worm-scraper

This version adds initial support for Worm's sequel, Ward.

I have not yet done a read-through of the book, so presumably many text fixups remain to be done. I hope to add them over time, but I can give no promises.

The command-line options have changed slightly:

• -s / --start-url is removed, replaced with -b / --book
• -c / --cache-directory is replaced with -c / --cache
• -b / --book-directory is replaced with -s / --staging

And a final bugfix: for Worm, the "description" of the generated eBook no longer has spurious linebreaks inserted.

This version raises the minimum required Node.js version to v12.10.0, and includes internal dependency updates. There should be no user-visible changes.

Added the --jobs option for the convert command, to allow specifying the number of concurrent conversion read/write jobs that are attempted. This doesn't appear to be that helpful, at least on my machine, as things seem to happen rather sequentially either way. But it's better than the previous architecture of hard-coding it to 10 jobs.

Fixed the package metadata to explicitly state one of the dependencies, instead of relying on it getting implicitly installed. This shouldn't matter in theory, but in practice it seems to have potentially caused an issue for at least one user (#14).

General fixups:

• Replaced four instances of "grey" with "gray" (including one "Grey Boy").

Included npm-shrinkwrap.json in the published package. This ensures that installations from the npm registry always have the exact same dependency versions as I am testing locally, which could increase reliability.

Individual fixups for Arc 11, Infestation, and Arc 12, Plague.

General fixups:

• More of the never-ending italics/quotes fixes.
• Fixed to use apostrophe instead of opening quote for "’Sup?".
• Standardized on the accent-including spelling of cliché.
• Hyphenated "hard-headed", "self-destruct(ing)", and "life-threatening".

Updated under-the-hood dependencies. (This should not generally be noticeable, except perhaps fewer error cases or better performance.)

Many new fixups to the text based on my most recent read-through.

Individual fixups are present through Arc 10, Sentinel; maybe I will move over the rest of the fixups at some point. (Currently they exist as highlights in my eBook reader.)

In addition to individual fixups, the following general fixups were applied:

• Removed the introductory pseduo-trigger warning.

• Removed italics from more foreign names.

• De-capitalized PRT power designations; early chapters especially over-capitalized them.

• De-capitalized "halberd", "north", "south", "east", and "west".

• De-capitalized words after colons.

• Capitalized "Birdcage".

• Capitalized place names, e.g. "Lord street" became "Lord Street".

• Added hyphens to "shoulder-mounted", "golden-skinned", "creepy-crawly", "well-armed".

• Added missing periods to the end of some sentences.

• Settled on "okay", "TV", and "signaled", replacing various other forms such as "ok", "T.V.", or "signalled".

• Fixed ampersands to become a spelled-out "and", where appropriate.

• Fixed mismatched single quotes.

• Fixed possessives of names ending in "s".

• Fixed more em dash usage.

• Fixed period and ellipsis usage, e.g. removing spaces before periods.

• Fixed all separators to uniformly use <hr>, instead of sometimes-left and sometimes-center aligned ■ characters.

• Tried to fix incorrect punctuation and capitalization in dialogue tags. For example,

  “I didn’t get much done,” Greg said, “I got distracted by...

  was corrected to

  “I didn’t get much done,” Greg said. “I got distracted by...

  and

  “Basically,” Alec said, “For your powers to manifest, ...

  was corrected to

  “Basically,” Alec said, “for your powers to manifest, ...

  This correction is probably incomplete and may require more individual fixups after Arc 10.

No user-facing changes; only behind-the scenes refactoring.

Marked as a major version because it increases minimum Node.js version requirement to at least v8.x.

Fixed an issue where non-Windows users could not run the scraper easily by following the README instructions.

Text fixes:

• Corrected some whitespace placement (mostly invisible) around italics.
• Removed sporadic italics from "Sifara" and "Thanda".
• Ensured "Yàngbǎn" always has its diacritic marks.

Initial release.