Fix some TZ issues after sanity checking status display#19
Open
alexis-cousein wants to merge 12 commits into
Open
Fix some TZ issues after sanity checking status display#19alexis-cousein wants to merge 12 commits into
alexis-cousein wants to merge 12 commits into
Conversation
alexis-cousein
force-pushed
the
fix_state_dt_tz
branch
from
cb72ab0 to
25a19cf
Compare
…ith Local Sidereal Time
Regression tests for SharedStateObj set_datetime / utc_datetime / local_datetime: a naive input is interpreted as UTC, an aware-local input is normalised to UTC in storage (so bare datetime() stays UTC), and UTC vs local read-back differ at a non-UTC location — the status-screen bug PR brickbots#508 fixes. The clock is frozen so datetime()'s drift term is zero and the assertions are exact.
…bots#508 ADR-0018 + Positioning CONTEXT.md glossary: shared_state civil datetime is stored timezone-aware in UTC, normalised at the set_datetime() boundary, and read via utc_datetime()/local_datetime(); distinguish civil datetime (the astronomical epoch) from measurement epoch (time.time() fix-age). Captures the GPS LST->LKT rename and the no-naive-local-persisted-data finding.
Docstrings on datetime/utc_datetime/local_datetime/set_datetime stating the ADR-0018 contract: set_datetime is the single normalising write boundary, callers pass tz-aware datetimes, and the stored value is always UTC-aware.
…orce DTZ Add PiFinder/timez.py as the single sanctioned source of datetimes (utc_now / local_now / utc_from_timestamp / parse / naive / utc) and enable ruff's DTZ (flake8-datetimez) family repo-wide, exempting only timez.py. Every bare datetime.now() / datetime(...) / strptime() / fromtimestamp() now routes through timez, so 'local or UTC?' is explicit at the call site and a naive value cannot silently reach the astronomy/ephemeris path. Also fixes a latent relabel bug in the comets.py freshness check (a file mtime read naive-local and stamped UTC -> now the true UTC instant via timez.utc_from_timestamp). See ADR-0018.
Civil-datetime UTC contract: docs, regression test, and timez/DTZ funnel
…directly creating them
alexis-cousein
force-pushed
the
fix_state_dt_tz
branch
2 times, most recently
from
5e02fd7 to
c72f470
Compare
alexis-cousein
force-pushed
the
fix_state_dt_tz
branch
from
c72f470 to
19296ea
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
I had noticed that when you input time/date "UTC TM" and "LCL TM" were always identical, and that when recovering time from GPS there was the expected gap between them, so I sanity checked some of the datetime timezone usage in the current code.
-The current date/time input routines are self-consistent but they store a datetime in shared_state that contains a timezone reference (to the local timezone the times were specified in), while when using the GPS everything is in UTC) .
-[bug] The status screen that displays "UTC TM" just expects the shared state to be in UTC.
Since other consumers of the shared state datetime might make the same mistake, I corrected this at both ends:
-When saving the shared state, always convert to a timezone-aware datetime that is UTC based, so that manual input time and GPS time will always behave similarly.
-Introduced a shared_state query routine utc_datetime in addition to local_datetime to avoid the ambiguous "datetime" (which just returns a valid datetime, but these can be in different timezones); the "TM UTC" visualisation now uses this.
Other changes:
-replaced pytz.timezone("UTC") by the initialized singleton pytz.utc -- more readable and faster, no need to reconstruct this object time and again.
-gps_fake now explicitly uses a UTC-timezone-stamped datetime.now(pytz.utc) , just to avoid sending "naive" datetimes around.
-catalogs.py was extracting a naive datetime.now() --expressed in the OS timezone-- and just tagged on "UTC" to make a timezone-aware datetime. This is only correct if the OS timezone is UTC, and that is unnecessarily dangerous: datetime.now(pytz.utc) is guaranteed to give you the actual time expressed in UTC, and tagged as such.
-[bug] The use of datetime.now() to get "today's date" in set_time (so that the screen for data input gets initialized to the correct date) to combine it with the time that was input is just wrong, because "today's date in UTC" might be different from the date in the target timezone (which we force before setting it), leading us to set the datetime 24 hours off in some cases. There, we need to use datetime.now(timezone) to avoid the wrong date being proposed in the datetime screen.