Skip to main content
The text object is an instance accessible from the main mupdfwv.MuPDFWebViewer instance as follows: 
const text = mupdf.text;
This assumes you have returned your instance name as mupdf from the initMuPDFWebViewer() promise!

​
text

The text object has the following methods: search(config: { keyword: string; caseSensitive?: boolean; useRegex?: boolean; pageRange?: string }) Searches for text.
  • Arguments:
    • config – The configuration object.
Config object:
  • Arguments:
    • keyword – (required) Search keyword.
    • caseSensitive – (optional) Whether to be case sensitive.
    • useRegex – (optional) Whether to use regular expressions.
    • pageRange – (optional) Page index range to search (e.g., “1-5, 7, 9-12”, “all”). If not set will search all pages.
  • Returns: Promise<{ results: { words: { prefix: string; keyword: string; suffix: string; redMarked: boolean; rects: mupdfwv.TRect[] }[], pageIndex: number }[] }.
Example 
mupdf.text.search({keyword:"hello"});

// returns e.g.

{
    results: {
        pageIndex:number,
        words: {
            prefix: string;
            keyword: string;
            suffix: string;
            rects: {
                    top: number;
                    left: number;
                    bottom: number;
                    right: number;
                }[];
        }[];
    }[];
}

​
getSelected

getSelected() Gets the currently selected text from a document. If there is no selected text then returns null.
  • Returns: Promise<{ text: string, pageIndex: number, rects: TRect[] }>.
Example 
async function getSelectedText() {
    return mupdf.text.getSelected();
}

function success(result) {
    if (result) {
        console.log(`Text is: ${result.text}`);
        console.log(`Page index is: ${result.pageIndex}`);
        let rects = result.rects;
        console.log(`rect[0] metrics:\ntop=${rects[0].top}\nright=${rects[0].right}\nbottom=${rects[0].bottom}\nleft=${rects[0].left}`);
    }
}

function failure(error) {
    console.error(`Error: ${error}`);
}

getSelectedText().then(success, failure);

​
locateSource

locateSource(config: { text: string }) Searches for the given text and returns its coordinates for highlighting. This is designed to be used with LLM-generated answers that include quoted source text, allowing you to map that text back to the PDF for highlighting. The input text is expected to be high-quality Markdown from an LLM response.
  • Arguments:
    • config – The configuration object.
Config object:
  • Arguments:
    • text – (required) Query text.
  • Returns: Promise<{ pageIndex: number, words: { rects: mupdfwv.TRect[] }[] }>.
Example 
mupdf.text.locateSource({text:text}).then(
    function success(data) {
        console.log(data);
    },
    function failure(error) {
        alert(`Error locating text: ${error}`);
    }
);
Refer to Source Locator: Example for a more detailed example.