Get link on webpage with specific text

[evanfuchs]

I've read various posts about Xpath and working with web elements. Could someone help me learn by sharing a macro that finds the link on Google.com with the text "About" and displays its URL?

For example, I can open the link using Click on the first web page item that matches an XPath with this xpath //a[contains(., "about")] - but I am just trying to learn the simple steps of 1) Finding a link by its text, and 2) returning the URL as a variable or display results.

[evanfuchs]

Sharing this progress in case it helps anyone. Firing this macro from Safari at google.com will find the link with the link text "About" and display its URL.

(function (strPath) {
    const link = document.evaluate(strPath, document, null, XPathResult.ANY_TYPE, null).iterateNext();
    if (link) {
        return link.href;
    } else {
        return null;
    }
})(document.kmvar.local_xPath);

[ComplexPoint]

And here's one way of collecting ALL links on the active page for which the label text contains a given string.

(Assumes Keyboard Maestro 11+)

ALL links containing text.kmmacros (4.8 KB)

---

Expand disclosure triangle to view JS source

const main = () =>
    Array.from(
        xPathMatches( kmvar.local_xPath )
    )
    .map(x => `[${x.textContent}](${x.href})`)
    .join("\n");


// --------------------- GENERIC ---------------------

// Just :: a -> Maybe a
const Just = x => ({
    type: "Maybe",
    Just: x
});

// Nothing :: Maybe a
const Nothing = () => ({
    type: "Maybe",
    Nothing: true
});

// xPathMatches :: XPath String -> Gen [Node]
const xPathMatches = xpath =>
    // A list of any matching nodes in the document.
    // (Can be made strict by wrapping in Array.from)
    unfoldr(x => {
        const v = x.iterateNext();

        return Boolean(v)
            ? Just([v, x])
            : Nothing();
    })(
        document.evaluate(
            xpath,
            document,
            null,
            XPathResult.ORDERED_NODE_ITERATOR_TYPE
        )
    );

// unfoldr :: (b -> Maybe (a, b)) -> b -> Gen [a]
const unfoldr = f =>
    // A lazy (generator) list unfolded from a seed value
    // by repeated application of f to a value until no
    // residue remains. Dual to fold/reduce.
    // f returns either Nothing or Just (value, residue).
    // For a strict output list,
    // wrap with `list` or Array.from
    x => (
        function* () {
            let maybePair = f(x);

            while (!maybePair.Nothing) {
                const valueResidue = maybePair.Just;

                yield valueResidue[0];
                maybePair = f(valueResidue[1]);
            }
        }()
    );

// MAIN ---
return main();

[evanfuchs]

I've only started with Javascript, so this is over my head in terms of learning for now, but I can certainly use it. I see you're adding all matches to an array and creating a list of links from the array variable. I'll come back to the code when I know a little more.

Thanks!

[ComplexPoint]

Note that the Modern Syntax option in Keyboard Maestro 11 (see options in small chevron menu to left of the const in the JS text field below)

1. Allows a simpler reference to Keyboard Maestro variable values,
2. uses a return expression to make a value available to the calling macro.

---

For example, to return only the first Xpath match:

FIRST link containing text.kmmacros (3.7 KB)

---

Expand disclosure triangle to view JS source

const 
    firstMatch =  document.evaluate(
        kmvar.local_xPath,
        document,
        null,
        XPathResult.ORDERED_NODE_ITERATOR_TYPE
    )
    .iterateNext()

return `[${firstMatch.textContent}](${firstMatch.href})`;