Missing features

Missing features

Abstract

“Missing” in PYNE is a moving frontier. As of 2026-09-14 (hoox-pyne 0.6.4), core v6 support remains very high (~99%+ for core language), with post-launch TV items closed (multiline strings, library export const, footprint mocks, risk enforcement, matrix LA, compile-mode strategy, drawing GC, …) plus open-source corpus parse/Runtime hardening, incremental bar-mode TA (including 0.3.10–0.3.12 kernels), alert engine + L2 webhooks, series caps, warm-compile, pending-fill averaging, timeframe.change (intraday buckets + 0.6.0 D/W/M exchange calendar), package Runtime SoT (pynescript.runtime; backend.runtime is a shim), compile object-mode residual recovery (UDF locals, nopython None/unicode, matrix handles, UDT returns, x = switch, drawing copy, int(na), round() extras, import stubs, color type dispatch, nested UDT methods), type-aware intraday risk enforcement on both brokers, intrabar var rollback on realtime ticks, request.security barmerge on HTF resample paths, and linter legacy-rule retirements. Remaining gaps cluster into optional set0x community-corpus plot parity (P1p official builtins closed), by-design platform differences, and editor-only features. Incremental ta.nvi/ta.pvi/aroon/dpo/donchian/kst and Supertrend mid±factor·ATR goldens have landed.

Authoritative narrative: docs/missing_features.md. Canonical IDs: docs/ROADMAP.md (2026-08-17).

Conceptual model

Rendering…

Interface surface

Recently completed (do not re-open as “missing”)

Summarized — full prose in source doc:

Still incomplete / by design

Post-v6 launch themes (historical gap list)

v6 launch items (dynamic requests, strict bool, enums, polylines, log.*, negative indices, truediv, …) are largely supported or intentionally stubbed where platform-bound.

2025–2026 monthly TV updates introduced footprint types, active inputs, multiline strings, export const, UDT sort fields — tracked and largely closed in this repo’s July 2026 work.

Corpus + performance snapshot (2026-08-09)

Open-source set01–04 (2477 scripts; measured locally, not shipped in-repo):

Not TradingView® platform parity. OK% excludes only path-listed intentional demos (never soft-suppress unknown failures).

Plans: .opencode/plans/2026-07-28-runtime-performance.md. Skill: .grok/skills/pynescript-perf/. Golden: tests/test_ta_incremental.py, tests/test_series_cap.py.

Internals

Invariants & edge cases

1. Closing a syntax gap requires grammar + builder + unparser + tests, not only an evaluator stub.
2. Mock data can green tests while production feeds differ — declare data source explicitly.
3. Compile does not fill foreign request.security data — multi-asset plots stay na on compile unless same-symbol simple OHLCV; do not treat invent-as-close as a fix.
4. Auto Fib needs pivots — without swing structure (or ZigZag library registration), expect structured insufficient-pivot errors, not fib lines.
5. “0 missing vs reference list” ≠ identical semantics for every edge case.
6. Perf changes must keep bar-by-bar semantics — no whole-script vectorization; prefer golden tests vs current oracle.
7. Update this annex when missing_features.md changes major status lines.

Worked examples

Detect multiline string support

from pynescript.ast.helper import parse, unparse
src = '''//@version=6
indicator("m")
s = """a
b"""
plot(1)
'''
tree = parse(src)
assert '"""' in unparse(tree) or "'''" in unparse(tree)

Footprint mock call

Prefer tests in tests/test_v6_features.py as executable specification for request.footprint methods.

Incremental TA parity smoke

.venv/bin/python -m pytest tests/test_ta_incremental.py -q
# Disable hot path if needed:
PYNE_TA_INCREMENTAL=0 python -c "from pynescript.runtime import Runtime; ..."

Failure modes

See also

• Implementation status
• Pine v6 surface
• Roadmap
• Compatibility
• Technical analysis (ta.*)
• Alerts
• Interpret ↔ compile parity