Extract Links from PDF
See where every link in a PDF goes, without visiting a single one.
What was found
Links
| No. | Page | Type | Link textLink | Address | Status |
|---|
No links match.
Change the PDF
Saves a changed copy. The original file is not touched, and the links are not opened.
About the Extract Links from PDF
Choose a PDF and get a table of every link in it: the page, what kind of link it is (a web address, an email address, a phone number, a jump to another page, a file), the words that carry the link and the address it goes to. Addresses that are only written out in the text and that nothing links to are listed too, marked “not clickable”.
Filter by kind, show each address once, export the table as CSV, or copy the addresses. You can also save a copy of the PDF in which those written-out addresses are clickable, or in which the links are removed. The tool only reads addresses: it never opens, fetches or checks a link, and the file never leaves your device.
How to use it
- Choose the PDF or drag it onto the box. A password-protected file asks for its password first.
- Read the table. Use Show to pick a kind of link, the search box to find an address, Only addresses that are not clickable to see what is only typed out, and One row per address to merge repeats.
- Press Download CSV for the table, or Copy addresses for a plain list.
- To change the PDF, choose Make written-out addresses clickable or one of the Remove options under Change the PDF, then press Save changed PDF.
Examples
Page 3 · Web address · “our documentation” → https://example.com/docs · Clickable Page 3 · Web address · www.example.org/report · Not clickable Page 5 · Page in this PDF · “back to contents” → Page 1 Page 7 · Web address · “https://good.example.com” → https://evil.example.net/login · The text shows good.example.com but the link goes to evil.example.net.
The last line is how a misleading link looks: its text names one site and it leads to another.
No.,Page,Type,Clickable,Link text,Address,Note 1,3,Web address,Yes,our documentation,https://example.com/docs, 2,3,Web address,No,www.example.org/report,www.example.org/report,
Common uses
- Checking that every link in a brochure, report or ebook goes where it should before you publish it.
- Finding web addresses that look like links but are not clickable, and fixing them in one step.
- Spotting links whose text shows one site and whose address goes to another, a sign of a misleading or phishing link, in a file you were sent.
- Making the list of sources a paper or report cites, to open in a link checker or reference manager.
- Removing all links, or only those that leave the document, before you share a file.
What counts as a link
A clickable link in a PDF is a link annotation (ISO 32000-1 §12.5.6.5) laid over the words, with an action: open a web address, start an email, jump to a page or a named place, open another file, or run a viewer command such as “next page”. The table lists each one with the text found under it (the start of it, when a link covers a whole block of text).
Links to pages in this PDF show their page number; one whose page does not exist is marked broken. Links that run a script or submit a form cannot be shown as addresses; they are listed as Script or form action. Open the file in PDF Sanitizer to see what they do. Addresses that use an unusual scheme (anything other than http, https, ftp, mailto or tel) are flagged, because an ordinary web address is what people expect.
A link is marked Text differs when its words are themselves an address of one site and the link goes to another. Brackets and full stops around the address are ignored, and an address that wraps over two lines (stored as one link per line) is judged as a whole.
Addresses that are only written out
The tool reads the page’s text from the exact position of every character, finds web addresses (starting with http://, https://, ftp:// or www.) and email addresses (also when written with mailto: in front), and checks whether a link covers them. Other kinds, such as a file: address, are left alone, and no link is ever made for one. One that nothing links to is not clickable. Addresses cut by a line break are joined up when the break comes after a “/”, “-”, “=”, “&”, “?” or “#”, or before one of those or a dot, and such a row says so. A sentence that ends with an address does not swallow the next word.
Bare domains such as example.com are not looked for unless you ask, because file names like readme.md look the same. When a link is added, “www.” addresses and bare domains get https:// in front.
Changing the PDF
Make written-out addresses clickable adds an invisible link over each address that is not clickable yet (one per line of text for an address that wraps). Look at the rows marked “Written across two lines” first: typesetters sometimes hyphenate a long address at the end of a line, and a hyphen that only marks the break would become part of the link. Remove deletes link annotations; the words stay, and so does any blue colour or underline that was drawn into the page. “Remove links to the web, email, phone and files” keeps links to pages and viewer commands, so a table of contents still works. The rest of the file is kept as it is. Changing a digitally signed file makes its signature show as invalid.
Sources
- ISO 32000-1:2008, Document management — Portable document format — Part 1: PDF 1.7, §12.5.6.5, Link annotations, and §12.6.4, Action types: web (URI), page (GoTo), file (GoToR, Launch) and script (JavaScript) actions.
Limitations
- Only link annotations and written-out text are read. Links in bookmarks, inside form buttons or drawn as pictures are not listed.
- An address inside a scanned image has no text to read. Run PDF OCR first; the summary says which pages have no text.
- An address that a line break cuts in the middle of a word cannot be joined safely, so only the first part is found.
- The tool reads addresses and never opens them, so it cannot say whether a web address still works.
- Script and form actions are not shown as addresses. Saving a changed copy needs an unprotected PDF; a password-protected one must be unlocked with Unlock PDF first.
Privacy
Everything happens in your browser. What you enter or open here is not uploaded or stored by MySmartCoPilot.
Frequently asked questions
Does this tool open or check the links?
No. It only reads the addresses in the file and writes them out as text; nothing is fetched, opened or sent. The addresses are shown as plain text, not as clickable links, so you cannot open one by accident in a file you do not trust.
Why is an address listed as “not clickable”?
The PDF shows the address as ordinary text and no link covers it. Many PDF programs make such addresses clickable when you open the file, so it may look like a link on screen, but nothing in the file makes it one. Make written-out addresses clickable adds the links.
What does “The text shows … but the link goes to …” mean?
The words of the link are themselves an address, and the link leads to a different site. That is how a misleading link is built. There are harmless cases, such as a short address that redirects, so look at the target before you decide.
Does removing the links remove the blue underlined text?
No. Links are separate objects laid over the page. When they are removed the words stay as they are, including any colour or underline that was drawn into the page.
Is my PDF uploaded?
No. It is read and changed by your browser on your device. Nothing is sent to MySmartCoPilot or anyone else.