[kvlonge]

Why

DuckShop's edge-case area is meant to exercise the Snowflake constructs real projects' tests rely on. Several common functions and syntax forms were still missing from it:

• text and encoding functions;
• bit and floor arithmetic;
• null-safe comparisons;
• comma joins after explicit joins;
• array edits;
• clock text conversions;
• XML reads.

This adds them with expectations captured on Snowflake.

Changes

• int_edge_sku_text_reads:
  • STARTSWITH, ENDSWITH, SUBSTRING, POSITION (with IN and with a comma), CHARINDEX, CHR, REGEXP_SUBSTR_ALL;
  • ::STRING and CAST ... AS STRING;
  • hex and Base64 functions;
  • a comment attached to a text literal.
• int_edge_amount_bit_reads:
  • FLOOR, BITAND, ZEROIFNULL, ::BIGINT, ::DOUBLE;
  • IS NOT DISTINCT FROM and EQUAL_NULL;
  • a comma join after an explicit join;
  • HASH_AGG read in two row orders.
• int_edge_basket_array_edits: ARRAY_APPEND, ARRAY_INTERSECTION, ARRAY_SLICE, TO_ARRAY and bracket array literals.
• int_edge_clock_text_reads:
  • DATE, TO_TIME, TRY_TO_TIMESTAMP;
  • TO_TIMESTAMP_LTZ read back at UTC;
  • CURRENT_DATE and SYSDATE, used only for facts that hold on any day.
• int_edge_basket_xml_reads: PARSE_XML, XMLGET and CHECK_XML.
• One unit test per model, plus a README update.

Verification

• sqb compile reports 0 errors and 0 warnings with the edge area. The core-copy checksum is unchanged.
• sqb format --check is clean.
• The new tests pass on Snowflake under the edge target. Two capture runs in a row produced identical files.
• pytest benchmarks/duckshop/tests passes, and ruff is clean.
• make check-ci passes.
• The public-repository hygiene scan passes.