pikers
/
piker
4
1
Fork 0
Code Issues 20 Pull Requests 11 Packages Projects Releases Wiki Activity

how_to_show_ur_pp: fixes for end-2-end order/position display #60

Merged
goodboy merged 11 commits from how_to_show_ur_pp into main 2026-01-07 19:32:55 +00:00
Conversation 0 Commits 11 Files Changed 10
goodboy commented 2026-01-07 18:24:57 +00:00
Owner

Since apparently we can’t (yet) rely on our fqme (fully-qualified-market-endpoint)as consistent, e2e uniquely matchable keys for various brokerd -> emsd -> chartd event relaying, namely bc (at least from the .brokers.ib backend..) the venue-token-part can change depending on,

  • where the order was cleared, possibly on multiple (sub-)venues which are not the primary listing venue.
  • how cleared orders are recorded in the broker backend’s blotter/ledger normally by time but with a varying venue field.
  • depending on piker’s txn-ledger processing (again normally sorted by record time stamp) the last venue is likely not the primary; the fqme might not match that of the data feed’s MktPair.fqme.

This problems becomes most obvious with our chart-trading UX where position markers/annotations might not be displayed on the correct live time-series view due to such a mismatch of MktPair.fqme -> BrokerPosition.fmqe.

So instead (at least for now) just always relay through the broker’s own unique-mkt-ID schema (what we dub the bs_mktid: backend-sys market ID) on a new (5b91b089) BrokerdPosition.bs_mktid field to guarantee consistency at least on a per-broker reporting basis.

Obviously ensure we always set it for .ib emitted pp msgs (58654915) and ensure the .ui.order_mode prioritize the mkt-match on .bs_mktid (388a9a4d).

Related/surrounding broker txn ledger improvements

  • (b6d70d50, 7b68444c) track and ignore any txn entries with invalid time-stamps (again such as can be recorded by the .ib backend..) by refining the .accounting.calc.iter_by_dt() impl with a handy _invalid: list passed to the embedded dyn_parse_to_dt() closure which allows us to either,

    • crash-handle such cases when a new debug: bool param is set OR,
    • at the least log.error() such txns to console.

  • d67ace75 avoid overriding Account.pps: dict entries by again using a bs_mktid to uniquely key Position entries in .accounting._pos.open_account() as delivered by Account.pps.

  • 0e9b50de _ems: tolerate and warn on already popped execs,

    such that we never crash on a status_msg > = book._active.pop(oid) in the ‘closed’ status handler whenever a double removal happens.

  • c609858f ui._remote_ctl: shield remote rect removals

    Since under trio-cancellation the .remove() is a checkpoint

  • f5850fe5 draft a ems/txn-ledger test which needs to eventually have checks on an example ledger file ensuring a position cleared over multiple venues (and thus currently mapping over multiple .fqme keys in its txn set) renders to a single, unified and correct position measure/report.

Since apparently we can't (yet) rely on our fqme (fully-qualified-market-endpoint)as consistent, e2e uniquely matchable keys for various `brokerd` -> `emsd` -> `chartd` event relaying, namely bc (at least from the `.brokers.ib` backend..) the venue-token-part can change depending on, - where the order was cleared, possibly on multiple (sub-)venues which are **not the primary** listing venue. - how cleared orders are recorded in the broker backend's blotter/ledger normally by time but with a varying venue field. - depending on `piker`'s txn-ledger processing (again normally sorted by record time stamp) the last venue is likely not the *primary*; the fqme might not match that of the data feed's `MktPair.fqme`. This problems becomes most obvious with our chart-trading UX where position markers/annotations might not be displayed on the correct live time-series view due to such a mismatch of `MktPair.fqme` -> `BrokerPosition.fmqe`. So instead (at least for now) just always relay through the broker's own unique-mkt-ID schema (what we dub the `bs_mktid`: backend-sys market ID) on a new (5b91b089) `BrokerdPosition.bs_mktid` field to guarantee consistency at least on a per-broker reporting basis. Obviously ensure we always set it for `.ib` emitted pp msgs (58654915) and ensure the `.ui.order_mode` prioritize the mkt-match on `.bs_mktid` (388a9a4d). --- #### Related/surrounding broker txn ledger improvements - (b6d70d50, 7b68444c) track and ignore any txn entries with invalid time-stamps (again such as can be recorded by the `.ib` backend..) by refining the `.accounting.calc.iter_by_dt()` impl with a handy `_invalid: list` passed to the embedded `dyn_parse_to_dt()` closure which allows us to either, - crash-handle such cases when a new `debug: bool` param is set OR, - at the least `log.error()` such txns to console. - d67ace75 avoid overriding `Account.pps: dict` entries by again using a `bs_mktid` to uniquely key `Position` entries in `.accounting._pos.open_account()` as delivered by `Account.pps`. - 0e9b50de `_ems`: tolerate and warn on already popped execs, > such that we never crash on a `status_msg > = book._active.pop(oid)` > in the 'closed' status handler whenever a double removal happens. - c609858f `ui._remote_ctl`: shield remote rect removals > Since under `trio`-cancellation the `.remove()` is a checkpoint - f5850fe5 draft a ems/txn-ledger test which needs to eventually have checks on an example ledger file ensuring a position cleared over multiple venues (and thus currently mapping over multiple `.fqme` keys in its txn set) renders to a single, unified and correct position measure/report. * hopefully i won't forget about actually finishing this suite in follow-up with https://www.pikers.dev/pikers/piker/issues/61 !!
goodboy force-pushed how_to_show_ur_pp from 7c11bdb859 to 58654915ac 2026-01-07 18:41:28 +00:00 Compare
goodboy added 1 commit 2026-01-07 19:22:04 +00:00
7b68444c7a accounting.calc: `.error()` on bad txn-time fields.. 
Since i'm seeing IB records with a `None` value and i don't want to be
debugging every time order-mode boots up..

Also use `pdb=debug` in `.open_ledger_dfs()`

Note, this had conflicts on `piker/accounting/calc.py` when rebasing
onto the refactored `brokers_refinery` history which were resolved
manually!
goodboy merged commit 9e82a46c0b into main 2026-01-07 19:32:55 +00:00
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels
No items
No Label
Milestone
Clear milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
No Assignees
1 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: pikers/piker#60
There is no content yet.
Delete Branch "how_to_show_ur_pp"

Deleting a branch is permanent. Although the deleted branch may exist for a short time before cleaning up, in most cases it CANNOT be undone. Continue?