Skip to content

Edit PDF not recognize text of a pdf when PDF have signed small image #3005

Open

Description

@aymn3787

Edit PDF not recognize text of a pdf when PDF have signed small images even i can select any text before choose edit text or image option

the images are interface with the text and overlay covert all page to one image and can't select any text

Also may update to recognize small image separately not to select all in one box

Image

Activity

  1. ticket-craft commented on Oct 17, 2026

    @ticket-craft

    Thanks for the report! To look into this, we need a sample file that shows the problem and the app version. Please add them in a comment here, or open a new ticket with the details.

    Thanks! This ticket describes several separate problems:

    • Text not editable when small images overlay the page
    • Request to detect small images separately instead of one box

    It helps a lot if each one has its own ticket, so they can be tracked and fixed separately.

    Possibly related: #1022 and #2203.

  2. aymn3787 commented on Oct 17, 2026

    @aymn3787
    Author

    @Ticket-Craft

    Unfortunately i cant share the file however problems are exactly as you summarize in your comment

    Below image illustration that one small image overlay all table and text and it is move with one box, note that i hide table data

    my app version is 0.8.0

    Image
  3. mtlcl410 commented on Oct 17, 2026

    @mtlcl410
    Contributor

    I could reproduce this without your file, using synthetic PDFs: a numeric table plus a small signature image with transparency, built in a few different ways. One of them gives exactly what your screenshot shows.

    When the table and the signature are drawn inside one Form XObject (a reusable block of page content), Edit a PDF treats the whole block as a single picture:

    • One box, with a resize handle, covers the table and the signature together.
    • None of the table's text is offered for editing (0 paragraphs), although it can still be selected and copied outside edit mode, as in your first screenshot.

    HTML-to-PDF converters (Chrome's print to PDF, for example) write a block of a page this way when something in it has transparency, and a signature image with a transparent background does. Your table is right-to-left, with a hand-drawn signature, which fits that kind of source.

    This is the known limitation tracked in #2964: text inside a Form XObject can't be edited yet, and the form is moved as one piece.

    Two related cases I also checked, which give a "too big" box even when the text is editable:

    • The signature alone inside a page-sized form (how stamping tools add a layer).
    • A signature PNG that is page-wide but mostly transparent.

    In both, the box and the click target cover the whole form or image, not just the visible signature. So clicking on text under it picks the image instead.

    No need to share the file. If you can, please say which program produced the PDF (Document Properties ▸ Description shows "PDF Producer"). That would confirm which of these it is.

  4. mtlcl410 commented on Oct 17, 2026

    @mtlcl410
    Contributor

    #3067 adds Ungroup for this case. In Edit text & images, right-click the big box and choose Ungroup. The table and signature are then drawn as ordinary page content, so each cell's text can be edited and the signature is its own picture. Undo puts the box back. The page looks exactly the same before and after.

    One caveat until #2155 lands: right-clicking artwork in edit mode can open a second menu on top of the artwork menu (#2637). If Ungroup is hidden behind it, move the pointer away and right-click again.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions