Feat/budgets - #752
Merged
Merged
Conversation
feat: add Simplified Chinese.
- zh_CN: 中位数字消费 -> 中位数消费; the stray 字 read as "median digit"
- zh_CN/zh_TW: add missing measure word to upcomingCharges
({count} 笔/筆预定扣款), matching every other count string
- en: deleteCloudBackupConfirmation said "This action is not
irreversible!" — a double negative telling users a destructive
backup deletion could be undone
Claude-Session: https://claude.ai/code/session_01JUAzk892137fKesAsZbGnN
Routine bump, along with the transitive test/meta/matcher/vector_math versions pub resolved with it. Claude-Session: https://claude.ai/code/session_014o8coDD67GU2x4P5nxSrFZ
Pending transactions were excluded from budget spend outright. That makes a budget read healthy right up until a scheduled payment clears, at which point it lurches. Rent dated the 28th is money the period is already committed to; leaving it out means the number answers "what has cleared" while the user is asking "how much room is left". So `spent` now includes pending, and everything derived from it -- ratio, status, insights, alerts -- is measured on the committed total. The pending slice stays recoverable as `pendingSpent`, because counting it is not the same as pretending it has already gone: every bar fills solid to `confirmedRatio` and continues to `ratio` as a lighter ghost. `projectedRatio` deliberately does not extrapolate the pending part. A scheduled lump sum is already dated, so running it through the run rate would count it once for being committed and again for every day left in the period -- rent on the 28th would project a 4.5x overshoot on the 2nd with nothing actually spent. Confirmed spend is extrapolated as before, the pending remainder is added at face value. BulletChart is rewritten alongside it. The track had 10% headroom past whichever of value/target was larger, so a bar sat at ~76% of its width while the label beside it read 84%, and the same budget drew differently depending on how far over it was. It now spans exactly 0..target: an overrun fills the track and recolours instead of rescaling it. The old qualitative band went with it -- it marked the target, which is now the end of the track -- and the target tick became a pace tick, neutral rather than `primary`, since a green accent colour made it the same hue as a healthy bar. payloadVersion goes to 3: `spent`/`ratio` mean something new, and a v2 extension would draw the larger figure as though all of it had cleared. Both extensions default `confirmedRatio` to `ratio` when decoding, so the degraded case is one solid bar rather than an all-ghost one. Glance has no alpha modifier, so the Android ghost is three pre-multiplied colour resources rather than a modifier on the status colour. Claude-Session: https://claude.ai/code/session_014o8coDD67GU2x4P5nxSrFZ
The Android budget widgets built a bare `Intent(context, MainActivity)`. That carries no action and no data, so every tap cold-opened the home tab -- a widget showing "Groceries, 112%" dropped you somewhere with no budget on it, and the roll-up sent you to the same place as the pinned one. The quick-entry widgets already deep-link through the `flow-mn` scheme the manifest accepts, so this just uses the same road. The intent is pinned to our own package. `flow-mn` is a custom scheme, which any installed app may register for, so without setPackage a widget tap could raise a chooser or open a Flow-lookalike. On the app side, the URI handler is an allowlist of three paths, not a catch-all forward. Anything on the device can claim a custom scheme, so forwarding an arbitrary path from one into the router would be an open redirect into any route in the app -- including the destructive ones. Unmatched URIs are logged and dropped. `/budgets/:id` still routes through the existing redirect, so an id that no longer resolves lands on the list instead of a dead page. The iOS roll-up moves from `/budgets` to `/stats/budgets` for the same reason the Android one points there: that widget is the overview in miniature, so a tap should expand what it shows rather than drop you on the adjacent list. The Android chrome changes with it. The widgets drew a rounded card inside an 8dp inset inside the launcher's own rounded container, which reads as a border around a screenshot rather than as the widget itself. One `appWidgetBackground()` surface now *is* the widget, at the system radius, so the corners match every other widget on the home screen instead of a hard-coded 16dp that only agrees with some launchers. Claude-Session: https://claude.ai/code/session_014o8coDD67GU2x4P5nxSrFZ
It sat as a ListTile above the budget cards, so the one action on the page looked like the first item in the list it creates into. Claude-Session: https://claude.ai/code/session_014o8coDD67GU2x4P5nxSrFZ
zh_CN.json was contributed against develop, which has no budgets, so merging this branch left it 66 keys short of en.json and the integrity test red. Machine-translated, like every other locale here, and worth a native pass before release. Inserted in en.json's key order rather than appended: json_integrity_test compares positionally, so scripts/translate_missing alone would not have satisfied it. Claude-Session: https://claude.ai/code/session_014o8coDD67GU2x4P5nxSrFZ
Claude-Session: https://claude.ai/code/session_014o8coDD67GU2x4P5nxSrFZ
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
3 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.
No description provided.