* Removed the itertools dependency
* Removed an unused feature flag
* Stubbed out a toml_query replacement
* Update dependencies.
* Bump env_logger.
* Use warp instead of iron for http server.
Iron does not appear to be maintained anymore. warp/hyper seems to be
reasonably maintained. Unfortunately this takes a few seconds more
to compile, but shouldn't be too bad.
One benefit is that there is no longer a need for a separate websocket
port, which makes it easier to run multiple servers at once.
* Update pulldown-cmark to 0.7
* Switch from error-chain to anyhow.
* Bump MSRV to 1.39.
* Update elasticlunr-rs.
Co-authored-by: Michael Bryan <michaelfbryan@gmail.com>
At present, code listings without a main function will be wrapped in one and
annotated with an allow lint check attribute provided by the following [code][]:
```
format!(
"\n# #![allow(unused_variables)]\n{}#fn main() {{\n{}#}}",
attrs, code
)
```
A broader lint check attribute such as `#![allow(unused)]` seems like it might
better fit the apparent intent of this code.
Addresses: https://github.com/rust-lang/mdBook/issues/1192
[code]: 769cc0a7c1/src/renderer/html_handlebars/hbs_renderer.rs (L635-L638)
Html page title for every chapter in the book is created by the following code:
```rust
let title: String;
{
let book_title = ctx
.data
.get("book_title")
.and_then(serde_json::Value::as_str)
.unwrap_or("");
title = ch.name.clone() + " - " + book_title;
}
```
If the `book.toml` file is missing, and `book_title` parameter is empty, there's an awkward ` - ` string hanging at the end of each chapter's title.
This PR adds a `match` case to handle that kind of situation.
* Using .to_lowercase() on the theme matching to avoid needed exact capitolization in the book.toml
* Changed the default-dark-theme from .to_string() to .to_lowercase() to match theme
* Option to display copy buttons.
- Added field to playpen data structure
- Communicate through window.playpen_copyable
- Javascript updated to check before displaying copy buttons.
* html -> html_config
Also:
- update description of copyable in source code.
- update description of line_numbers (my last PR to this repository)
- Added line numbers to config struct
- Added playpen_line_numbers field to hbs renderer.
- Added section to set `window.playpen_line_numbers = true` in page template
- Use line number global variable to show line numbers when required.
Use case: when trying to `mdbook test` a file that has many `include`
directives, and a test fails, the line numbers in the `rustdoc` output
don't match the line numbers in the original markdown file.
Turning on the markdown renderer implemented here lets you see what is
being passed to `rustdoc` by saving the markdown after the preprocessors
have run.
This renderer could be helpful for debugging many preprocessors, but
it's probably not useful in the general case, so it's turned off by
default.
* Update to pulldown-cmark 0.4.1.
* Update to pulldown-cmark 0.5.2.
* Remove pulldown-cmark-to-cmark dependency.
Since it is not compatible with the new pulldown-cmark. This example isn't
directly usable, anyways, and I think the no-op example sufficiently shows how
to make a preprocessor.
* cargo fmt
* Fix example link.
Previously, additional JavaScript files inside a directory were
correctly copied (with their parent created), but the link to it was
stripped of that parent.
There's no need for that (and it was not done for CSS)
This class will supress the "play" button in the html backend (which you
can also do with the "ignore" class), but it will still let the code be
tested via `mdbook test` (which is not possible with the "ignore" class).
This is useful for code examples that don't really do much (and so the
user doesn't gain much from running them), but as an author you still
want to test them to guard against syntax errors and typos and the like.
This allows the search index to be loaded asynchronously, and should
use fewer resources as it doesn't have to execute the JS.
JS loading is kept as a fallback for CORS issues with file:// URIs in Chrome.