Desktop app guide

A tour of the LEXO desktop app: the pages strip, preview, text pane, and Tune panel, plus editing, running OCR, proofreading, exporting, and keyboard shortcuts.

Start the app with:

lexo gui

The desktop app follows one flow: open → tune → run → proofread → export. It drives the same engine as the CLI, so results match.

LEXO welcome screen with buttons to import a PDF or image, select an image set, or batch OCR PDFs. — The welcome screen. Its How it works list is the whole workflow in four lines.

Opening documents

Import PDF or Image (Ctrl+O) opens a PDF or a single image.
Select Image Set (Ctrl+Shift+O) opens several images as one document, one page per image. PNG, JPEG, TIFF, and BMP are supported.
Batch OCR PDFs (Ctrl+Alt+O) runs OCR across many PDFs without opening them. See Batch OCR.

Opening a file copies it to a temporary working copy. Every edit applies to that copy. Your original file is untouched until you choose File → Save or Save As.

The window

LEXO main window: pages strip on the left, page preview, text pane, and Tune panel on the right. — Pages strip, preview, text pane, and the Tune panel with its Run, Edit, Pages, and PDF tabs.

Pages strip (left): a thumbnail for every page. Badges show status: blue for text-layer pages, green for OCR’d pages, amber while working, red for failures. Select several thumbnails with Ctrl- or Shift-click.
Preview (centre): the current page. Scroll over it to move between pages; hold Ctrl and scroll to zoom, then drag to pan. Fit returns to the whole page.
Text pane: the page’s text, editable. Copy (Ctrl+Shift+C) copies it.
Tune panel (right): four tabs, described below.

Both side panels can be resized, closed, and reopened from the View menu.

Run tab

Choose a mode:

Text layer copies the text a digital PDF already contains. Instant, offline, no account.
Google Docs OCR reads scanned pages and images through your Google account. Needs the one-time setup.

Then click Run (Ctrl+R). Cancel (Esc) stops a run. The progress bar counts pages as they finish, and each page’s text appears as soon as it is ready.

Google Docs OCR running at 25 percent, with thumbnails marked done or pending. — A run in progress. Finished pages are badged in the strip and readable immediately.

If some pages still fail after automatic retries, Retry Failed Pages (Ctrl+Shift+R) re-runs only those.

If you aren’t signed in when you start OCR, LEXO offers to sign you in and then retries the run.

Export as sets the format (plain text, Markdown, or JSONL) and Export (Ctrl+E) saves the file. See Export formats.

Edit tab: rotate and crop

Apply to sets the scope for every edit on this tab: This page, All pages, or Selected pages.

Rotate: Left 90° or Right 90°.
Crop: switching to the Edit tab shows a crop box on the preview. Drag its edges, or draw a new box, to remove headers, page numbers, and margins. Margins also appear as percentages you can type. Click Apply crop.

A blue crop box over a scanned Burmese page, with top, bottom, left, and right margins in the Edit tab. — Cropping before OCR keeps running heads and page numbers out of your text.

Cropping a page that is already cropped insets from the current crop rather than starting over.

Pages tab

Run Pages limits a run to some pages: type all, odd, even, or a range such as 1-3,7,10-, or use the quick buttons. Matching thumbnails are selected as you type.
Extract selected to new file… (Ctrl+Shift+E) saves the selected pages as a new PDF.
Remove selected pages (Delete) drops them from the working copy.

Reorder pages with Move Up and Move Down (Ctrl+Shift+↑ / Ctrl+Shift+↓), or from the thumbnail right-click menu, which also has rotate, extract, and remove.

PDF tab: spreads, splitting, appending

Show split line puts a vertical line on the preview. Drag it to the gutter of a two-up scan, then click Split two-up at line. Each page becomes two, left half first.
Split into files… cuts the PDF into several files of N pages each.
Append PDF(s)… adds other PDFs to the end of this one.

A two-page spread with a draggable vertical split line in the preview. — Two-up spreads are split by dragging a line, not by typing a ratio.

Edits run in the background with a busy indicator, so the window stays responsive on long documents.

Keyboard shortcuts

Action	Shortcut
Open PDF or image	`Ctrl+O`
Open image set	`Ctrl+Shift+O`
Batch OCR PDFs	`Ctrl+Alt+O`
Save / Save As	`Ctrl+S` / `Ctrl+Shift+S`
Export text	`Ctrl+E`
Close document	`Ctrl+W`
Run / Cancel	`Ctrl+R` / `Esc`
Retry failed pages	`Ctrl+Shift+R`
Copy page text	`Ctrl+Shift+C`
Move pages up / down	`Ctrl+Shift+↑` / `Ctrl+Shift+↓`
Extract selected pages	`Ctrl+Shift+E`
Remove selected pages	`Delete`
About Lexo	`F1`

On macOS, Qt maps Ctrl in these shortcuts to ⌘ Command.

Housekeeping

Account shows whether you’re signed in, and has Sign in with Google, Sign out, and Copy Data Folder Path.
Help → Clean Up Temporary Files… removes leftover working copies from earlier sessions.
Help → Check for Updates… compares your version with PyPI.