SnowPro Specialty: Gen AI free question 9 / 10 · Domain 4: Snowflake Document Processing
An analyst needs only pages 3, 4 and 5 of a 40-page PDF, counting the cover as page 1. Which page_filter value makes AI_PARSE_DOCUMENT process exactly those pages?
- [{'start': 3, 'end': 5}]
- [{'start': 2, 'end': 4}]
- [{'start': 2, 'end': 5}]
- [{'start': 3, 'end': 6}]
Show the answer and explanation
Answer: C
Why this is correct
Under "Arguments", the reference says 'page_filter' is an array of ranges, each an object with start and end fields that specify the first (inclusive) and last (exclusive) page in the range, and that page indexes start at 0. Pages 3 to 5 have indexes 2, 3 and 4, so start is 2 and end is the index after the last wanted page, 5. The reference's own example, {'start': 0, 'end': 1}, specifies the first page.
Why the other options are wrong
"[{'start': 3, 'end': 5}]" … With indexes starting at 0, start 3 is the fourth page. This range covers the fourth and fifth pages and leaves out page 3.
"[{'start': 3, 'end': 6}]" … Start 3 is the fourth page, so this range covers pages 4, 5 and 6 of the document, one page later than needed.
"[{'start': 2, 'end': 4}]" … The end value is exclusive, so end 4 stops after index 3. This range covers pages 3 and 4 and misses page 5.
No separate page_split
The reference says specifying page_filter implies page_split, so there is no need to set page_split in the same options object. Filtering suits the case the user guide mentions, where you know which pages hold the information you need.
Go to the SnowPro Specialty: Gen AI practice app (10 free questions; purchase unlocks all questions)