Tier 17.4: F3.2 sub-tier 4 — fact_option_order_events columns (30 cols)

[jamezycesar-collab]

Summary

Fourth sub-tier of WS2 F3.2 follow-on. Adds 30 spec-mapping columns to fact_option_order_events across all four dialects.

Files changed

• ddl/option/02_option_gold_delta.sql - 30 columns added
• ddl/option/04_option_gold_hive.sql - 30 columns added (Hive line comments to avoid validator COMMENT terminator)
• ddl/option/05_option_silver_fabric_lakehouse.sql - 30 columns added
• ddl/option/06_option_fabric_warehouse.sql - 30 columns added
• guardrails/known_field_mapping_gaps.csv - 81 → 51 rows
• CHANGELOG.md - Tier 17.4 entry

30 columns added (PDF-verified)

Group	Columns	Spec section
Order key + ts	order_key_date, event_timestamp	§5.1.1 rows 6, 9
Manual flags	manual_flag, manual_order_key_date, manual_order_id, electronic_dup_flag, electronic_timestamp	§5.1.1 rows 10-14
Order metadata	dept_type, min_qty, solicitation_flag, rfq_id, representative_ind, exch_origin_code	§5.1.1 rows 15, 19, 28, 30, 32 + §5.1.3 row 28
Party / account	firm_designated_id, account_holder_type, affiliate_flag	§5.1.1 rows 24-26
Routing	sender_imid, receiver_imid, sender_type, originating_imid, destination, destination_type, routed_order_id, session, route_rejected_flag, multi_leg_ind, paired_order_id	§5.1.3 rows 9, 14-18, 27, 30, 34 + §5.1.4 rows 15, 17
Modify / parent	prior_order_key_date, parent_order_key_date, request_timestamp	§5.1.6.1 row 9, §5.1.7 rows 9, 33

No mapping CSV fixes

All 30 columns passed PDF verification with OK status (verifier output in TIER_17_VERIFICATION.csv). The mapping CSV's section refs are already correct.

Validator quirk found and worked around

The validator's _CREATE_TABLE_RE regex includes COMMENT in its terminator alternation. When the Hive variant initially used per-column COMMENT 'section ...' documentation, the regex misinterpreted the first column-level COMMENT appearing after a DECIMAL(38, 18)) close-paren as the table-level terminator — silently truncating column discovery and triggering false parity errors.

Workaround in this PR: Hive uses -- section ... line comments instead of inline COMMENT 'string'. Delta / Fabric Lakehouse / Fabric Warehouse variants use -- end-of-line comments (their conventional style anyway).

Followup (tooling debt): harden the validator's regex to require column-list close-paren before considering COMMENT as a body terminator. Tracking outside this PR.

Coverage

All 8 guardrails:                  pass (exit 0)
known_field_mapping_gaps.csv:    51 (was 81; -30)
  F3.1 phantom-table rows:        0 (unchanged - cleared in Tier 13-16)
  F3.2 missing-column rows:      51 (was 81)
DDL files in parity scope:        22 (unchanged)
Tables in 2+ dialects:            42 (unchanged - same tables, new columns)
New parity violations:             0

F3.2 burndown progress

Sub-tier	Host table	Cols	Allowlist	Status
Tier 17.1	fact_option_executions	3	92 → 89	✅
Tier 17.2	fact_option_allocations	3	89 → 86	✅
Tier 17.3	fact_multileg_option_legs	5	86 → 81	✅
Tier 17.4 (this)	fact_option_order_events	30	81 → 51	✅
Tier 17.5	fact_multileg_option_events	51	51 → 0	next (final)

Test plan

• All 8 guardrails pass (exit 0)
• No new parity violations
• All 30 fact_option_order_events mapping rows now resolve in DDL
• PDF verification: all 30 cols match mapping CSV section + row
• Recurrence-safe precheck in push script