Add filtering for not mine (STATS) swaps

[tonymorony]

It would be wonderful to use DB and filtration/pagination functionality not only for my_orders swap statuses but also for collected swap statuses on seed nodes

ref: #33

[artemii235]

@tonymorony Thanks for creating the issue! Could you please provide more details about the expected SQLite table data/schema? I guess we need the following columns: UUID, maker_coin, taker_coin, maker_amount, taker_amount, started_at (maybe separate for maker and taker too), but maybe something else is required?

[tonymorony]

yeah, that looks like enough data. also separation between successful and failed would be helpful

just rechecked data from the swap status structure we using:

first_event = swap["events"][0]["event"]["data"]

first_event["taker_amount"]
first_event["maker_amount"]
["events"][0]["timestamp"]
first_event['uuid']
first_event['started_at']

The most tricky matter we got in our stats tool so far is to determine the list of pairs with >0 swaps - we should scan all possible permutations. Is it possible to create such an "alive pairs" list during import into DB? it'll help a lot

[artemii235]

Is it possible to create such an "alive pairs" list during import into DB? it'll help a lot

Sure, just one question: should we add both RICK/MORTY and MORTY/RICK there? Or will it be sufficient to order tickers alphabetically and insert just MORTY/RICK?

[tonymorony]

counting logic using both ways now (according to dexstats.info not only we counting both ways) - so would be great to include it as RICK/MORTY + MORTY/RICK if there are both directions trades (if it's not a huge complication to list creation logic)

[artemii235]

@tonymorony I have implemented the indexing of swaps statistics to the SQLite DB. You can use the build from http://195.201.0.6/mm2.1-stats-swaps-index-filter/ until the corresponding PR is merged. The SQLite file path is %MM2_DB_DIR%/MM2.db. The table name with statistics is stats_swaps.

The most tricky matter we got in our stats tool so far is to determine the list of pairs with >0 swaps

You can now get them using the SELECT DISTINCT maker_coin, taker_coin FROM stats_swaps; query.

[tonymorony]

Works like a charm, thank you!

Parser (WIP) with this DB usage https://github.com/tonymorony/dexstats_sqlite_py