PDF Page Manager
Reorder or extract PDF pages and rotate or split them locally in your browser with typed ranges, page-flow checks, and a source-to-output manifest.{{ summaryTitle }}
{{ summaryLine }}
{{ operation_mode === 'split_packets' ? 'Packet archive' : 'Managed PDF' }}
The chart renderer is unavailable. The same page route remains available in the manifest.
| Packet | Output page | Source file | Source page | Selector | Action | Copy |
|---|---|---|---|---|---|---|
| {{ row.packet }} | {{ row.outputPage }} | {{ row.sourceName }} | {{ row.sourcePage }} | {{ row.selector }} | {{ row.action }} |
Page work becomes difficult when visible numbering and actual PDF order diverge. A report may show Roman numerals in its front matter, restart numbering in an appendix, or display a printed “12” on the viewer's fourteenth page. Reordering and splitting operate on the PDF page positions, so the viewer sequence must be checked before writing a page plan.
Several common jobs share the same underlying operation: copy selected source pages into a new sequence. Reordering changes their order; extraction leaves unwanted pages out; duplication copies the same source page more than once; rotation changes a copied page's viewing angle; splitting sends selected pages into several new PDFs.
- Arrange defines one exact reading order, including intentional repeats.
- Extract keeps a chosen subset and requires a check for missing notices, exhibits, or cross-references.
- Split defines packet boundaries as well as page membership.
- Rotate adds a turn to selected copied pages without changing their source content.
Copying pages creates a new document rather than editing the originals. That is useful for preserving source files, but it can leave document-level features behind. Bookmarks, forms, annotations, accessibility structure, attachments, scripts, page labels, and signatures require separate review. Any content change can also affect the validity or permitted-edit state of a digitally signed PDF.
Multiple input files form one combined source sequence in file order. Moving a file changes every combined page number after it. Freeze that source order before finalizing selectors, then keep a manifest that maps each output position back to its source file and source page.
How to Use This Tool:
Build the page route first, then load up to 12 authorized PDFs in the order they should be numbered.
- Choose local files and reorder the source list before editing selectors. Combined page 1 starts in the first file and numbering continues through each later file.
- Select Arrange pages into one PDF or Split pages into a packet ZIP. For splitting, choose custom ranges, one PDF per page, or a fixed number of pages per PDF.
- Enter the Page order or packet selectors. Use
all,reverse,odd,even, a page number, or a range such as4-9,-6, or12-. - Choose whether repeated pages keep typed order or become ascending and unique. Set rotation only after confirming whether it targets arranged output positions or combined source positions.
- Check the planned page count, omissions, duplicates, page flow, and manifest. Correct an out-of-range selector or raise the page guard deliberately if the build button remains unavailable.
- Build the PDF or ZIP locally, download it, and inspect the first, boundary, rotated, repeated, and last pages in a separate viewer.
Interpreting Results:
Page flow summarizes how many pages from each source are copied normally, copied with added rotation, or omitted. Page manifest is the exact audit trail: packet, output position, source file, source page, selector, and action.
Duplicate references are not automatically errors because repeated divider sheets and covers can be intentional. An omission count is also not automatically wrong in an extraction. Compare both against the intended route, and treat a successful build as file creation evidence rather than proof that the reading order or document semantics are correct.
Technical Details:
All selectors use one-based combined page positions. If the first source has 8 pages and the second has 6, combined pages 1-8 belong to the first file and 9-14 belong to the second. A selected position resolves to one source index and one source page before any output is built.
Transformation Core
| Selector | Resolved sequence | Notes |
|---|---|---|
all | Every combined page from first to last | Ascending order |
reverse | Every combined page from last to first | Preserved only in typed-order mode |
odd or even | Positions of the chosen parity | Parity uses combined positions |
7 | Combined page 7 | Must exist |
3-8 | 3, 4, 5, 6, 7, 8 | Both endpoints included |
8-3 | 8, 7, 6, 5, 4, 3 | Reverse ranges are valid |
-5 or 10- | First through 5, or 10 through last | Open endpoint uses the document boundary |
Arrange mode concatenates the comma-separated selector results into one route. Typed order preserves repeats and reverse ranges. Ascending unique removes duplicate positions and sorts the resolved route numerically.
Split mode applies one of three packet rules. Custom ranges makes each comma-separated selector one packet. Every page makes a one-page packet for every combined source position. Fixed interval groups consecutive positions into packets of 1 through 100 pages, with a shorter final packet when necessary.
Rotation rule
Rotation accepts 0, 90, 180, or 270 degrees. In arrange mode, the rotation selector addresses resolved output positions after repeats and reordering. In split mode, it addresses combined source positions before packet creation. A blank rotation selector means every selected page. The requested turn is added to the copied page's existing rotation modulo 360 degrees.
Guards and output construction
The page guard accepts 1 through 1,000 planned copies and defaults to 500. The per-file guard accepts 10 through 250 MB and defaults to 150 MB; files over the chosen guard are ignored before parsing. A route above the page guard is rejected rather than truncated.
Each selected page is copied into a newly created PDF. Arrange mode writes one PDF. Split mode writes one PDF per packet and places those files in a compressed ZIP. Only the new document title is chosen from the output name or first source name; other source metadata is not deliberately copied as a complete document record.
Privacy and Accuracy Notes:
PDF bytes stay in browser memory and are not uploaded. The route settings can be shared separately, but file bytes, page geometry, and generated output are excluded from shared state, chart data, manifests, and clipboard text.
- Encrypted, malformed, or very large PDFs may fail to load or exhaust available browser memory.
- The process does not repair, OCR, redact, compress, or validate source PDFs.
- Keep the originals and use unsigned copies when document signatures or certification matter.
- Review forms, links, bookmarks, tags, annotations, attachments, and accessibility after page copying.
Worked Examples:
Move an appendix ahead of a section
For two ordered files with 8 and 6 pages, the plan 1-4, 9-12, 5-8 creates a 12-page PDF: the first four pages of the first source, four pages from the second source, then the remaining four pages of the first. The manifest is the fastest way to confirm the source switch at output pages 5 and 9.
Build packets of four pages
With 14 combined pages, fixed-interval splitting at 4 pages produces packets covering 1-4, 5-8, 9-12, and 13-14. The last packet is valid with two pages; it is not padded or discarded.
References:
- PDFDocument API, PDF-LIB.
- PDF-UX: Page Labels, PDF Association.
- Permissions and limitations of signed PDFs, Adobe, December 2025.