Extraction rules

wuchale doesn’t force you to only use the predefined extraction rules. Those are just defaults. What gets extracted is decided by the heuristic function which you can customize. It is implemented by the adapters so you can configure it on a per-adapter basis. Sensible default heuristic functions are provided out of the box. They implement the following rules (where applicable).

Default rules

General rule (applies anywhere)

If the message contains no letters used in any natural language (e.g., just numbers or symbols), it is ignored.

Ignored elements

These are elements where all markup or attribute is ignored. They are:

• style
• path
• code
• pre

In markup (<p>Text</p>)

All textual content is extracted unless the element is ignored.

Examples:

<p>This is extracted</p>

In attribute (<div title="Info">)

• The element must not be ignored.
• If the first character is a lowercase English letter ([a-z]), it is ignored.
• The method attribute of a form is ignored.
• Otherwise, it is extracted.

Examples:

<img alt="Profile Picture" class="not-extracted" />

In script (<script> and .js/ts)

• If the value is not inside a function definition, it is ignored.
• If the value is inside console.*() or fetch calls, it is ignored.
• If the first character is a lowercase English letter ([a-z]) or is any non-letter character, it is ignored.
• Otherwise, it is extracted.

Examples:

const message = 'This is NOT extracted' // it's not inside a function

function foo() {
    // extracted
    const extracted = 'Hello!'
    const nonExtracted = '-starts with non letter'
}

Per-adapter defaults

In addition to the above rules, the adapter may have additional restrictions to provide default rules as good as possible. You can see them in the specific adapter’s page.

Global overrides

If you need more control, you can supply your own heuristic function in the configuration. Custom heuristics can return undefined or null to fall back to the default. For convenience, the default heuristic is exported by the package.

With that, it is easy to handle just one case or a few cases and let the default heuristic handle the rest. For example, if we want to ignore all title attributes that contain the + character, the custom heuristic would be:

wuchale.config.js

    //...
    adapters: {
        main: svelte({
            heuristic: (msg, details) => {
                if (details.attribute === 'title' && msg.includes('+')) {
                    return false
                }
            }
        })
    },
    //...

And the rest of the checks will be handled with the default.

Specific overrides

If you don’t want to modify the global heuristic but want to ignore or include just a few messages, you can use comment directives.

• @wc-ignore — skips extraction
• @wc-include — forces extraction