in_the_world_is_carme...</a> <i>at time of the talk</i> (2 years ago) used/discussed these sites:<p><a href="https://matrix.itasoftware.com/" rel="nofollow">https://matrix.itasoftware.com/</a> to get detailed fare rules & restrictions<p><a href="https://www.expertflyer.com" rel="nofollow">https://www.expertflyer.com</a> & <a href="https://itunes.apple.com/us/app/seat-alerts/id533533342" rel="nofollow">https://itunes.apple.com/us/app/seat-alerts/id533533342</a><p><a href="https://www.checkmytrip.com/" rel="nofollow">https://www.checkmytrip.com/</a> allows entering name & booking-reference and gives all detail<p><a href="https://www.viewtrip.com/" rel="nofollow">https://www.viewtrip.com/</a><p><a href="https://tripcase.com" rel="nofollow">https://tripcase.com</a> will use name + booking-reference to get detailed PNR information including the first name<p><a href="https://www.fly.kiev.ua" rel="nofollow">https://www.fly.kiev.ua</a> can make a flight reservation without a payment, still gives an 6char booking-reference
Looks cool! The problem of enumerating the best routes between two given airports is indeed an interesting problem. My favorite discussion of the problem is from Carl de Marcken: <a href="http://www.ai.mit.edu/courses/6.034f/psets/ps1/airtravel.pdf" rel="nofollow">http://www.ai.mit.edu/courses/6.034f/psets/ps1/airtravel.pdf</a><p>I've been involved in solving a similar problem, on the award flight search side. There are about 12,000 direct commercial routes offering award redemption, and the airline websites have different limitations: some only let you search one-way or only round-trip, or only one cabin at a time, or only give hints about available quantities. So you have to formulate the minimal set of queries that will give you complete information. And then given an origin and destination, figure out the valid routings, like you've discovered.<p>The other really interesting challenge with many of the airline websites are the extreme measures they go to, to block you from crawling their data. My favorite is British Airways, which runs 800 different tests to fingerprint your browser, from generating audio files to drawing to an off-screen canvas, all inside a custom VM with encrypted byte code. (Similar to how I've seen Google obfuscate the ReCaptcha code base.) Surprising to hear that Ryanair makes it so easy to access prices in comparison!
Off-topic but related: <a href="https://scottscheapflights.com" rel="nofollow">https://scottscheapflights.com</a> is _amazing_. The Premium sub is worth every penny...my wife and I will be doing a ton more overseas travel in 2019 purely because of the deals this site finds.
As a CS student, I agree it is really refreshing to see basic graph algorithms to be used in a practical, non academic/industry context. I never really see these basic algorithms used in small side projects, so thank you for this!
We (me and friend) are working on the same problem, however we are stuck at the next step. Namely the user interface: how do you let the user search through all the possibilities.<p>We created some data structures that allow almost instant searching of such routes and we have scrapers running regularly on Ryanair, easyJet, wizzair, and Transavia. You can query the algorithm here: <a href="https://algo.tripchemy.com/routes/TSF?year=2019&month=02&day=10&length=20" rel="nofollow">https://algo.tripchemy.com/routes/TSF?year=2019&month=02&day...</a><p>Where you want to have a trip of 20 days (4 lovatons, 5 flights) from 10 Feb. Starting and returning to TSF. We also added some scoring for niceness of airports.
back when I was a student, I would sometimes spend hours on Ryanair's website to find that type of deals.<p>My base was Paris and I wanted to go to Stockholm for example. The direct Ryanair flight from Paris to Stockholm was expensive (Let's say 150 euros).
I would manage to find two cheap flights that go for example from Paris to Rome and another one from Rome to Stockholm, both of those for 10 Euros each. The key is to use the big Ryanair hubs as gateway (Milan and Brussels would work really well back in the days)<p>I would basically lose my whole day traveling to save 80 euros.
Sometimes it backfires though, and when it backfires you are on your own. Once my first flight got super delayed, and I got stuck in Milan for the night. I managed to have Ryanair rebook my second flight for the day.<p>I think Ryanair wants to be able to propose high prices for direct flights and prevent people from finding those cheaper indirect flights. That's why they don't make it easy to find them.
I use <a href="https://airwander.com/" rel="nofollow">https://airwander.com/</a> when flying overseas, I don't know how well it works for low cost airlines though.<p>It's a flight search engine that let you filter what connections there are between the cities you want to travel and help you book a single flight that is going to stay there as a stopover. Sometimes it finds some weird flights that doesn't make any sense if you're just trying to reach your destination but if you're just traveling it's quite nice.<p>Last time I used it I found a overseas flight that stopped in Lisbon but on the way back made a connection in Bologna that was +200 cheaper then the normal flight.
Great work, this is really cool!<p>For multi city flights, I've found Google flights does things really well from a UX standpoint, and I believe they support Ryanair.
Why is there no interface that allows searching for flights without specifying <i>exact</i> dates? All I want to do is fly from A to B to A. Maybe stay a week, maybe two. I'm flexible - it doesnt have to be exactly N days starting on the M'th.<p>Even services that "allegedly" advertise they have that, still require manually clicking every day combination.<p>I'd fly more often <i>if it was easy</i>. The few times I tried, it would be just faster to drive than go through all possible 900+ day combinations.
Finding a cycle of a given length can actually be done in time `O(e^k n^2.3)` where `k` is the length of the cycle, using the cool technique of color coding.<p>Basically, to find a cycle of length 6, you repeatedly color every node in one of 6 colors at random. Then you look for a cycle with all different colored nodes, which can be done efficiently using dynamic programming.<p>You succeed in finding a particular 6 cycle exactly if each node in it has gotten a different color, but the chance of this is not bad. In the case of the OP it's even better, since there are probably many such cycles to find.
Somewhat off-topic and shameless plug, I'm working on a meta flight-search engine that finds the cheapest flights from multiple origins to the same destination (i.e. me and some friends who live in another city want to fly to Hong Kong with no particular date in mind, just want the cheapest flights and with overlapping departure/return dates). Does anyone else have a similar need or know of anything that can do this?
@jonluca<p>If you are looking to take this further, you can find some more info on Ryanair API endpoints here:<p><a href="https://gist.github.com/vool/bbd64eeee313d27a82ab" rel="nofollow">https://gist.github.com/vool/bbd64eeee313d27a82ab</a>
ryan air weekend flight search ... pretty handy<p><a href="https://github.com/cloudsquall/flightsearch" rel="nofollow">https://github.com/cloudsquall/flightsearch</a>