Copying visible text does not guarantee that the hidden destination travels with it. Clipboard-aware applications can provide several representations at once, and the receiving application chooses which representation to read.
Plain text contains only characters
If a cell displays “Project brief” but links to a web address, a plain-text copy may contain only those two words. There is no URL for an extractor to recover. Exporting to CSV and pasting through a plain-text editor commonly causes this loss.
HTML carries anchor destinations
An HTML clipboard representation can include an anchor element with an href destination. Both the visible label and destination survive. Browser-based document and spreadsheet editors often provide HTML when you copy directly from their editing view.
Rich text varies by application
Desktop editors may use RTF or an application-specific clipboard format. A browser tool cannot always read every proprietary representation. If the receiving page shows labels but detects no links, try copying from the application's web version or exporting to HTML instead.
PDFs are unpredictable
A PDF may contain real link annotations, visible URL text, or simply styled text that looks clickable. Selection and copy behavior depends on the viewer. For a small document, test one known hyperlink before processing the full file.
A diagnostic sequence
- Copy one item whose destination you already know.
- Paste into the tool and check whether it remains clickable.
- If not, avoid a plain-text intermediary.
- Try the source application's browser version.
- For visible raw URLs, use plain-text URL detection.
- For hidden links that still disappear, export the source to HTML or use its native link-export feature.
The important distinction is whether the URL exists in the copied data at all. No extraction algorithm can reconstruct a hidden destination after the source application has reduced it to plain text.