The filters `|linebreaks`, `|linebreaksbr` and `|paragraphbreaks`
generate HTML code to be embedded in a page. Having to specify that the
output of these filters is `|safe` is cumbersome. Also, these filters
need to operate on already escaped HTML data. This could be done by
writing `{{ s|escape|linebreaks|safe }}`.
This PR does the input and output formatting escaping for the user. The
input gets escaped for HTML (invariant of the selected escaper), and the
output gets marked as HTML safe.
This PR reimplements the code generation for `{% filter %}` blocks, so
that the data is written directly into its destination writer, without
using a buffer. This way it behaves like a specialized
`{{ expr|filter1|filter2 }}` would, if the `expr` was a `Template` that
contained the body of the filter block.
By using codepoint entities like `'&'` → `"&"`, we have a much
smaller lookup table (58 bytes instead of 29× pointer size ~= 232
bytes). This makes the cache happy, and the benchmark run about ~20%
faster.
```text
$ cargo bench --bench escape
Escaping time: [3.4087 µs 3.4126 µs 3.4168 µs]
change: [-19.790% -19.580% -19.354%] (p = 0.00 < 0.05)
Performance has improved.
```
This PR configures the formatting with a few defaults that aid
readability, in my opinion. One drawback of adding this file is that
rustfmt uses unstable features, now, and you have to use nightly to run
it:
```sh
cargo +nightly fmt --all
```
This PR adds an optional argument to the `|tojson` filter, which
controls if the serialized JSON data gets prettified or not. The
arguments works the same as flask's [`|tojson`][flask] filter, which
passes the argument to python's [`json.dumps()`][python]:
* Omitting the argument, providing a negative integer, or `None`, then
compact JSON data is generated.
* Providing a non-negative integer, then this amount of ASCII spaces is
used to indent the data. (Capped to 16 characters.)
* Providing a string, then this string is used as prefix. I attempts are
made to ensure that the prefix actually consists of whitespaces,
because chances are, that if you provide e.g. `&nsbp;`, then you are
doing it intentionally.
This is a breaking change, because it changes the default behavior to
not prettify the data. This is done intentionally, because this is how
it works in flask.
[flask]: https://jinja.palletsprojects.com/en/3.1.x/templates/#jinja-filters.tojson
[python]: https://docs.python.org/3/library/json.html#json.dump
