- Extended shopping_total cache TTL from 1h to 24h
- Added inline price fallback: when cache is empty/stale, computes total
from shopping_price_cache.json (no AI calls); joins shopping_list with
products to get canonical shopping_name; tries both v3 and legacy v0
key formats to maximise cache hit rate; works in both internal and
Bring shopping modes (removed isShoppingBringMode guard — table is
always populated by sync)
- Fixed haInventorySensor + haRefreshPrices: shopping_list has no
quantity/unit/checked columns; changed to SELECT name with
COALESCE(p.shopping_name, sl.name) join, defaults qty=1/unit=pz
- Extend isExpiringSoon threshold: 3d -> 7d
- Expired items: add isRegular/buyCount>=2 guard so one-off
expired products don't appear in shopping list (expiry
banner already covers them)
- Expiring-soon block: require isRegular for 7-day window;
add 'willExpireBeforeUsed' check (daysLeft > daysToExpiry);
new reason string 'Scade in Ngg — ricompra' when stock is
adequate but won't be consumed in time
When adding a new pack of a product that already has an opened row
in inventory (opened_at IS NOT NULL), the previous code merged the
new stock into the opened row, corrupting opened_at tracking and
hiding the second pack from the anomaly model.
Now: search only for sealed rows (opened_at IS NULL) to merge into.
If only opened rows exist, INSERT a new sealed row instead.
getConsumptionPredictions now aggregates total qty across all
inventory rows for the same product_id before flagging.
If totalQtyAllRows >= expectedQty, the anomaly is suppressed
(stock is healthy, just split across opened+sealed rows).
Also uses aggregated total as the displayed actual_qty.
Products like salt/spices that are never marked per-use now get
consumption rate estimated from the average time between restocks:
avgCycleDays = (lastIn - firstIn) / (buyCount - 1)
estimatedDaysLeft = avgCycleDays - daysSinceLastBuy
Requirements: buyCount >= 3, dailyRate == 0, avgCycle >= 7 days.
Appears in smart shopping list with reason 'Finisce tra ~Ngg (ciclo medio Mgg)'.
Also marks buy-cycle products as isRegular so stock checks apply.
- Add _getFpCachePath(), _loadFpCache(), _saveFpCache() helpers
- Check data/reported_issue_fps.json before GitHub Search API
(falls back to /tmp/ when data/ is not writable)
- Save new issue number to cache immediately after creation
- Apply 30-minute comment throttle per fingerprint
- Fall back to GitHub Search on first run / cache miss
Fixes root cause of ~50 duplicate issues (#134 duplicates #135-#183)
caused by GitHub Search API indexing delay.
2026-05-29 06:02:27 +00:00
3 changed files with 262 additions and 39 deletions
@@ -11,6 +11,42 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **Recipe scraps tips** — During cooking steps, detect "waste" generated (peels, cores, bones, eggshells, coffee grounds, citrus zest, etc.) and surface AI-powered tips on how to reuse them (compost, natural cleaner, broth, candied peel, etc.). Could be shown as an optional collapsible hint card below the step that generates the scrap.
## [1.7.33] - 2026-05-29
### Fixed
- **HA sensor `shopping_total` always null** — `haInventorySensor` was reading `shopping_total_cache.json` with a 1-hour TTL (cache populated only by the JS frontend, so it was often empty). Extended TTL to 24 hours and added an inline fallback: when the cache is absent or stale, the sensor now computes the total directly from `shopping_price_cache.json` without any AI calls. Queries `shopping_list` joined to `products` for the canonical `shopping_name`, then looks up both v3 and legacy v0 cache key formats to maximise hit rate. Works in both internal and Bring shopping modes.
- **HA `ha_refresh_prices` using non-existent columns** — `haInventorySensor` and `haRefreshPrices` were querying `quantity`, `unit`, `checked` from `shopping_list` — columns that do not exist in that table (schema: `id, name, raw_name, specification, added_at, sort_order`). Changed to `SELECT name` with `shopping_name` join and default `qty=1 / unit=pz`.
## [1.7.32] - 2026-05-29
### Changed
- **Smarter expiry u2192 shopping list logic** — The "expiring soon" threshold is now 7 days (was 3), giving enough time to plan the next shopping trip. Items expiring soon are only flagged for restocking when the user is a **regular buyer** (`isRegular`) and either stock is low (<50%) or the consumption rate predicts the item will expire before being used. Non-regular products keep the old 3-day safety-net. Expired items are now only added to the shopping list when `isRegular || buyCount >= 2` — products that expired unused without ever being a staple no longer pollute the list; the expiry banner handles them.
## [1.7.31] - 2026-05-29
### Fixed
- **New pack merges into opened pack on add** — `addToInventory` was looking for ANY existing row for the same product+location and adding the new quantity to it. This caused a newly purchased sealed pack to be silently merged with an already-opened pack, collapsing two physically distinct containers into one row and corrupting the `opened_at` timestamp. The fix now searches only for a **sealed** (unopened) row (`opened_at IS NULL`) to merge into. If only opened rows exist, a new sealed row is created instead — keeping the two packs separate and allowing the anomaly model and shelf-life tracker to work correctly.
## [1.7.30] - 2026-05-29
### Fixed
- **False consumption anomaly with multi-row stock** — The anomaly detection banner was evaluating each inventory row in isolation. Products split across multiple rows (e.g. one opened pack with 1 pz + one sealed pack with 6 pz) incorrectly triggered a "consumed faster than expected" warning because only the opened row (1 pz) was compared against the model. The check now aggregates the total quantity across all rows for the same product before deciding to flag an anomaly. If the combined total ≥ expected remaining, the anomaly is suppressed.
## [1.7.29] - 2026-05-29
### Added
- **Buy-cycle consumption prediction** — Products that are never tracked per-use (salt, spices, cleaning supplies, etc.) now use the average time between restocks as a proxy for consumption rate. When a product has ≥ 3 purchase events and no individual `out` events, EverShelf calculates the average buy cycle (`(lastBuy - firstBuy) / (buyCount - 1)`) and estimates how many days of stock remain in the current cycle. The product appears in the smart shopping list with a reason like "Finisce tra ~12gg (ciclo medio 75gg)" before it runs out, rather than only after. These products are now also treated as `isRegular` so all stock-level urgency checks apply correctly.
## [1.7.28] - 2026-05-30
### Fixed
- **Duplicate auto-reported issues** — The GitHub issue reporter was relying solely on the GitHub Search API for deduplication. Because search indexing has a several-minutes lag, rapid error recurrences each created a new issue before the previous one was indexed, producing ~50 duplicate issues. The reporter now uses a local file cache (`data/reported_issue_fps.json`, with `/tmp/` fallback when `data/` is not writable) as the primary deduplication store. A 30-minute per-fingerprint comment throttle is also applied to prevent flooding an existing issue. GitHub Search is used only on first run or after a cache miss. Closes [#134](https://github.com/dadaloop82/EverShelf/issues/134) (and all duplicates #135–#183).
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.