```javascript
// jQuery-lite:
const q = (x,y=document) => y.querySelector(x)
const qq = (x,y=document) => Array.from(y.querySelectorAll(x))
// For taking a chunk of a webpage and making it the whole page.
const maxi = (x,y=document) => { let a; if( a = q(x,y) ) { document.body.innerHTML = a.innerHTML } }
const kill = (x,y=document) => qq(x,y).forEach(x => x.remove())
```
scraping example: get hrefs for all pdf links in a div element with class `download`
```javascript
copy(Array.from(document.querySelectorAll("div.download a"))
  .map(x => x.getAttribute("href"))
  .filter(x => x.match(/aw/\.pdf/))
  .join("\n"))
```
## Assembling contents for headers
The basic idea is this:
```javascript
cs = [0,0,0,0,0,0]
a = qq("h1, h2, h3, h4, h5, h6")
a.forEach(x => {
    tag = x.tagName
    lvl = tag.substr(1)|0
    for(let i=lvl;i<6;i++) {
        cs[i] = 0;
    }
    cs[lvl-1]++
    console.log(cs.join(".")+" "+x.textContent)
})
```
Then we can:
* See the highest level of heading (e.g. if there are only h2's and below) -- don't number things 0.1 0.2 etc
* Make a list of `(element,headingLevel,counters)` and from counters we extract e.g. the first 3 numbers for a h3, assuming
there is an h1, but e.g. if there are only h2's and below, we take the first 2 numbers after the first.
* Assemble a contents page -- if we assign `element` to a property of the element in the contents, we can use `scrollIntoView`
rather than a `a href='#name'` construct.
We will use an option `contents=yes` to signify that the page is to build and display a contents page.
We will use [CSS]() with an added class on the main content div to make things wider if we have a contents page.