“See page 4” is a fragile research instruction. In a selected-page packet, page 4 might be page 2 of a supplement, printed as “S-1.” None of those numbers tells you whether a share count is a year-end balance or an average over the year. A useful note keeps these meanings apart.
Write a mapping, not a new citation
Take the synthetic recipe A:1-3,B:2,A:5. The fourth output page comes from B’s second physical PDF page. If B prints that page as S-1, record all three labels when they matter:
Reading copy position: 4 Source: supplement.pdf, physical page 2 Printed label: S-1 (if shown) Source version: synthetic supplement, v1 Financial period: not supplied by this fixture
Reordering does not create a new source or a new financial period. Keep the original filename and version alongside the convenience copy. A saved recipe can reproduce the order; it cannot recover a missing original or prove which assertion the page supports.
Label a financial input on its own terms
Now consider a real research note with two fields: “shares at period end” and “weighted-average shares for EPS.” They may legitimately differ. Moving the supporting page to the front of a packet does not make either number the right input for every calculation.
Share Slice’s method uses period-end scenario shares and fixed income. If a note compares its output with reported diluted EPS, write down the difference in definitions before interpreting a discrepancy. Never infer a financial period from the page’s place in the packet.
Use each check for one job
- Use Page Recipe to preview and export physical page order, then visually inspect the new file.
- Use Paper Markdown Check to catch missing local resources and supported structural mistakes in your explanatory note.
- Read the original document to establish period, units and what a passage actually says. No clean report performs that reading for you.
If you annotate the resulting reading copy in the related KomoPDF editor, call it an annotated copy and retain the original. Page copying and annotation are not evidence that an original digital signature remains valid. A careful workflow leaves the reader enough labels to find the same place and enough assumptions to question the same calculation.