How do I check the drawing index against the sheet title blocks with AI before issuing a drawing set?

Read this drawing set [upload the PDF] and build two tables. Index: one row per line of the drawing index on the cover sheet with sheet number, sheet title and current revision exactly as written. TitleBlocks: one row per PDF page with page number, sheet number, sheet title, latest revision number and revision date, and issue date, read from the title block only. Copy values exactly, write NOT SHOWN instead of guessing, and give a confidence word (clear, likely, unsure) for anything hard to read. Do not fix or reconcile anything.

EXAMPLE DATA ONLY. Here are my Index and TitleBlocks tabs [paste the example rows] and the set issue date I typed: [type yours]. Write Excel formulas (or a step-by-step Copilot plan) that match the tabs on sheet number after trimming spaces and ignoring case, and flag: on the index but not in the set, in the set but not on the index, duplicate sheet numbers in the set, title mismatch, revision mismatch, and an issue date different from my set date. Use only my data.

Index tab (example, read from the cover sheet)
Sheet | Title | Rev
A-001 | Cover Sheet and Drawing Index | 2
A-101 | First Floor Plan | 3
A-102 | Second Floor Plan | 3
A-201 | Exterior Elevations | 2
A-301 | Building Sections | 1
A-501 | Wall Sections and Details | 2

TitleBlocks tab (example, one row per PDF page)
Page | Sheet | Title | Rev | Issue date
1 | A-001 | Cover Sheet and Drawing Index | 2 | 2026-10-30
2 | A-101 | First Floor Plan | 3 | 2026-10-30
3 | A-102 | Second Floor Plan | 2 | 2026-10-30
4 | A-201 | North and South Elevations | 2 | 2026-10-30
5 | A-401 | Enlarged Toilet Room Plans | 1 | 2026-10-30
6 | A-501 | Wall Sections and Details | 2 | 2026-10-16
7 | A-501 | Wall Sections and Details | 2 | 2026-10-30

Set issue date (I typed): 2026-10-30

EXAMPLE DATA ONLY (not from a real project). Index tab: A-001 Cover Sheet and Drawing Index rev 2; A-101 First Floor Plan rev 3; A-102 Second Floor Plan rev 3; A-201 Exterior Elevations rev 2; A-301 Building Sections rev 1; A-501 Wall Sections and Details rev 2. TitleBlocks tab: seven PDF pages with sheet number, title, revision and issue date. Set issue date I typed: 2026-10-30. Write Excel formulas (or a step-by-step Copilot plan) that match the tabs on sheet number and flag: on the index but not in the set, in the set but not on the index, duplicate sheet numbers, title mismatch, revision mismatch, and an issue date different from my set date. Use only my data; do not decide which revision is correct.

Expected results on this example (output of the script below on these rows)
A-102 REV MISMATCH: index 3 vs title block 2 (page 3) → likely printed from the old revision; re-export the sheet
A-201 TITLE MISMATCH: index 'Exterior Elevations' vs title block 'North and South Elevations' → update one so they agree
A-301 ON INDEX, NOT IN SET → dropped from the print set, or removed and still listed
A-401 IN SET (page 5), NOT ON INDEX → added late; add it to the index or pull it
A-501 DUPLICATE IN SET, pages 6, 7 → leftover re-print
A-501 DATE MISMATCH: page 6 dated 2026-10-16, set is 2026-10-30 → page 6 is the stale copy

Suggested Excel logic (write against your own sheet layout)
- Normalize sheet numbers in a helper column: UPPER(TRIM([@Sheet])).
- On index, not in set: COUNTIF(TitleBlocks[Key], [@Key])=0 on the Index tab.
- In set, not on index: COUNTIF(Index[Key], [@Key])=0 on the TitleBlocks tab.
- Duplicate in set: COUNTIF(TitleBlocks[Key], [@Key])>1.
- Title and revision: XLOOKUP([@Key], Index[Key], Index[Title]) and Index[Rev], compared with EXACT after TRIM and UPPER.
- Date: [@IssueDate]<>SetDate.

Same check as a short Python script (our own code; reads two CSVs you export)
import csv
from collections import defaultdict
SET_DATE = "2026-10-30"   # type the issue date of this set
norm = lambda s: " ".join(s.split()).upper()
index = {norm(r["Sheet"]): r for r in csv.DictReader(open("index.csv"))}
pages = defaultdict(list)
for r in csv.DictReader(open("titleblocks.csv")):
    pages[norm(r["Sheet"])].append(r)
for sheet in sorted(set(index) | set(pages)):
    i, p = index.get(sheet), pages.get(sheet, [])
    if not p:
        print(sheet, "ON INDEX, NOT IN SET"); continue
    if not i:
        print(sheet, "IN SET (page", p[0]["Page"] + "), NOT ON INDEX"); continue
    if len(p) > 1:
        print(sheet, "DUPLICATE IN SET, pages", ", ".join(r["Page"] for r in p))
    for r in p:
        if norm(r["Title"]) != norm(i["Title"]):
            print(sheet, f"TITLE MISMATCH: index '{i['Title']}' vs title block '{r['Title']}'")
        if r["Rev"].strip() != i["Rev"].strip():
            print(sheet, f"REV MISMATCH: index {i['Rev']} vs title block {r['Rev']} (page {r['Page']})")
        if r["IssueDate"].strip() != SET_DATE:
            print(sheet, f"DATE MISMATCH: page {r['Page']} dated {r['IssueDate']}, set is {SET_DATE}")
Do not decide which revision is current — only flag for human review.