wnget 0.3.1

web novel getter, a web scraping and ebook generation tool.

Rarely will web novel sites provide any means to read their contents off line, and this is precisely why wnget came to be. It is a tool to scrape web novels from blogs, and optionally convert them to epub format. It provides several options to configure the exact behaviour, while at the same time trying to provide sane defaults. The strings for the next/previous navigation links, as well as CSS class for title/content containers can be configured, among other settings.

Besides the main wnget utility, the suite include wnbook, to generate ebooks from the downloaded contents, and wnlocal, to rewrite the links of html files that can be resolved to a local file using the same rules applied by wnget.

Installation

$ pip install wnget

If you happened to have cloned the repo and are playing with the code, you probably want to install wnget in “editable” mode while you’re working on it. This is so it becomes both installed and editable in project form.

Assuming you’re in the root of the project, just run:

$ pip install -e .

Configuration

Any argument that starts with ‘–’ can also be set in a config file (./wnget.conf or ~/.wnget.conf or specified via -s). The recognized syntax for setting (key, value) pairs is based on the INI and YAML formats (e.g. key=value or foo=TRUE). If an argument is specified in more than one place, then commandline values override config file values which override defaults.

The default configuration looks like this:

keeplinks: no
firsttitle: no
next: "Next Chapter"
previous: "Previous Chapter"
titleclass: entry-title
contentclass: entry-content
epub: default_ebook
limit: 0 # No limit if < 1

Since every web novel site will have a unique theme, it is unlikely the default configurations shipped with _wnget_ will work out of the box (unless you happen to be addicted to the same ones as yours truly). Feel free to extend data/sites.yaml for your favorite sites and issue a pull request.

Usage

$ wnget --help
usage: wnget [-h] [-s file] [-k] [-f] [-n caption] [-p caption] [-t css_class]
             [-c css_class] [-e title] [-l N] [-v]
             first_url [last_url]

positional arguments:
  first_url             first URL to crawl
  last_url              optional last URL to crawl (stops after reaching)

optional arguments:
  -h, --help            show this help message and exit
  -s file, --settings file
                        config file path
  -k, --keeplinks       rewrite and keep navigation links in HTML content
  -f, --firsttitle      keep first title match instead of trying to be smart
                        about it
  -n caption, --next caption
                        next link caption (default: 'Next Chapter')
  -p caption, --previous caption
                        previous link caption (default: 'Previous Chapter')
  -t css_class, --titleclass css_class
                        title container class (default: 'entry-title')
  -c css_class, --contentclass css_class
                        content container class (default: 'entry-content')
  -e title, --epub title
                        create Epub with this title (will use cover.jpg/png if
                        found)
  -l N, --limit N       crawl at most N pages
  -v, --version         show program's version number and exit

Usage examples

Invoke each command without arguments to display help information.

The scraper looks for a given CSS class in the title and content containers, and those can be set manually to suit your web novel site of choice. It also looks for links with default strings to find the next and previous chapters, and this can also be set by hand.

To scrape all chapters of a given web novel, following links, and saving each chapter in a different html file in the current directory:

$ wnget http://example.com/first_chapter_link

Or, for more advanced uses, downloading all chapters until a given link is retrieved, and generate an EPUB with the loot:

$ wnget -e "My Web Novel" \
  http://example.com/first_chapter_link  \
  http://example.com/first_chapter_link

Additionally, the ebook functionality can be used directly through the wnbook standalone utility. Just provide the HTML index file and a name for the book, and it will generate an ebook with all referenced resources in the working directory:

$ wnbook index.html "My Web Novel"

Also, if a cover.png or cover.jpg file is present, it will be used as cover page. Its use as standalone command will often prove more flexible, as it exposes features not normally used by the main application, while allowing some manual tweaking of the downloaded contents and index files.

Here, generating a book with relative paths, and custom filename, cover image, and language/author metadata:

$ wnbook ../index.html "I Shall Seal the Heavens (我欲封天)" \
  --filename=issth.epub --language=zh --author="Ergen (耳根)" \
  --cover ~/images/MengHao.png

And finally, to rewrite links within a file so that they point to already downloaded resources, wnlocal can be used.

Either to print the converted file to stdout:

$ wnlocal introduction.html

Or to write it back to disk:

$ wnlocal introduction.html newfile.html

TODO

• Make content selection more flexible: by tag, class, caption, or xpath.

• Add option for elements to be removed during the parsing stage.

• Add interactive title selection mode (and ability to repeat choice).

• In-content image support (download, and store locally).