Files
accounted/app/api/import/articles/execute/route.ts
T
Jakob Wennberg ec27228a8e style: remove em/en dashes repo-wide, add CLAUDE.md rule against them (#890)
Em dashes (—) and en dashes (–) had spread across comments, docs, tests,
and a few UI strings, reading as AI-generated boilerplate rather than
house style. Replaced each with punctuation matching its context: colon
for explanatory clauses, comma for asides, plain hyphen for numeric/legal
ranges (e.g. "21-23§"), "to"/"till" for date ranges, parentheses for
paired-dash asides. messages/en.json and messages/sv.json were fixed by
hand together to keep sv/en in sync.

Left untouched where the dash is the functional subject rather than
decorative punctuation: date-range-parser.ts's separator regex,
charset-repair.ts's CP1252 byte-mapping table (and its test), the SIE
encoding mojibake docs, generic-csv.ts's minus-sign normalizer, the
agent system-prompt files that already instruct against em dashes, and
a golden iXBRL test fixture compared byte-for-byte.

Also fixes two bugs surfaced along the way: an off-by-one in
ApiKeysPanel's scope-label split (a leftover from an earlier partial
pass), and a charset-repair test that had lost the literal en-dash it
exists to verify.

Regenerated the agent atom seed migration (skills:generate) since 27
SKILL.md files changed. Added a CLAUDE.md rule against em/en dashes,
with an explicit carve-out for the functional-dash cases above.

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-04 15:58:06 +02:00

234 lines
8.4 KiB
TypeScript

import { NextResponse } from 'next/server'
import { ensureInitialized } from '@/lib/init'
import { eventBus } from '@/lib/events'
import { validateBody } from '@/lib/api/validate'
import { ArticleImportExecuteSchema } from '@/lib/api/schemas'
import { fetchAllRows } from '@/lib/supabase/fetch-all'
import { withRouteContext } from '@/lib/api/with-route-context'
import { errorResponseFromCode } from '@/lib/errors/get-structured-error'
import { ensureArticleNumber } from '@/lib/articles/ensure-article-number'
import { checkRevenueAccount, type RevenueAccountStatus } from '@/lib/articles/validate-revenue-account'
import type { Article } from '@/types'
import type { ArticleImportExecuteResult } from '@/lib/import/articles/types'
ensureInitialized()
interface ExistingArticle {
id: string
name: string
article_number: string | null
}
function nameKey(value: string | null): string | null {
if (!value) return null
return value.trim().toLowerCase() || null
}
/**
* POST /api/import/articles/execute
*
* Imports validated article rows. Duplicates (matched by article number, then
* by name) are either updated (merge: only non-empty fields overwrite) or
* skipped based on `update_duplicates`. An optional BAS revenue-account override
* is kept only when it is an active class-3 account; unknown/inactive accounts
* are dropped (with a warning) rather than mutating the chart of accounts.
*/
export const POST = withRouteContext(
'register_import.articles.execute',
async (request, ctx) => {
const { user, supabase, companyId, log, requestId } = ctx
const result = await validateBody(request, ArticleImportExecuteSchema, {
log,
operation: 'register_import.articles.execute',
})
if (!result.success) return result.response
const { rows, update_duplicates } = result.data
const opLog = log.child({ rowCount: rows.length, updateDuplicates: update_duplicates })
if (rows.length === 0) {
return errorResponseFromCode('REG_IMPORT_NO_ROWS', opLog, { requestId })
}
try {
const existingRaw = await fetchAllRows(({ from, to }) =>
supabase
.from('articles')
.select('id, name, article_number')
.eq('company_id', companyId)
.range(from, to),
)
const existing = existingRaw as unknown as ExistingArticle[]
const byNumber = new Map<string, ExistingArticle>()
const byName = new Map<string, ExistingArticle>()
for (const a of existing) {
if (a.article_number) byNumber.set(a.article_number, a)
const nk = nameKey(a.name)
if (nk && !byName.has(nk)) byName.set(nk, a)
}
// Revenue-account validation is cached per distinct account so a large
// import doesn't re-query the chart for every row.
const accountStatusCache = new Map<string, RevenueAccountStatus>()
const droppedAccounts = new Set<string>()
const warnings: string[] = []
const resolveRevenueAccount = async (acc: string | null): Promise<string | null> => {
if (!acc) return null
let status = accountStatusCache.get(acc)
if (!status) {
status = await checkRevenueAccount(supabase, companyId!, acc)
accountStatusCache.set(acc, status)
}
if (status === 'ok') return acc
if (!droppedAccounts.has(acc)) {
droppedAccounts.add(acc)
warnings.push(
status === 'activatable'
? `Försäljningskonto ${acc} är inte aktiverat i kontoplanen. Artiklar importerades utan kontoöverstyrning.`
: `Försäljningskonto ${acc} är ogiltigt, ignorerades.`,
)
}
return null
}
const created: Article[] = []
const updated: Article[] = []
let skipped = 0
const errors: { row_index: number; name: string; reason: string }[] = []
for (const row of rows) {
const nk = nameKey(row.name)
const match =
(row.article_number ? byNumber.get(row.article_number) : undefined) ??
(nk ? byName.get(nk) : undefined) ??
null
const revenueAccount = await resolveRevenueAccount(row.revenue_account)
if (match) {
if (!update_duplicates) {
skipped++
continue
}
// Merge mode: overwrite only fields the file clearly carries a value
// for. type/unit/vat_rate carry parser defaults that can't be told
// apart from "absent", so they are left untouched to avoid clobbering.
const merged: Record<string, unknown> = {}
if (row.name) merged.name = row.name
if (row.name_en) merged.name_en = row.name_en
if (row.price_excl_vat > 0) merged.price_excl_vat = row.price_excl_vat
if (row.cost_price !== null) merged.cost_price = row.cost_price
if (row.ean) merged.ean = row.ean
if (row.housework_type) merged.housework_type = row.housework_type
if (row.notes) merged.notes = row.notes
if (revenueAccount) merged.revenue_account = revenueAccount
if (Object.keys(merged).length === 0) {
skipped++
continue
}
const { data, error } = await supabase
.from('articles')
.update(merged)
.eq('id', match.id)
.eq('company_id', companyId)
.select()
.single()
if (error) {
errors.push({ row_index: row.row_index, name: row.name, reason: error.message })
continue
}
if (data) updated.push(data as Article)
continue
}
// No match, create.
const { data, error } = await supabase
.from('articles')
.insert({
user_id: user.id,
company_id: companyId,
name: row.name,
name_en: row.name_en,
type: row.type,
unit: row.unit || 'st',
price_excl_vat: row.price_excl_vat,
vat_rate: row.vat_rate,
revenue_account: revenueAccount,
cost_price: row.cost_price,
ean: row.ean,
housework_type: row.housework_type,
notes: row.notes,
article_number: row.article_number,
})
.select()
.single()
if (error) {
// Unique violation on (company_id, article_number): treat as a soft
// skip (manual number collided with an existing or in-batch article).
if (error.code === '23505') {
skipped++
continue
}
errors.push({ row_index: row.row_index, name: row.name, reason: error.message })
continue
}
if (data) {
// Auto-number when the file didn't supply one. Non-fatal: an
// unnumbered article is still usable and can be numbered later.
if (!data.article_number) {
try {
data.article_number = await ensureArticleNumber(supabase, companyId!, data.id)
} catch (err) {
opLog.warn('article number assignment failed', err as Error, { articleId: data.id })
}
}
created.push(data as Article)
// Track newly inserted number + name so later rows in the same batch
// dedup against them too.
const newArticle = data as ExistingArticle
if (newArticle.article_number) byNumber.set(newArticle.article_number, newArticle)
const nk = nameKey(newArticle.name)
if (nk && !byName.has(nk)) byName.set(nk, newArticle)
}
}
// Emit events for downstream listeners (non-blocking).
for (const a of created) {
await eventBus.emit({
type: 'article.created',
payload: { article: a, companyId: companyId!, userId: user.id },
})
}
const response: ArticleImportExecuteResult = {
success: errors.length === 0,
created: created.length,
updated: updated.length,
skipped,
failed: errors.length,
errors,
warnings,
}
opLog.info('article import complete', response)
return NextResponse.json({ data: response })
} catch (err) {
opLog.error('article import execute failed', err as Error)
return errorResponseFromCode('REG_IMPORT_EXECUTE_FAILED', opLog, {
requestId,
details: { reason: err instanceof Error ? err.message : 'unknown' },
})
}
},
{ requireWrite: true },
)