XML Syntax

The parts of an XML document

<?xml version="1.0" encoding="UTF-8"?>          <!-- 1. XML declaration (optional, must be first) -->
<!-- 2. A comment -->
<library>                                        <!-- 3. The root element: exactly one -->
  <book id="b1" lang="en">                       <!-- 4. Start tag with attributes -->
    <title>Dune</title>                          <!-- 5. Element with text content -->
    <available/>                                 <!-- 6. Empty element (self-closing) -->
  </book>                                        <!-- 7. End tag -->
  <?print-style compact?>                        <!-- 8. Processing instruction -->
</library>

• The declaration states the XML version and character encoding. It's optional (UTF-8 is the default), but if present it must be the very first thing in the file, with no whitespace before it.
• There is exactly one root element; everything else lives inside it.
• Every start tag needs a matching end tag (<title>…</title>), or the element is written self-closing (<available/>).
• Attributes are name="value" pairs in the start tag; values are always quoted (single or double quotes).
• Comments are <!-- … --> and can't contain --.
• Processing instructions (<?target data?>) pass hints to applications, like <?xml-stylesheet?>.

Names

Element and attribute names are case-sensitive (<Book> and <book> are different), must start with a letter or underscore, and may contain letters, digits, -, _ and .. Names starting with xml (in any case) are reserved.

The document is a tree

A parser turns the text into a tree of nodes. Walking it shows the structure:

from lxml import etree

doc = etree.fromstring(b"""<library>
  <book id="b1" lang="en">
    <title>Dune</title>
    <available/>
  </book>
  <!-- more books soon -->
</library>""")

def show(node, depth=0):
    pad = "  " * depth
    if isinstance(node, etree._Comment):
        print(f"{pad}comment: {node.text.strip()}")
        return
    attrs = " ".join(f'{k}="{v}"' for k, v in node.attrib.items())
    text = (node.text or "").strip()
    print(f"{pad}<{node.tag}{' ' + attrs if attrs else ''}>{' text=' + repr(text) if text else ''}")
    for child in node:
        show(child, depth + 1)

show(doc)
print("root:", doc.tag, "| children:", len(doc))

<library>
  <book id="b1" lang="en">
    <title> text='Dune'
    <available>
  comment: more books soon
root: library | children: 2

Whitespace between elements is kept as text nodes. Most programs ignore it, but it matters for documents (mixed content) and when you compare or sign XML.