The transaction-row parser in services/membership/sources.go used a
helper (`getVal`) that did `fmt.Sprint(row[i])` before passing to
`matching.FormatDate`. The Sheets API returns date-formatted cells
as `float64` (Sheets serial-day numbers); pre-stringifying defeated
`FormatDate`'s `case float64:` dispatch, so values like 46147 leaked
through unchanged as the string "46147" instead of being converted
to "2026-05-05".
Surfaced by `make parity` (M5.4) — every `transactions[].date` on
/api/adults and /api/juniors differed between Python and Go. Python
side passes the raw value through directly (`isinstance(val, (int,
float))` in scripts/match_payments.py format_date), so it was always
correct.
Added a `getRaw` helper that returns row[i] without stringifying;
only the date column needs it. Extended TestLoadTransactions with
a numeric-serial-day row to lock in the regression.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Previously both backends defaulted to `CacheDir=tmp` and used the
same cache keys (`attendance_regular`, `attendance_juniors`,
`payments_transactions`, `exceptions_dict`) but stored different
shapes: Python caches post-processed view-model tuples
(e.g. `(members, sorted_months)`), Go caches raw sheet rows.
Whichever backend wrote last poisoned the cache for the other,
producing `ValueError: too many values to unpack (expected 2,
got 68)` on Python's /adults after the Go side populated the
file with 68 raw CSV rows.
This breaks the M5.4 `make parity` workflow that requires both
backends running side-by-side.
Fix: change Go's default to `tmp/go` so the two cache trees
never overlap. `CACHE_DIR` env var override still works.
`os.MkdirAll` already handles creating the new subdirectory on
first write.
Recovery for users with poisoned `tmp/`: hit /flush-cache on
the Python side once after pulling, then restart the Go server.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Add internal/web/api package with Go structs for every /api/X route:
AdultsResponse, JuniorsResponse, PaymentsResponse, VersionResponse.
All fields carry explicit json: tags matching the Python view-model keys.
Key design choices:
- member_data / month_labels / raw_payments are nested objects (not
the pre-serialised JSON strings used in Jinja templates)
- Expected{Value int; Unknown bool} with custom MarshalJSON emits int
or the string "?" for junior single-attendance months
- RawTransaction covers the full 11-column payments sheet row
schemagen_test.go reflects all four response types via
github.com/invopop/jsonschema and golden-compares against committed
schemas in tests/fixtures/api-schema/. The JSONSchema() method on
Expected lives in the test file so the prod binary has no jsonschema
dependency.
Closes M5.1 in docs/plans/2026-05-03-2349-go-backend-rewrite-progress.md.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Fio's transparent account page now serves dates as DD.MM.YY (e.g.
07.05.26) rather than the previously expected 4-digit-year format.
Extends parseCzechDate to try all eight layout variants: padded and
non-padded, dot and slash separators, 4-digit and 2-digit years.
Go maps 2-digit year 00-68 → 2000-2068, so 26 → 2026.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Wires slog.SetDefault to honour LOG_LEVEL in all CLI commands and adds
debug logs on the Fio fetch path so a silent "fetched 0 transaction(s)"
can be diagnosed without code changes:
- fio.New: which client variant (api/transparent) was selected
- apiClient: GET URL (token redacted as ****), HTTP status, body bytes,
parsed transaction count
- transparentClient: GET URL, HTTP status, body bytes, plus parser
stats (raw rows from second table, kept, dropped_bad_date,
dropped_nonpositive_amount)
Also suppresses the --print-fio-table block when zero transactions were
fetched, so the bare header no longer prints under that condition.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Prints an aligned tabwriter table of every Fio transaction in the
look-back window, with a STATUS column showing NEW (would be appended)
or DUP (already in sheet). Only fires when --dry-run is also set, so
it can't affect real syncs. Refactors Sync ID computation into a single
pre-pass shared by both the table printer and the row builder.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
extractSecondTableRows tracked a boolean inTarget flag and exited on
the first </table> token while inside the target. Any nested <table>
(e.g. pagination markup in the real Fio page) would cause an early
return before reading any data rows, explaining the 0-transaction report.
Fixed by tracking targetDepth instead: depth increments on every <table>
inside the target and we only return when it reaches 0 again.
parseCzechDate also only tried zero-padded layouts ("02.01.2006").
The real Fio transparent page emits non-padded dates ("7.5.2026");
added "2.1.2006" and "2/1/2006" as the preferred layouts.
Also adds a dry-run diagnostic line ("fetched N transaction(s) from Fio")
so the fetch vs dedup split is visible without reading logs.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
When token is empty, New falls back to transparentClient with the
caller-supplied hc. main.go passes nil, so the first Do() call panicked.
Default to http.DefaultClient when hc is nil.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Mirror fuj infer's read-only mode: SyncOpts.DryRun skips WriteHeader,
AppendValues, and SortByDateColumn, printing one "Dry run: would …"
line per planned operation instead. ID-dedup still runs so the output
reflects exactly what the next real sync would write.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Add internal/services/membership package: AttendanceLoader,
TransactionLoader, ExceptionLoader interfaces + NewStubSources stub
(returns ErrIOPending until M4 lands real Sheets loaders).
FeesReport and ReconcileReport orchestrate domain/fees + domain/reconcile
and write fixed-width text reports matching Python calculate_fees.py and
match_payments.py print_report output. 13 unit tests cover all formatter
branches and orchestration wiring via fake loaders.
cmd/fuj/main.go: fees and reconcile subcommands now dispatch; sync/infer
retain the [M4] placeholder.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
New go/internal/domain/matching package porting three helpers from
scripts/match_payments.py:
- BuildNameVariants: normalized ASCII variants from a member name (nickname
in parens, last/first split, len<3 filtered); variants[0] is always the
full base name — MatchMembers relies on this invariant.
- MatchMembers: auto/review confidence matching with an exact-name
short-circuit pass that prevents nickname substrings (tov) from firing
inside longer surnames (ottova); common-surname filter for review tier.
- FormatDate: nil/empty/""/serial int/float64 (since 1899-12-30, fractional
days supported)/YYYY-MM-DD passthrough/garbage → never errors.
- InferTransactionDetails: composes BuildNameVariants+MatchMembers+
ParseMonthReferences; falls back to sender-only member match and
date-derived month when text carries no signal.
21 table-driven tests; all expected values verified against live Python
on 2026-05-06. go-build, go-test, go-lint all clean.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
SHA-256 dedup hash from sync_fio_to_sheets.py generate_sync_id.
Key subtlety: Python str(float) emits "500.0" for whole-valued floats
and switches to scientific notation at |f|>=1e16 or |f|<1e-4 —
replicated via formatAmount using 'f'/'e' format selection.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Port scripts/infer_payments.py parse_czk_amount to Go as
internal/domain/money.ParseCZK. Preserves the Czech-locale heuristic
(comma = decimal sep; 2+ dots = thousand seps; single dot = decimal)
and returns (float64, error) so callers can opt into Python's
silent-zero contract via v, _ := money.ParseCZK(s).
All expected values verified against live Python on 2026-05-06.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Ports calculate_fee and calculate_junior_fee from scripts/attendance.py
into a new go/internal/domain/fees package. Introduces the Expected type
(Value int, Unknown bool) for the junior "?" sentinel, keeping the Go
API strictly typed instead of mirroring Python's str|int return.
All 20 table-driven tests pass with -race; golangci-lint clean.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Three-pass regex parser matching python/czech_utils.py parse_month_references:
1. Numeric slash notation — "11+12/2025", "01/26"; 2-digit year → +2000
2. Dot notation — "12.2025" (4-digit year only)
3. Czech month names — range walk (listopad-leden wrap logic) then
standalone with m≥10 → defaultYear-1 heuristic; longest-match
alternation (sorted desc by name length) handles cervenec vs cerven
35 table-driven tests, all expected outputs verified against live Python
on 2026-05-05 before locking. Plan at
docs/plans/2026-05-05-2337-go-rewrite-m2-2-parse-month-references.md.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Adds internal/domain/czech.Normalize, the first pure-domain function in
the Go rewrite (M2 milestone). Matches Python czech_utils.normalize byte-
for-byte: NFKD decompose via golang.org/x/text/unicode/norm, drop Mn-
category combining marks (unicode.Mn, not IsMark, to match Python's
unicodedata.combining() semantics), then strings.ToLower.
Includes 13-case table-driven test; all inputs spot-checked against the
Python implementation before locking. Adds golang.org/x/text v0.36.0 as
first external dependency.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
