Add money: statement-driven personal finance tracker

A data directory holds one folder per account. Statements dropped into
those folders are parsed into a rebuildable SQLite index, categorised by
ordered glob rules in rules.toml, and browsed or hand-tagged in a Bubble
Tea TUI. Movements between the user's own accounts are marked as
transfers by the same rules and excluded from spending totals.

Manual tags and transfer marks are stored separately from the rule-derived
ones and always win, so editing rules.toml and re-running retag never
destroys hand edits.

Parsers are pluggable. Three are ported from the Python extractors they
replace -- nlb and traderepublic read PDFs via pdftotext -layout, revolut
reads the CSV export -- alongside a configurable-column CSV parser and a
cmd parser that shells out to an external script.

Both ports fix two latent bugs in the originals: the sign character class
rejected the typographic minus U+2212 that some PDF fonts emit, and NLB's
hardcoded continuation indent broke when pdftotext compressed runs of
spaces, so the threshold is now measured from the description column.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
2026-08-09 00:37:20 +02:00
co-authored by Claude Opus 5
commit b0026c5a79
30 changed files with 5048 additions and 0 deletions
+221
View File
@@ -0,0 +1,221 @@
package importer
import (
"os"
"path/filepath"
"strings"
"testing"
"git.petrovv.com/nikola/money/internal/config"
"git.petrovv.com/nikola/money/internal/parser"
"git.petrovv.com/nikola/money/internal/rules"
"git.petrovv.com/nikola/money/internal/store"
)
const accountTOML = `
name = "Checking"
currency = "EUR"
parser = "csv"
[csv]
skip_rows = 1
date = { col = 0, layout = "2006-01-02" }
description = { col = 1 }
amount = { col = 2 }
`
// newRoot builds a data root with one account and the given statement files.
func newRoot(t *testing.T, statements map[string]string) (string, *store.DB, []*config.Account, *rules.Engine) {
t.Helper()
root := t.TempDir()
dir := filepath.Join(root, "checking")
if err := os.MkdirAll(dir, 0o755); err != nil {
t.Fatal(err)
}
write(t, filepath.Join(dir, config.AccountFile), accountTOML)
for name, body := range statements {
write(t, filepath.Join(dir, name), body)
}
db, err := store.Open(config.IndexPath(root))
if err != nil {
t.Fatal(err)
}
t.Cleanup(func() { db.Close() })
accounts, err := config.LoadAccounts(root)
if err != nil {
t.Fatal(err)
}
engine := rules.New(&config.Rules{Rule: []config.Rule{{Match: "*LIDL*", Tag: "groceries"}}})
return root, db, accounts, engine
}
func write(t *testing.T, path, body string) {
t.Helper()
if err := os.WriteFile(path, []byte(body), 0o644); err != nil {
t.Fatal(err)
}
}
func mustRun(t *testing.T, root string, db *store.DB, accounts []*config.Account, e *rules.Engine, opts Options) Result {
t.Helper()
res, err := Run(root, db, accounts, e, opts)
if err != nil {
t.Fatal(err)
}
for _, f := range res.Errs() {
t.Fatalf("import of %s failed: %v", f.Path, f.Err)
}
return res
}
// Identical lines inside one statement are distinct transactions; the same
// line seen again in an overlapping statement is not.
func TestDedupe(t *testing.T) {
root, db, accounts, engine := newRoot(t, map[string]string{
"2026-01.csv": `date,description,amount
2026-01-06,LIDL SOFIA,-45.20
2026-01-06,LIDL SOFIA,-45.20
2026-01-10,RENT,-800.00
`,
})
res := mustRun(t, root, db, accounts, engine, Options{})
if _, added, _ := res.Total(); added != 3 {
t.Fatalf("first import added %d rows, want 3 (identical same-day lines must both survive)", added)
}
// Unchanged file: skipped without even parsing.
res = mustRun(t, root, db, accounts, engine, Options{})
if parsed, added, _ := res.Total(); parsed != 0 || added != 0 {
t.Errorf("re-import parsed %d and added %d, want 0 and 0", parsed, added)
}
// Same file, parsed again: every row is recognised as a duplicate.
res = mustRun(t, root, db, accounts, engine, Options{Force: true})
if _, added, skipped := res.Total(); added != 0 || skipped != 3 {
t.Errorf("forced re-import added %d, skipped %d; want 0 and 3", added, skipped)
}
// An overlapping statement contributes only its genuinely new rows.
write(t, filepath.Join(root, "checking", "2026-02.csv"), `date,description,amount
2026-01-10,RENT,-800.00
2026-02-10,RENT,-800.00
`)
accounts, err := config.LoadAccounts(root)
if err != nil {
t.Fatal(err)
}
res = mustRun(t, root, db, accounts, engine, Options{})
if _, added, skipped := res.Total(); added != 1 || skipped != 1 {
t.Errorf("overlapping import added %d, skipped %d; want 1 and 1", added, skipped)
}
txns, err := db.Transactions(store.Filter{})
if err != nil {
t.Fatal(err)
}
if len(txns) != 4 {
t.Errorf("index holds %d transactions, want 4", len(txns))
}
}
// Rules are applied as rows are inserted, so a fresh import is already tagged.
func TestImportAppliesRules(t *testing.T) {
root, db, accounts, engine := newRoot(t, map[string]string{
"2026-01.csv": `date,description,amount
2026-01-06,LIDL SOFIA,-45.20
2026-01-10,RENT,-800.00
`,
})
res := mustRun(t, root, db, accounts, engine, Options{})
if res.Retagged != 0 {
t.Errorf("retag changed %d rows after import, want 0: rules should already be applied", res.Retagged)
}
untagged, err := db.Transactions(store.Filter{Untagged: true})
if err != nil {
t.Fatal(err)
}
if len(untagged) != 1 || untagged[0].Description != "RENT" {
t.Errorf("untagged = %+v, want only RENT", untagged)
}
}
// A broken statement must be reported without aborting the rest of the run.
func TestBadFileIsReportedNotFatal(t *testing.T) {
root, db, accounts, engine := newRoot(t, map[string]string{
"good.csv": `date,description,amount
2026-01-06,LIDL SOFIA,-45.20
`,
"bad.csv": `date,description,amount
not-a-date,BROKEN,-1.00
`,
})
res, err := Run(root, db, accounts, engine, Options{})
if err != nil {
t.Fatalf("Run returned a fatal error, want a per-file report: %v", err)
}
failures := res.Errs()
if len(failures) != 1 || filepath.Base(failures[0].Path) != "bad.csv" {
t.Fatalf("failures = %+v, want exactly bad.csv", failures)
}
if _, added, _ := res.Total(); added != 1 {
t.Errorf("added %d rows, want 1 from good.csv", added)
}
}
func TestCheckBalances(t *testing.T) {
balance := func(v int64) *int64 { return &v }
good := []parser.RawTxn{
{Date: "2026-01-01", Description: "A", AmountMinor: -1000, BalanceMinor: balance(9000)},
{Date: "2026-01-02", Description: "B", AmountMinor: -500, BalanceMinor: balance(8500)},
{Date: "2026-01-03", Description: "C", AmountMinor: 2000, BalanceMinor: balance(10500)},
}
if w := checkBalances(good, 2); len(w) != 0 {
t.Errorf("a consistent chain produced warnings: %v", w)
}
// A missed row shows up as a break at the row after it.
broken := []parser.RawTxn{
{Date: "2026-01-01", Description: "A", AmountMinor: -1000, BalanceMinor: balance(9000)},
{Date: "2026-01-02", Description: "B", AmountMinor: -500, BalanceMinor: balance(7000)},
}
w := checkBalances(broken, 2)
if len(w) != 1 {
t.Fatalf("got %d warnings, want 1: %v", len(w), w)
}
if !strings.Contains(w[0], "70.00") || !strings.Contains(w[0], "85.00") {
t.Errorf("warning = %q, want both the reported and the expected balance", w[0])
}
// Statements that report no balances are not checked.
none := []parser.RawTxn{
{Date: "2026-01-01", Description: "A", AmountMinor: -1000},
{Date: "2026-01-02", Description: "B", AmountMinor: -500},
}
if w := checkBalances(none, 2); len(w) != 0 {
t.Errorf("statements without balances produced warnings: %v", w)
}
}
// Only files matching include globs are treated as statements.
func TestIncludeGlobs(t *testing.T) {
root, db, accounts, engine := newRoot(t, map[string]string{
"2026-01.csv": `date,description,amount
2026-01-06,LIDL SOFIA,-45.20
`,
"notes.txt": "not a statement",
})
accounts[0].Include = []string{"*.csv"}
res := mustRun(t, root, db, accounts, engine, Options{})
if len(res.Files) != 1 {
t.Fatalf("processed %d files, want 1: %+v", len(res.Files), res.Files)
}
if _, added, _ := res.Total(); added != 1 {
t.Errorf("added %d rows, want 1", added)
}
}