• Formatting a book with CSS
• Pages …
• Hyphenation, text, fonts…
• Crop, Bleed, Spreads, Binding & Finishing…
• Distribution, and Metadata

@page {
  size: A4;
  /* or,
   * width: 21cm;
   * height: 29.7cm;
   */
}
@page:left {
  @top-left {
    content: "Page " counter(page);
  }
}
@page:right {
  @top-right {
    content: "Page " counter(page);
  }
}

• Formatted content inside the page body generally has to come from the document, not the style sheet …
• Use position: fixed on the style for an element in the document and it will appear on every page for its container. …
• You can set strings in an element, and use them in CSS (procedural!) …
• Experimental: you can move an element out of the document into running heads/feet.

• CSS can't reorder content, so you have to include the ToC in the document. …
• Page numbers and dot leaders (if wanted) are supplied by the formatter.

div.contents ul li a:after {
  content: " " leader(dotted) " " target-counter(attr(href, url), page);
}

<div class="contents">
  <h1>Table of Contents</h1>
  <ul><li><a href="#preface">Preface</a></li>
      <li><a href="#c1">Chapter One</a></li>
  </ul></div>

Table of Contents

Preface . . . . . . . . . . . 7
Chapter One  . . . . . . . . 11

• The ToC example used a CSS selector:
  div.contents ul li a:after
• CSS selectors have “specificity” - all the rules for an element are applied, least specific first …
• This is different from XSLT, where an entire template is applied or not applied, rather than parts of a template being used to override parts of other templates. …
• Debugging selectors can be hard!

• Both XSL-FO and CSS position outer boxes, you don’t position text by baseline. …
• In a browser situation you don’t know which font wll be used, so you can’t get reliable positioning.
• The Digital Publishing IG has worked on requirements here.

• A back-of-the-book index requires collation and sorting of index terms and merging of references to the same page.
• You can do the collation and sorting in XSLT and put the index in the document, but you can’t merge adjacent entries.
• You could use JavaScript to merge entries, but there’s not yet any API to do this.