PDF Viewer with Search and Text Extraction

Open a PDF and the viewer shows one page at a time with previous and next buttons, a box to jump to a page number and the left and right arrow keys. A thumbnail list on the side renders lazily, so only the thumbnails you scroll to are drawn. Zoom runs from 50 to 300 percent, with a fit-to-width option that suits phones and narrow windows.

Showing a single page keeps the viewer light. Rendering all 500 pages of a manual at once would use a lot of memory, so the viewer draws the current page and frees the memory again when you move on. Document info lists the page count, title, author, creator, producer, creation and modification dates, PDF version and file size, which is useful when you need to know which tool produced a file.

Type a word in the search box and the viewer scans the text of every page, reports how many matches it found, and highlights them on the current page. Use next and previous match to jump between hits, even across pages. Highlight positions are approximate for PDFs that use unusual fonts or heavily custom spacing, so treat the box as a pointer to the spot rather than a pixel-exact outline.

Extract text gives you the text of the whole document in page order, which you can copy or download as a .txt file. That is the way to get text out, because text selection directly on the rendered page is not available. Layout is lost in extraction: columns, tables and headers can come out in reading order that differs from what you see, so check the result before using it in a script.

The viewer uses Mozilla pdf.js, the same engine behind Firefox's built-in reader. If a PDF is password-protected, a password box appears; the password is used only by pdf.js on your device. A corrupt file or a file that is not a PDF gives a plain message instead of a blank page. Files up to about 100 MB are accepted, and the whole file is read into browser memory, so a very large file on a phone may be slow.

A scanned PDF is usually just pictures of pages. It has no text layer, so search finds nothing and extraction returns nothing, because the tool does not do OCR. Forms, annotations and embedded JavaScript are not interactive here. If search works in your desktop reader but not here, the file probably stores its text in a way that is hard to read; try extracting the text to see what is actually there.