I know, this thread is a couple of years old, but I am looking to acquire a new set of CME historical market depth data in the FIX/FAST format, and I am wondering if anyone has written a FIX/FAST->JBT converter. If so, please let me know. Thanks.
On Wednesday, November 4, 2009 10:22:24 PM UTC-5, shaggsthestud wrote: > > Oh, and I also meant to mention - perhaps QuickFIX of some other > library could be used to generate a second implementation completely > separate from mine. The best test is to have two completely separate > parsers, and see if they give the same results. > > I can also see it being a complete waste of time.. but... data > integrity is pretty important for a trading system (but then again, > what isn't?). > > On Nov 3, 5:11 am, nonlinear5 <[email protected]> wrote: > > Instead of implementing FIX parser from scratch, you may want to reuse > > this open source package:http://www.quickfixj.org/ > > > > On Nov 3, 3:02 am, shaggsthestud <[email protected]> wrote: > > > > > I wasn't expecting this, but maybe I should: the file has all of the > > > futures contracts in it. I thought it was just for one expiration > > > date.. The plot thickens. > > > > > On Nov 2, 11:17 pm, shaggsthestud <[email protected]> wrote: > > > > > > Ok, I have made some progress. I can do the basic FIX processing: > > > > check message length and checksum, split up messages into > field/value > > > > pairs, check that message send order is preserved, divide a message > > > > into a header and groups of message blocks, throw exceptions. > > > > > > I need to clean up the code a bit (I am a bit of a Java newb), and > > > > then I will start decoding the individual message blocks. > > > > > > Some general notes observations: > > > > - The FIX specs I was able to find didn't tell me much. I signed up > > > > for a site membership at fixprotocol.org, and the PDF was a bit > > > > helpful, but not really. > > > > - The Wikipedia page told me about as much, and more (such as the > > > > checksum and length calcs), and it took only a few minutes to read. > > > > - I glanced over the basic idea from those and have been using the > CME > > > > doc to interpret the data. The structure doesn't look rigidly > defined > > > > anywhere in the text, so I am just going buy the patterns in the > > > > data. (For example, all data blocks seem to start with tag 279, so > > > > far. I wrote good exceptions to catch irregularities). > > > > - The FIX protocol states that messages should start with > "8=FIX...". > > > > The CME data does not start like that (although it does start with a > > > > tag with similiar meaning, I guess). I'm not sure what to make of > > > > that but I am ignoring that for now. > > > > - Each message (which take up 1 line of the data file) can have > > > > several "message blocks" inside it. Each block represents a market > > > > data update. So within the message you could have, for example, 5 > > > > different updates to the book, and/or different types of messages as > > > > well. > > > > > > On Oct 30, 10:11 am, shaggsthestud <[email protected]> wrote: > > > > > > > If no one has started, I am looking into it. Looks like a meaty > > > > > project. I don't want to start from scratch if someone else is > doing > > > > > it already. > > > > > > > Anyone started yet? -- You received this message because you are subscribed to the Google Groups "JBookTrader" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/jbooktrader?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
