Seed provenance and inventory: source links and seed lots (#50)
Build image / build-and-push (push) Successful in 4s
Gadfly review (reusable) / review (pull_request) Successful in 11m8s
Adversarial Review (Gadfly) / review (pull_request) Successful in 11m9s

"German Red Garlic" now links back to the Johnny's page it came from, and pansy
knows how much is left.

Two modelling calls, both of which look like omissions unless stated.

The plant catalog stays FLAT — no species/variety hierarchy. A row in your
catalog just is the thing you plant; the built-in "Garlic" is a generic starting
point you clone. A hierarchy buys taxonomy nobody asked for and complicates
every join.

Inventory belongs to a PURCHASE, not to a variety. You might buy the same garlic
from two vendors in two years at two prices and two germination rates; quantity
columns on plants would collapse all of that into one number and lose it. So
lots get their own table, and owner_id sits on the lot rather than being
inherited from the plant — you can buy generic built-in Garlic from Johnny's and
the record is still yours alone. Lots are never shared along with a garden.

`remaining` is derived, never stored: quantity minus the summed effective count
of the active plantings attributed to the lot. The alternative — decrementing a
column when you plant — drifts the moment anything edits or removes a planting
behind its back, and once it has drifted there's no way to tell what the true
number was. Removing a plop gives the seed back with nothing having to remember
to. The usage query returns raw radius/count/spacing rather than a SUM, because
the effective-count formula lives in the service and reproducing it in SQL would
guarantee the two drift apart.

source_url is validated as http(s) with a host, on both plants and lots. It
renders as a clickable link, so that check is load-bearing rather than tidiness:
without it a stored javascript: URL becomes script execution the moment someone
clicks it. Schemeless and relative URLs are refused too — they'd resolve against
pansy's own origin, which is never what a vendor link means.

A planting's lot must be the actor's AND a lot of the plant being planted.
Attributing a garlic planting to a tomato lot would silently corrupt every
remaining count downstream, so it's refused rather than stored, and re-checked
on update in case a plant swap invalidated a previously-valid attribution.

Deleting a lot leaves its plantings alone with a null link: the plants really
were in the ground, whatever happened to the record of the packet.

Closes #50

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_01H3zbym8Doka2d7D48maSgZ
This commit is contained in:
2026-07-21 01:20:50 -04:00
co-authored by Claude Opus 4.8
parent 3ec77a1099
commit 81481ede2f
15 changed files with 1223 additions and 31 deletions
@@ -0,0 +1,53 @@
-- Seed provenance and inventory (#50): where "German Red Garlic" came from, and
-- how much of it is left.
--
-- Two modelling calls worth stating, because both look like omissions otherwise:
--
-- 1. The plant catalog stays FLAT. No species→variety hierarchy. A row in your
-- catalog just is the thing you plant — "German Red Garlic" is a plant row,
-- and the built-in "Garlic" is a generic starting point you clone. A
-- hierarchy buys taxonomy nobody asked for and complicates every join.
--
-- 2. Inventory belongs to a PURCHASE, not to a variety. You might buy the same
-- garlic from two vendors in two years at two prices and two germination
-- rates. Quantity columns on plants would collapse all of that into one
-- number and lose it, so lots get their own table.
--
-- Nothing here is destructive: plants gains two defaulted columns, the built-ins
-- keep working with empty strings, and plantings gains a nullable link.
ALTER TABLE plants ADD COLUMN source_url TEXT NOT NULL DEFAULT '';
ALTER TABLE plants ADD COLUMN vendor TEXT NOT NULL DEFAULT '';
-- owner_id sits on the LOT rather than being inherited from the plant, because a
-- lot may reference a built-in plant (you can buy generic Garlic from Johnny's)
-- while still being nobody's business but yours. Lots are never shared along
-- with a garden.
CREATE TABLE seed_lots (
id INTEGER PRIMARY KEY,
owner_id INTEGER NOT NULL REFERENCES users (id) ON DELETE CASCADE,
plant_id INTEGER NOT NULL REFERENCES plants (id) ON DELETE CASCADE,
vendor TEXT NOT NULL DEFAULT '',
source_url TEXT NOT NULL DEFAULT '',
sku TEXT NOT NULL DEFAULT '',
lot_code TEXT NOT NULL DEFAULT '',
purchased_at TEXT, -- 'YYYY-MM-DD'
packed_for_year INTEGER,
quantity REAL NOT NULL DEFAULT 0,
unit TEXT NOT NULL CHECK (unit IN ('seeds','grams','ounces','packets','bulbs','plants')),
cost_cents INTEGER,
germination_pct REAL,
notes TEXT NOT NULL DEFAULT '',
version INTEGER NOT NULL DEFAULT 1,
created_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%SZ', 'now')),
updated_at TEXT NOT NULL DEFAULT (strftime('%Y-%m-%dT%H:%M:%SZ', 'now'))
);
CREATE INDEX idx_seed_lots_owner ON seed_lots (owner_id);
CREATE INDEX idx_seed_lots_plant ON seed_lots (plant_id);
-- Which lot a planting came out of. ON DELETE SET NULL rather than CASCADE:
-- retiring a lot must never destroy planting history — the plants really were in
-- the ground, whatever happened to the record of the packet.
ALTER TABLE plantings ADD COLUMN seed_lot_id INTEGER REFERENCES seed_lots (id) ON DELETE SET NULL;
CREATE INDEX idx_plantings_seed_lot ON plantings (seed_lot_id) WHERE seed_lot_id IS NOT NULL;
+10 -10
View File
@@ -12,13 +12,13 @@ import (
// plantingColumns lists plantings columns in the order scanPlanting expects.
// Used unqualified for direct selects; the /full read below qualifies with pl.
const plantingColumns = `id, object_id, plant_id, x_cm, y_cm, radius_cm, count, label,
planted_at, removed_at, version, created_at, updated_at`
planted_at, removed_at, seed_lot_id, version, created_at, updated_at`
func scanPlanting(s scanner) (*domain.Planting, error) {
var p domain.Planting
if err := s.Scan(
&p.ID, &p.ObjectID, &p.PlantID, &p.XCM, &p.YCM, &p.RadiusCM,
&p.Count, &p.Label, &p.PlantedAt, &p.RemovedAt,
&p.Count, &p.Label, &p.PlantedAt, &p.RemovedAt, &p.SeedLotID,
&p.Version, &p.CreatedAt, &p.UpdatedAt,
); err != nil {
return nil, err
@@ -89,12 +89,12 @@ func (d *DB) RestorePlanting(ctx context.Context, p *domain.Planting) (*domain.P
restored, err := scanPlanting(d.sql.QueryRowContext(ctx,
`INSERT INTO plantings
(id, object_id, plant_id, x_cm, y_cm, radius_cm, count, label, planted_at, removed_at,
version, created_at, updated_at)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?,
seed_lot_id, version, created_at, updated_at)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?,
strftime('%Y-%m-%dT%H:%M:%SZ', 'now'))
RETURNING `+plantingColumns,
p.ID, p.ObjectID, p.PlantID, p.XCM, p.YCM, p.RadiusCM, p.Count, p.Label,
p.PlantedAt, p.RemovedAt, p.Version, p.CreatedAt,
p.PlantedAt, p.RemovedAt, p.SeedLotID, p.Version, p.CreatedAt,
))
if err != nil {
return nil, fmt.Errorf("store: restore planting: %w", err)
@@ -153,14 +153,14 @@ func (d *DB) GetPlanting(ctx context.Context, id int64) (*domain.Planting, error
// — with plantingInsertArgs supplying its parameters — so all three stay in step
// with the table. removed_at is never set here: a new plop is active, and "clear
// bed" sets removed_at later via UpdatePlanting.
const plantingInsert = `INSERT INTO plantings (object_id, plant_id, x_cm, y_cm, radius_cm, count, label, planted_at)
VALUES (?, ?, ?, ?, ?, ?, ?, ?)
const plantingInsert = `INSERT INTO plantings (object_id, plant_id, x_cm, y_cm, radius_cm, count, label, planted_at, seed_lot_id)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
RETURNING ` + plantingColumns
// plantingInsertArgs builds plantingInsert's parameters, parenting p to objectID
// (p's own object normally, and the copied object when copying a garden).
func plantingInsertArgs(objectID int64, p *domain.Planting) []any {
return []any{objectID, p.PlantID, p.XCM, p.YCM, p.RadiusCM, p.Count, p.Label, p.PlantedAt}
return []any{objectID, p.PlantID, p.XCM, p.YCM, p.RadiusCM, p.Count, p.Label, p.PlantedAt, p.SeedLotID}
}
// CreatePlanting inserts a plop (fields already validated by the service) and
@@ -207,12 +207,12 @@ func (d *DB) UpdatePlanting(ctx context.Context, p *domain.Planting) (*domain.Pl
updated, err := scanPlanting(d.sql.QueryRowContext(ctx,
`UPDATE plantings
SET plant_id = ?, x_cm = ?, y_cm = ?, radius_cm = ?, count = ?, label = ?,
planted_at = ?, removed_at = ?,
planted_at = ?, removed_at = ?, seed_lot_id = ?,
version = version + 1,
updated_at = strftime('%Y-%m-%dT%H:%M:%SZ', 'now')
WHERE id = ? AND version = ?
RETURNING `+plantingColumns,
p.PlantID, p.XCM, p.YCM, p.RadiusCM, p.Count, p.Label, p.PlantedAt, p.RemovedAt,
p.PlantID, p.XCM, p.YCM, p.RadiusCM, p.Count, p.Label, p.PlantedAt, p.RemovedAt, p.SeedLotID,
p.ID, p.Version,
))
if errors.Is(err, sql.ErrNoRows) {
+9 -7
View File
@@ -11,13 +11,13 @@ import (
// plantColumns lists plants columns in the order scanPlant expects.
const plantColumns = `id, owner_id, name, category, spacing_cm, color, icon,
days_to_maturity, notes, version, created_at, updated_at`
days_to_maturity, source_url, vendor, notes, version, created_at, updated_at`
func scanPlant(s scanner) (*domain.Plant, error) {
var p domain.Plant
if err := s.Scan(
&p.ID, &p.OwnerID, &p.Name, &p.Category, &p.SpacingCM, &p.Color, &p.Icon,
&p.DaysToMaturity, &p.Notes, &p.Version, &p.CreatedAt, &p.UpdatedAt,
&p.DaysToMaturity, &p.SourceURL, &p.Vendor, &p.Notes, &p.Version, &p.CreatedAt, &p.UpdatedAt,
); err != nil {
return nil, err
}
@@ -111,10 +111,11 @@ func (d *DB) GetPlant(ctx context.Context, id int64) (*domain.Plant, error) {
// migration only, never through this path.
func (d *DB) CreatePlant(ctx context.Context, p *domain.Plant) (*domain.Plant, error) {
created, err := scanPlant(d.sql.QueryRowContext(ctx,
`INSERT INTO plants (owner_id, name, category, spacing_cm, color, icon, days_to_maturity, notes)
VALUES (?, ?, ?, ?, ?, ?, ?, ?)
`INSERT INTO plants (owner_id, name, category, spacing_cm, color, icon, days_to_maturity, source_url, vendor, notes)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
RETURNING `+plantColumns,
p.OwnerID, p.Name, p.Category, p.SpacingCM, p.Color, p.Icon, p.DaysToMaturity, p.Notes,
p.OwnerID, p.Name, p.Category, p.SpacingCM, p.Color, p.Icon, p.DaysToMaturity,
p.SourceURL, p.Vendor, p.Notes,
))
if err != nil {
return nil, fmt.Errorf("store: insert plant: %w", err)
@@ -130,12 +131,13 @@ func (d *DB) UpdatePlant(ctx context.Context, p *domain.Plant) (*domain.Plant, e
updated, err := scanPlant(d.sql.QueryRowContext(ctx,
`UPDATE plants
SET name = ?, category = ?, spacing_cm = ?, color = ?, icon = ?,
days_to_maturity = ?, notes = ?,
days_to_maturity = ?, source_url = ?, vendor = ?, notes = ?,
version = version + 1,
updated_at = strftime('%Y-%m-%dT%H:%M:%SZ', 'now')
WHERE id = ? AND version = ?
RETURNING `+plantColumns,
p.Name, p.Category, p.SpacingCM, p.Color, p.Icon, p.DaysToMaturity, p.Notes,
p.Name, p.Category, p.SpacingCM, p.Color, p.Icon, p.DaysToMaturity,
p.SourceURL, p.Vendor, p.Notes,
p.ID, p.Version,
))
if errors.Is(err, sql.ErrNoRows) {
+184
View File
@@ -0,0 +1,184 @@
package store
import (
"context"
"database/sql"
"errors"
"fmt"
"gitea.stevedudenhoeffer.com/steve/pansy/internal/domain"
)
// seedLotColumns lists seed_lots columns in the order scanSeedLot expects.
const seedLotColumns = `id, owner_id, plant_id, vendor, source_url, sku, lot_code,
purchased_at, packed_for_year, quantity, unit, cost_cents, germination_pct, notes,
version, created_at, updated_at`
func scanSeedLot(s scanner) (*domain.SeedLot, error) {
var l domain.SeedLot
if err := s.Scan(
&l.ID, &l.OwnerID, &l.PlantID, &l.Vendor, &l.SourceURL, &l.SKU, &l.LotCode,
&l.PurchasedAt, &l.PackedForYear, &l.Quantity, &l.Unit, &l.CostCents,
&l.GerminationPct, &l.Notes, &l.Version, &l.CreatedAt, &l.UpdatedAt,
); err != nil {
return nil, err
}
return &l, nil
}
// CreateSeedLot inserts a lot (fields already validated by the service).
func (d *DB) CreateSeedLot(ctx context.Context, l *domain.SeedLot) (*domain.SeedLot, error) {
created, err := scanSeedLot(d.sql.QueryRowContext(ctx,
`INSERT INTO seed_lots
(owner_id, plant_id, vendor, source_url, sku, lot_code, purchased_at,
packed_for_year, quantity, unit, cost_cents, germination_pct, notes)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
RETURNING `+seedLotColumns,
l.OwnerID, l.PlantID, l.Vendor, l.SourceURL, l.SKU, l.LotCode, l.PurchasedAt,
l.PackedForYear, l.Quantity, l.Unit, l.CostCents, l.GerminationPct, l.Notes,
))
if err != nil {
return nil, fmt.Errorf("store: insert seed lot: %w", err)
}
return created, nil
}
// GetSeedLot returns a lot by id, or domain.ErrNotFound. Ownership is the
// service's check, not this one's.
func (d *DB) GetSeedLot(ctx context.Context, id int64) (*domain.SeedLot, error) {
l, err := scanSeedLot(d.sql.QueryRowContext(ctx,
`SELECT `+seedLotColumns+` FROM seed_lots WHERE id = ?`, id))
if errors.Is(err, sql.ErrNoRows) {
return nil, domain.ErrNotFound
}
if err != nil {
return nil, fmt.Errorf("store: get seed lot: %w", err)
}
return l, nil
}
// ListSeedLotsForOwner returns every lot the actor owns, newest purchase first
// (undated lots last), then by id. Optionally narrowed to one plant. Always a
// non-nil slice.
func (d *DB) ListSeedLotsForOwner(ctx context.Context, ownerID int64, plantID *int64) ([]domain.SeedLot, error) {
query := `SELECT ` + seedLotColumns + ` FROM seed_lots WHERE owner_id = ?`
args := []any{ownerID}
if plantID != nil {
query += ` AND plant_id = ?`
args = append(args, *plantID)
}
// NULL purchased_at sorts last: an undated lot is the least useful to see first.
query += ` ORDER BY purchased_at IS NULL, purchased_at DESC, id DESC`
rows, err := d.sql.QueryContext(ctx, query, args...)
if err != nil {
return nil, fmt.Errorf("store: list seed lots: %w", err)
}
defer rows.Close()
lots := []domain.SeedLot{}
for rows.Next() {
l, err := scanSeedLot(rows)
if err != nil {
return nil, fmt.Errorf("store: scan seed lot: %w", err)
}
lots = append(lots, *l)
}
if err := rows.Err(); err != nil {
return nil, fmt.Errorf("store: iterate seed lots: %w", err)
}
return lots, nil
}
// UpdateSeedLot applies a version-guarded update of every mutable column (the
// service merges partial patches onto the current row first). Same contract as
// the other mutable resources: the updated row, or (current, ErrVersionConflict).
// plant_id and owner_id are immutable — re-pointing a purchase at a different
// variety or user would rewrite history rather than correct it.
func (d *DB) UpdateSeedLot(ctx context.Context, l *domain.SeedLot) (*domain.SeedLot, error) {
updated, err := scanSeedLot(d.sql.QueryRowContext(ctx,
`UPDATE seed_lots
SET vendor = ?, source_url = ?, sku = ?, lot_code = ?, purchased_at = ?,
packed_for_year = ?, quantity = ?, unit = ?, cost_cents = ?,
germination_pct = ?, notes = ?,
version = version + 1,
updated_at = strftime('%Y-%m-%dT%H:%M:%SZ', 'now')
WHERE id = ? AND version = ?
RETURNING `+seedLotColumns,
l.Vendor, l.SourceURL, l.SKU, l.LotCode, l.PurchasedAt, l.PackedForYear,
l.Quantity, l.Unit, l.CostCents, l.GerminationPct, l.Notes,
l.ID, l.Version,
))
if errors.Is(err, sql.ErrNoRows) {
current, gerr := d.GetSeedLot(ctx, l.ID)
if gerr != nil {
return nil, gerr
}
return current, domain.ErrVersionConflict
}
if err != nil {
return nil, fmt.Errorf("store: update seed lot: %w", err)
}
return updated, nil
}
// DeleteSeedLot removes a lot. Plantings attributed to it survive with a NULL
// seed_lot_id (see the migration): the plants really were in the ground,
// whatever happened to the record of the packet.
func (d *DB) DeleteSeedLot(ctx context.Context, id int64) error {
res, err := d.sql.ExecContext(ctx, `DELETE FROM seed_lots WHERE id = ?`, id)
if err != nil {
return fmt.Errorf("store: delete seed lot: %w", err)
}
n, err := res.RowsAffected()
if err != nil {
return fmt.Errorf("store: seed lot delete rows: %w", err)
}
if n == 0 {
return domain.ErrNotFound
}
return nil
}
// SeedLotUse is one active planting attributed to a lot, carrying just enough to
// compute its effective plant count.
type SeedLotUse struct {
LotID int64
RadiusCM float64
Count *int
SpacingCM float64
}
// SeedLotUsage returns the ACTIVE plantings attributed to any of the owner's
// lots — the "used" half of a lot's remaining quantity.
//
// It returns the raw ingredients rather than a SUM, because the effective count
// of a planting is its explicit count when it has one and the derived
// π·r²/spacing² formula otherwise. That formula lives in the service; reproducing
// it in SQL would guarantee the two drift apart.
func (d *DB) SeedLotUsage(ctx context.Context, ownerID int64) ([]SeedLotUse, error) {
rows, err := d.sql.QueryContext(ctx,
`SELECT pl.seed_lot_id, pl.radius_cm, pl.count, p.spacing_cm
FROM plantings pl
JOIN seed_lots sl ON sl.id = pl.seed_lot_id
JOIN plants p ON p.id = pl.plant_id
WHERE sl.owner_id = ? AND pl.removed_at IS NULL`,
ownerID)
if err != nil {
return nil, fmt.Errorf("store: seed lot usage: %w", err)
}
defer rows.Close()
uses := []SeedLotUse{}
for rows.Next() {
var u SeedLotUse
if err := rows.Scan(&u.LotID, &u.RadiusCM, &u.Count, &u.SpacingCM); err != nil {
return nil, fmt.Errorf("store: scan seed lot usage: %w", err)
}
uses = append(uses, u)
}
if err := rows.Err(); err != nil {
return nil, fmt.Errorf("store: iterate seed lot usage: %w", err)
}
return uses, nil
}