Repository navigation
consistent typing for lists - #457
Merged
Merged
Conversation
devtools now derives `tagged` on unions and records: a union is tagged if every arm begins with a keyword. LAK's `bedleak` and MAW's `head_limit` become untagged unions of a word (NONE, OFF) and a number. Track both: - Codegen tells a union column (each arm one value: OBS's id, `bedleak`) from a keystring (arms are line forms) by its arms, not by `tagged`, which LAK's `laksetting` (arms led by a lake number) no longer sets. - A keyword arm types as a literal: `Union[Literal["NONE"], float]`. It reads inline, matched in any case, else as a number. - An arm of a keyword and one value names the value after the keyword, like a tagged scalar arm, so MAW's `HeadLimit.head_limit` keeps its name. - `_is_instance` checks every arm of a union, so `str` fields no longer need to accept numbers. - The grammar prefixes a tagged `valid` string's name itself, since the union made of it no longer says it's tagged. Grammars are unchanged. - Bump modflow-devtools to 7df4ae97. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A tagged list's record (SCALE <factor>) keeps its value's name. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- A union arm types by its kind: a keyword is a literal word, an untagged value its own type, and a record or tagged value a class. OC's ocsetting is `Literal['ALL', 'FIRST', 'LAST'] | Oc.Frequency | Oc.Steps`; `Oc.All`, `Oc.First` and `Oc.Last` are gone. This replaces `is_column_union`. - A list's element class mirrors the DFN's element. A record with a union field (LKT, LKE, MAW, SFR, SPC, TVK, TVS) is one row class holding the index and the union, not one class per arm with the index copied in: `Lkt.StressPeriodData(ifno=0, laksetting=Lkt.Status(status="active"))`. Flat tuples like `(0, "STATUS", "ACTIVE")` still work. LAK is wrapped too, until devtools makes its element the union itself, as PRP's is. - A row's children include those in its union arm (SFR's CROSS_SECTION TAB6 file). - Static list columns no longer take a block prefix for sharing a name with a period setting keyword; period settings aren't package attributes, so nothing collided. - Drop the "keystring" term (v1's union type) from code, docs and tests. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
devtools now makes LAK's period element its laksetting union (modflowpy#391), so LAK's rows are its arms again, as before. Also put `# noqa: E501` only on long union annotations, not on every column ruff wraps. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
`Maw.StressPeriodData(ifno=0, mawsetting=("status", "active"))` was
accepted and written as a raw tuple. A union field with arm classes now
converts, on init and on assignment, a tuple to the arm its keyword
picks, or a word to the word, and raises if no arm takes the value.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
devtools' `mf sync` now fails if any registry does, and the mf6/example models registry is missing its models.toml. Fail on a DFN sync failure, but only warn on a models one for now. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
pwsh exits with the last native command's code, so Windows still failed on the models sync. Pass GITHUB_TOKEN for the DFN sync's API call, which devtools will send once it authenticates that lookup. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
wpbonelli
changed the title
wpbonelli
marked this pull request as ready for review
October 13, 2026 00:25
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
wpbonelli
added a commit
that referenced
this pull request
Oct 13, 2026
Follow-up to #457. The OC reshape missed PRP which has the same form # before Prp(stress_period_data={0: [Prp.All(), Prp.Frequency(2)]}) # after Prp(stress_period_data={0: ["ALL", Prp.Frequency(2)]})
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
1 participant
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.
Use MODFLOW-ORG/modflow-devtools#389 and MODFLOW-ORG/modflow-devtools#391 for consistent list field typing
Up to now some packages' period data has been typed as a list of records, where the record is just a leading feature number followed by a union. Codegen used to copy the index into each of the union arms to artificially produce a list of unions. This gave a slightly less verbose API at the cost of harder-to-predict typing. Remove this so the generated field typing follows the DFNs predictably.
Plain tuples like
(0, "STATUS", "ACTIVE")can still be passed instead of the typed form. This all applies to MAW, LKT, LKE, SFR, SPC, TVK and TVSOther changes: