## html_to_text()


Derive plain text from `html`.


Usage

``` python
html_to_text(html)
```


It writes a link as `label <url>`, or as the label alone when the label is already the URL or the `mailto:` address, or when the URL is a fragment such as `#top`. It keeps the line breaks and indentation of a `<pre>` block, such as a code block or a log. It starts a list item with `-`. It writes a table row on one line with `|` between its cells when the cells hold only inline content. A block or a `<br>` inside a cell starts a new line, so the paragraphs of a layout table stay apart. It writes empty cells too, so it never shifts a value into the wrong column. It writes an image as its alt text in brackets. It drops `<style>`, `<script>`, `<title>`, and comments. It decodes entities. It returns `""` for HTML that holds no text, such as a lone image with no alt text.

It never raises on malformed HTML. The output is best effort and not a contract, so it may change in a minor version. Pass `text=` to [Message](Message.md#epistole.Message) when the exact text matters. See ADR-0008.


## Examples

``` python
from epistole import html_to_text

html = '<p>The report is [online](https://example.com/q3).</p>'
html_to_text(html)  # "The report is online <https://example.com/q3>."
```