zfin/src/data/staleness.zig
Emil Lerch 53b807d12e upgrade to zig 0.16.0
IO-as-an-interface refactor across the codebase. The big shifts:
- std.io → std.Io, std.fs → std.Io.Dir/File, std.process.Child → spawn/run.
- Juicy Main: pub fn main(init: std.process.Init) gives gpa, io, arena,
  environ_map up front. main.zig + the build/ scripts use it directly.
- Threading io through everywhere that touches the outside world (HTTP,
  files, stderr, sleep, terminal detection). Functions taking `io` now
  announce side effects at the call site — the smell is the feature.
- date math takes `as_of: Date`, not `today: Date`. Caller resolves
  `--as-of` flag vs wall-clock at the boundary; the function operates
  on whatever date it's given. Every "today" parameter renamed and
  the as_of: ?Date + today: Date pattern collapsed.
- now_s: i64 (or before_s/after_s pairs) for sub-second metadata
  fields like snapshot captured_at, audit cadence, formatAge/fmtTimeAgo.
  Also pure and testable.
- legitimate Timestamp.now callers (cache TTL math, FetchResult
  timestamps, rate limiter, per-frame TUI "now" captures) gain
  `// wall-clock required: ...` comments justifying the read.

Test discovery: replaced the local refAllDeclsRecursive with bare
std.testing.refAllDecls(@This()). Sema-pulling main.zig's top-level
decls reaches every test file transitively through the import graph;
no explicit _ = @import(...) lines needed.

Cleanup along the way:
- Dropped DataService.allocator()/io() accessor methods; renamed the
  fields to drop the base_ prefix. Callers use self.allocator and
  self.io directly.
- Dropped now-vestigial io parameters from buildSnapshot,
  analyzePortfolio, compareSchwabSummary, compareAccounts,
  buildPortfolioData, divs.display, quote.display, parsePortfolioOpts,
  aggregateLiveStocks, renderEarningsLines, capitalGainsIndicator,
  aggregateDripLots, printLotRow, portfolio.display, printSnapNote.
- Dropped the unused contributions.computeAttribution date-form
  wrapper (only computeAttributionSpec is called).
- formatAge/fmtTimeAgo take (before_s, after_s) instead of io and
  reading the clock internally.
- parseProjectionsConfig uses an internal stack-buffer
  FixedBufferAllocator instead of an allocator parameter.
- ThreadSafeAllocator wrappers in cache concurrency tests dropped
  (0.16's DebugAllocator is thread-safe by default).
- analyzePortfolio bug surfaced by the rename: snapshot.zig was
  passing wall-clock today instead of as_of, mis-valuing cash/CDs
  for historical backfills.

83 new unit tests added due to removal of IO, bringing coverage from 58%
-> 64%
2026-05-09 22:40:33 -07:00

299 lines
10 KiB
Zig

//! Staleness nags for hand-maintained data sources.
//!
//! Some datasets in this codebase require periodic human maintenance
//! (e.g. appending the latest year's T-bill rate, replacing Shiller's
//! `ie_data.csv`). Each such source declares a `pub const <name>_last_updated: Date`
//! constant which a human bumps when the refresh happens.
//!
//! This module registers those sources and prints a warning to stderr
//! on every `zfin` invocation once the annual refresh window opens
//! and the data is still stale. The refresh window is expressed as a
//! single `(month, day)` per year — the earliest date by which fresh
//! upstream data is expected to be available — not as a rolling
//! day-count.
//!
//! ### Nagging semantics
//!
//! Given today's date and an entry's `(due_month, due_day, last_updated)`:
//!
//! 1. Compute `this_years_due = Date.fromYmd(today.year(), due_month, due_day)`.
//! 2. If `today < this_years_due` — not yet nag season, silent.
//! 3. If `last_updated >= this_years_due` — already refreshed this cycle, silent.
//! 4. Otherwise — nag.
//!
//! The nag keeps firing every invocation until the human bumps
//! `last_updated` past `this_years_due`.
//!
//! ### Adding a new source
//!
//! 1. Add `pub const <name>_last_updated: Date = Date.fromYmd(YYYY, MM, DD);`
//! to the module that owns the data.
//! 2. Append a new `StaleEntry` to `entries` below.
//! 3. Document the refresh procedure in the owning module's doc
//! comment (or a `TODO` at the top).
const std = @import("std");
const Date = @import("../models/date.zig").Date;
const risk = @import("../analytics/risk.zig");
const shiller = @import("shiller.zig");
/// A hand-maintained data source that nags once a year if it hasn't
/// been refreshed by its annual `(due_month, due_day)`.
pub const StaleEntry = struct {
/// Human-readable name surfaced in the warning.
name: []const u8,
/// The last time a human refreshed this data. Bumped by hand.
last_updated: Date,
/// Calendar month (1-12) by which fresh upstream data is expected.
due_month: u8,
/// Calendar day (1-31) within `due_month`.
due_day: u8,
/// Source file to point the user at for refresh instructions.
source_file: []const u8,
};
/// Registry of hand-maintained data sources in this codebase. Passed
/// to `check` from `main.zig`; tests construct their own slices.
pub const entries = [_]StaleEntry{
.{
.name = "T-bill risk-free rate table",
.last_updated = risk.tbill_rates_last_updated,
.due_month = 1,
.due_day = 31,
.source_file = "src/analytics/risk.zig",
},
.{
.name = "Shiller annual returns (ie_data.csv)",
.last_updated = shiller.ie_data_last_updated,
.due_month = 4,
.due_day = 1,
.source_file = "src/data/shiller.zig",
},
};
/// Write a warning line for each entry in `entries` that is overdue
/// for refresh as of `as_of`. Writes nothing when every entry is
/// fresh. Entries are processed in order; multiple warnings are
/// separated by a blank line.
pub fn check(
writer: *std.Io.Writer,
as_of: Date,
list: []const StaleEntry,
) !void {
var wrote_any = false;
for (list) |entry| {
const this_years_due = Date.fromYmd(
as_of.year(),
entry.due_month,
entry.due_day,
);
// Not yet nag season.
if (as_of.lessThan(this_years_due)) continue;
// Already refreshed this cycle.
if (!entry.last_updated.lessThan(this_years_due)) continue;
if (wrote_any) try writer.writeAll("\n");
wrote_any = true;
var due_buf: [10]u8 = undefined;
var upd_buf: [10]u8 = undefined;
const due_str = this_years_due.format(&due_buf);
const upd_str = entry.last_updated.format(&upd_buf);
try writer.print(
"warning: {s} is overdue for refresh.\n" ++
" Expected by {s}; last updated {s}.\n" ++
" See {s} for refresh instructions.\n",
.{ entry.name, due_str, upd_str, entry.source_file },
);
}
}
// ── Tests ────────────────────────────────────────────────────
test "silent before due date" {
var buf: [1024]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{.{
.name = "Test data",
.last_updated = Date.fromYmd(2025, 3, 15),
.due_month = 4,
.due_day = 1,
.source_file = "src/test.zig",
}};
try check(&w, Date.fromYmd(2026, 3, 31), &list);
try std.testing.expectEqualStrings("", w.buffered());
}
test "nag on due date when stale" {
var buf: [1024]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{.{
.name = "Test data",
.last_updated = Date.fromYmd(2025, 3, 15),
.due_month = 4,
.due_day = 1,
.source_file = "src/test.zig",
}};
try check(&w, Date.fromYmd(2026, 4, 1), &list);
const expected =
"warning: Test data is overdue for refresh.\n" ++
" Expected by 2026-04-01; last updated 2025-03-15.\n" ++
" See src/test.zig for refresh instructions.\n";
try std.testing.expectEqualStrings(expected, w.buffered());
}
test "nag after due date when stale" {
var buf: [1024]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{.{
.name = "Test data",
.last_updated = Date.fromYmd(2025, 3, 15),
.due_month = 4,
.due_day = 1,
.source_file = "src/test.zig",
}};
try check(&w, Date.fromYmd(2026, 6, 15), &list);
const expected =
"warning: Test data is overdue for refresh.\n" ++
" Expected by 2026-04-01; last updated 2025-03-15.\n" ++
" See src/test.zig for refresh instructions.\n";
try std.testing.expectEqualStrings(expected, w.buffered());
}
test "silent when already refreshed this cycle" {
var buf: [1024]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{.{
.name = "Test data",
.last_updated = Date.fromYmd(2026, 4, 2),
.due_month = 4,
.due_day = 1,
.source_file = "src/test.zig",
}};
try check(&w, Date.fromYmd(2026, 6, 15), &list);
try std.testing.expectEqualStrings("", w.buffered());
}
test "silent when last_updated equals this year's due date" {
var buf: [1024]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{.{
.name = "Test data",
.last_updated = Date.fromYmd(2026, 4, 1),
.due_month = 4,
.due_day = 1,
.source_file = "src/test.zig",
}};
try check(&w, Date.fromYmd(2026, 4, 1), &list);
try std.testing.expectEqualStrings("", w.buffered());
}
test "two-cycle stale still fires" {
var buf: [1024]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{.{
.name = "Test data",
.last_updated = Date.fromYmd(2024, 2, 1),
.due_month = 4,
.due_day = 1,
.source_file = "src/test.zig",
}};
try check(&w, Date.fromYmd(2026, 6, 15), &list);
const expected =
"warning: Test data is overdue for refresh.\n" ++
" Expected by 2026-04-01; last updated 2024-02-01.\n" ++
" See src/test.zig for refresh instructions.\n";
try std.testing.expectEqualStrings(expected, w.buffered());
}
test "empty list is a no-op" {
var buf: [16]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
try check(&w, Date.fromYmd(2030, 12, 31), &[_]StaleEntry{});
try std.testing.expectEqualStrings("", w.buffered());
}
test "multiple stale entries separated by blank line" {
var buf: [2048]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{
.{
.name = "Alpha",
.last_updated = Date.fromYmd(2024, 12, 1),
.due_month = 1,
.due_day = 31,
.source_file = "src/alpha.zig",
},
.{
.name = "Beta",
.last_updated = Date.fromYmd(2025, 3, 15),
.due_month = 4,
.due_day = 1,
.source_file = "src/beta.zig",
},
};
try check(&w, Date.fromYmd(2026, 6, 15), &list);
const expected =
"warning: Alpha is overdue for refresh.\n" ++
" Expected by 2026-01-31; last updated 2024-12-01.\n" ++
" See src/alpha.zig for refresh instructions.\n" ++
"\n" ++
"warning: Beta is overdue for refresh.\n" ++
" Expected by 2026-04-01; last updated 2025-03-15.\n" ++
" See src/beta.zig for refresh instructions.\n";
try std.testing.expectEqualStrings(expected, w.buffered());
}
test "fresh entry does not emit blank line before later stale entry" {
var buf: [2048]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{
.{
.name = "Alpha",
.last_updated = Date.fromYmd(2026, 2, 1), // fresh
.due_month = 1,
.due_day = 31,
.source_file = "src/alpha.zig",
},
.{
.name = "Beta",
.last_updated = Date.fromYmd(2025, 3, 15), // stale
.due_month = 4,
.due_day = 1,
.source_file = "src/beta.zig",
},
};
try check(&w, Date.fromYmd(2026, 6, 15), &list);
const expected =
"warning: Beta is overdue for refresh.\n" ++
" Expected by 2026-04-01; last updated 2025-03-15.\n" ++
" See src/beta.zig for refresh instructions.\n";
try std.testing.expectEqualStrings(expected, w.buffered());
}
test "silent when today is one day before due" {
var buf: [256]u8 = undefined;
var w = std.Io.Writer.fixed(&buf);
const list = [_]StaleEntry{.{
.name = "Test data",
.last_updated = Date.fromYmd(2020, 1, 1),
.due_month = 4,
.due_day = 1,
.source_file = "src/test.zig",
}};
try check(&w, Date.fromYmd(2026, 3, 31), &list);
try std.testing.expectEqualStrings("", w.buffered());
}
test "real registry compiles and is non-empty" {
// Guard that the registry stays wired up; doesn't assert any
// particular nag behavior (real dates drift over time).
try std.testing.expect(entries.len >= 2);
for (entries) |e| {
try std.testing.expect(e.name.len > 0);
try std.testing.expect(e.source_file.len > 0);
try std.testing.expect(e.due_month >= 1 and e.due_month <= 12);
try std.testing.expect(e.due_day >= 1 and e.due_day <= 31);
}
}