The practical answer
The current PDF tool selects physical positions starting at 1, removes duplicate selections and outputs them in ascending original-file order. Input order is not a reordering command, and printed labels are not parsed. Reopen the downloaded PDF to verify its count, order and actual content.
Map physical positions to visible page content
Open the original and count from its first physical page. Covers, contents pages, blanks and appendices all occupy positions. Roman numerals and chapter labels printed on a page are content, not direct instructions for the extraction field.
The following six-page specimen is an original editorial exercise, not a real contract or archive. Give every page a clear heading so its identity survives the handoff. The underlying PDF-LIB page-copying API supports copying pages; deduplication and sorting are additional behavior of this KitNelo interface.
| Physical page | Printed label | Content |
|---|---|---|
| 1 | Cover | Cover |
| 2 | i | Contents |
| 3 | 1 | First body section |
| 4 | 2 | Second body section |
| 5 | A-1 | First appendix page |
| 6 | A-2 | Second appendix page |
Which four pages does 6,2,2,4-5 produce?
In the PDF tool, select Extract PDF pages, load the specimen and enter 6,2,2,4-5. It selects physical pages 2, 4, 5 and 6. Repeating 2 does not create an extra copy. The output should show i, 2, A-1 and A-2, in that order, across four pages.
The original’s fourth physical page displays “2”, but you must enter 4 to select it. Page 2 of the new document is another numbering system: it now contains original physical page 4. A handoff note such as “original 4 → new 2” prevents subsequent comments from mixing the two systems.
- Preserve the original and make a page map using clear headings or a small specimen.
- Choose extraction mode, load one PDF and enter numeric ranges separated by ordinary English commas.
- Process and download a separate file; a success message is not a reason to delete the original.
- Reopen the download and verify all four labels, page count and orientation.
| New file page | Original physical page | Visible label |
|---|---|---|
| 1 | 2 | i |
| 2 | 4 | 2 |
| 3 | 5 | A-1 |
| 4 | 6 | A-2 |
A range selects a set; it does not play pages backward
The current parser also expands 5-4 into 4 and 5, then sorts the final selection. That is not reverse-order export. If you require 6, 2, 4, 5 as a new sequence, use a workflow that genuinely supports reordering instead of rearranging tokens in this field.
Use ordinary commas. Zero, out-of-range numbers, a printed label such as i, and unrecognized punctuation fail. An empty selection cannot produce a zero-page output either. Verify the syntax with a small specimen before processing a large file.
Trying to put the appendix first
Enter 6,2 and assume that a result beginning with 2 has been randomly shuffled.
Recognize that this interface extracts in original physical order; choose a reordering workflow if a new sequence is required.
After counting, inspect the pages and handoff context
Compare the new PDF with the original side by side, checking text, images, orientation and page boundaries. For mixed page sizes, consult the A4 and Letter PDF checks. A viewer fitting everything to its window does not establish equal physical sizes.
Page numbers and watermarks are separate delivery choices; see the PDF review-copy numbering guide. If new labels such as “1 / 4” are needed, first approve the extraction, then prepare the review copy while retaining the old-to-new map.
Acceptance record for the four-page copy
- The output contains four pages, rather than a count inferred from input tokens.
- Visible labels are i, 2, A-1 and A-2 in sequence.
- Original and new physical positions are mapped.
- Necessary text, images, orientation and boundaries remain intact.
- The file name and handoff note distinguish the source from its extracted copy.
Visible-page extraction is not a redaction guarantee
Page selection does not prove that all hidden data, attachments, annotations or sensitive content have been securely removed. This exercise also makes no promise that bookmarks, interactive forms and other document-level structures survive unchanged. Verify those requirements separately.
Processing is local to the browser. Complex, damaged, encrypted or device-intensive files may fail. Preserve the original and actual error, and isolate the issue with a readable small specimen. Do not record delivery as complete without reopening the downloaded result.
Try 4,2,2: sorted positions with duplicates removed
Use the four-page mapping-four.pdf: 1=COVER, 2=CONTENTS, 3=BODY-1, 4=BODY-2. This specimen tests selection syntax independently of the earlier six-page illustration.
- Import it and enter 4,2,2 using ordinary commas in the extraction field.
- Download, reopen and record both page markers and the physical count.
- Run a separate comparison with 4-2: endpoints are normalized to positions 2 through 4, not played backward.
Reference result and recovery decision
4,2,2 should yield CONTENTS then BODY-2: physical positions 2 and 4, with only one copy of 2. The range 4-2 should yield CONTENTS, BODY-1, BODY-2.
If the requirement is BODY-2 first with two copies of the contents page, this selection field does not fulfill it. Extract, then use a page editor supporting reordering and duplication, and verify the entire resulting sequence.
The try 4,2,2: sorted positions with duplicates removed exercise is an authored teaching case. Retain source, output and review decisions separately; the shared ZIP is preparation material, not a verified user outcome.
Common questions
Why not enter 1 when the body starts at page 1?
The tool’s 1 is the first physical page. A cover and contents page can place printed body page 1 at physical position 3. Build the map first.
Can I recover unselected pages from the extracted file?
Return to the preserved original and extract again. Do not rely on the subset to restore pages it excluded. Keep source and handoff copies separately.
What if I need two copies of one page?
Repeated tokens are deduplicated. Use a workflow with an explicit page-duplication operation and verify the resulting count.
Further reading & sources
Put the guide into practice
Open the relevant tool, preview the result and check your exported file.
