Age | Commit message (Collapse) | Author |
|
|
|
|
|
It looks like rustfmt does not support being called from a virtual
workspace anymore, so we actually call it twice, one for each package
of the workspace.
|
|
|
|
|
|
This is specially useful for NaNoWriMo adventures. (:
|
|
Prior to this patch, these methods were always constructing a
`String'-based result, which was fine when all we wanted to do with
said result is use it with a template engine, but not that fine
when we aim to provide a tool to count words.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This script shall be installed manually with eg., a symbolic link.
|
|
The main idea behind these options is to make it possible to integrate
a celtchar book inside a larger static website generated by other
means (like Hakyll or soupault, for instance).
|
|
|
|
|
|
In this new iteration of libceltchar, there is now a very generic
`BookWriter` that the `EpubWriter` extends. This paves the way towards
a static website generator. In addition, we start using `rustfmt`,
which means the diff is actually a bit larger that what it shall be.
|
|
We provide an additional command `deps' that basically print one
dependency of —an ogmarkup document used to build a— celtchar
document. This way, `celtchar deps' can be used e.g., in a Makefile.
If a filename cannot be converted into a valid rust string, the
message <invalid utf8 filename> is print.
|
|
In addition to define the typography to use when generating ogmarkup
document, celtchar also sets the ebook metadata accordingly. This
patch required to update ogmkarkup for [ogmarkup::compile] works with
Typography objects of unknown size.
|
|
The rationale behind this choice is to provide a less cumbersome entry
point for loading a Book project in memory.
|
|
At some point, we will want to create a website to write ogmarkup
document in a collaborative manner. To prepare that time, we abstract
away the loading process of a Book project with a dedicated trait.
In addition, the project is now divided into a library on the one
hand, and the celtchar executable on the other hand. The celtchar
executable uses its own implementation of the [Loader] trait to load a
Book project from the filesystem, as before. This means, currently,
the related functions are not provided by the celtchar library.
|
|
|
|
There is now an optional option `numbering' one can set in the
Book.toml file. If set to true, numbers will appear consistently in
the book (even if no title is set). If unset, only the book title,
when it exists, will be set.
|
|
Previously, the content of the epub was generated in a dedicated build
directory, and it was necessary to make the final step (generating the
archive) manually. With this patch, the archive is generated, but no
compression method is performed.
We add a new dependency to the `zip-rs' crate, since the latter
appears to be the go-to solution for generating zip archives.
|
|
The purpose of this trait is to abstract the medium onto which the
epub is generated. Previously, the use of the filesystem was
hard-coded. With this patch, we add an abstraction level which will
eventually allows us from using the same code to generate a zip
archive or an in-file-system directory.
|
|
There is currently no way to overload these files, e.g., you cannot
specify your own font to be used for example. This might (probably
will actually) change in the future.
|
|
|
|
The reference is used is http://validator.idpf.org/
|
|
|
|
|
|
Default font is et-book, from https://github.com/edwardtufte/et-book.
It is a very beautiful font, but there is no support for bold+italic.
Therefore, so we will probably have to change it in the future.
|
|
|
|
|
|
The template engine of choice is Tera, mostly because I already knew
how to use it.
|
|
|
|
|