Resource Loading Code Changes

[liambloom]

• [ ] Remove ResourceLocator class; instead do bookUrl.toUri().resolve(resourceUri).toUrl().getInputStream() (Note: this would involve making book have a URL)
• [ ] Possibly store resource base in meta, rather than locator.*
• [ ] Check if book and resources are below the base URI (URL?) in BookParser rather than Checker*
• [ ] Check to make sure resource path does not have a relative path that goes outside of base at ANY time (e.g. "../foo" or "/path/to/base/foo"). Keep in mind that URI#resolve will normalize relative path.
• [ ] Make it work with jar URIs. (See code example below, and/or this stackoverflow explination)

// URL works fine with jars, but URI gets confused, and so resolve doesn't work properly
new URL("jar:file:///C:/foo/bar.jar!/foo").getPath() // => "file:///C:/foo/bar.jar!/foo"
new URI("jar:file:///C:/foo/bar.jar!/foo").getPath() // => null
new URI("jar:file:///C:/foo/bar.jar!/foo").resolve("./buz") // => "./buz"

// Non-jar URIs work fine
new URI("file:///C:/foo/bar.jar").resolve("./buz") // => file:/C:/foo/buz

* Items 2 and 3 may not be neccesary if I do:

• [ ] Automatically calculate the resource base from the Book and Resource URI/URLs

Notes:

• Resources should be URIs,
• Books should be URLs,
• IDK which the base should be if it's not calculated (I think a URI because it would be specified in the book, possibly relative to the book)

[liambloom]

Make it work with JAR/ZIP, by, if it is opaque and one of those two, creating a filesystem for the zip, changing the book URL (or the base, maybe?) to point in the zip filesystem, and resolving paths in the correct fs (zip fs or default)