Fix missing images in mobi format. (#20)

This commit is contained in:
Nemo 2019-01-31 15:49:41 +05:30 committed by GitHub
commit 41c6c3eff2
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
6 changed files with 49 additions and 23 deletions

View File

@ -15,6 +15,7 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
ruby-dev \
wget \
zlib1g-dev \
file \
&& gem install bundler --no-ri --no-rdoc \
&& apt-get clean \
&& rm -rf /var/lib/apt/lists/*

View File

@ -39,7 +39,7 @@ Requirements:
- Ruby
- `gem install bundler`
- `gem install nokogiri`
- `bundle install`
- `brew install pandoc`
- `brew cask install calibre`
- `brew install wget`

View File

@ -2,7 +2,13 @@
set -euo pipefail
IFS=$'\n\t'
TOC_URL="https://landing.google.com/sre/sre-book/toc/index.html"
# Vars.
export BOOK_NAME="sre-book"
export BOOK_NAME_FULL="Site Reliability Engineering"
BOOK_FILE="google-${BOOK_NAME}"
TOC_URL="https://landing.google.com/sre/${BOOK_NAME}/toc/index.html"
IMGS_DOMAIN="lh3.googleusercontent.com"
# Make sure that links are relative \
# # Remove the /sre/ directories
# Save stuff in html/ directory
@ -11,36 +17,55 @@ TOC_URL="https://landing.google.com/sre/sre-book/toc/index.html"
# Images are hosted elsewhere, download them as well.
# We need to go up a level from /toc/ where we start
wget \
--convert-links \
--convert-links \
--directory-prefix=html \
--page-requisites \
--adjust-extension \
--span-hosts \
--trust-server-names \
--backup-converted \
--mirror \
--no-verbose \
--recursive \
--domains=lh3.googleusercontent.com,landing.google.com https://landing.google.com/sre/sre-book/toc/index.html
--page-requisites \
--adjust-extension \
--span-hosts \
--trust-server-names \
--backup-converted \
--mirror \
--no-verbose \
--recursive \
--domains=${IMGS_DOMAIN},landing.google.com ${TOC_URL}
#
MODE=${1:-}
if [ "$MODE" != "docker" ];then
bundle install
fi
ruby generate.rb
# Add extension to files.
# That because `pandoc` cannot generate the right `mime type` without the extension.
# https://github.com/captn3m0/google-sre-ebook/issues/19
IMGS_FILES="$(ls html/${IMGS_DOMAIN}/*)"
for FILE_NAME_FULL in ${IMGS_FILES}; do
pushd html/landing.google.com/sre/sre-book/toc
pandoc --from=html --to=epub --output=../../../../../google-sre.epub \
--epub-metadata=../../../../../metadata.xml \
--epub-cover-image=../../../../../cover.jpg \
# Get file vars.
FILE_NAME_BASE="$(basename ${FILE_NAME_FULL})"
FILE_TYPE=$(file -b -- "${FILE_NAME_FULL}" | cut -f1 -d " ")
# Rename and replace file.
mv "${FILE_NAME_FULL}" "${FILE_NAME_FULL}.${FILE_TYPE,,}" &&
grep -rl "${FILE_NAME_BASE}" ./html | xargs sed -i "s/${FILE_NAME_BASE}/${FILE_NAME_BASE}.${FILE_TYPE,,}/g"
done
bundle exec ruby generate.rb
pushd html/landing.google.com/sre/${BOOK_NAME}/toc
pandoc --from=html --to=epub \
--output=../../../../../${BOOK_FILE}.epub \
--epub-metadata=../../../../../${BOOK_NAME}.xml \
--epub-cover-image=../../../../../${BOOK_NAME}.jpg \
complete.html
popd
ebook-convert google-sre.epub google-sre.mobi
ebook-convert google-sre.epub google-sre.pdf
for EXTENSION in mobi pdf; do
ebook-convert ${BOOK_FILE}.epub ${BOOK_FILE}.${EXTENSION}
done
if [ "$1"=="docker" ]; then
chown -v $(id -u):$(id -g) google-sre.*
mv -f google-sre.* /output
chown -v $(id -u):$(id -g) ${BOOK_FILE}.*
mv -f ${BOOK_FILE}.* /output
fi

View File

@ -4,7 +4,7 @@ require 'fileutils'
# First we get the list of all the book sections:
Dir.chdir("html/landing.google.com/sre/sre-book/toc")
Dir.chdir("html/landing.google.com/sre/%s/toc" % ENV['BOOK_NAME'])
chapter_links = Nokogiri::HTML(open("index.html"))
.css('#drop-down a')
.map {|l| l.attribute('href').value}
@ -13,7 +13,7 @@ html = <<EOT
<!DOCTYPE html>
<html>
<head>
<title>Site Reliability Engineering</title>
<title>#{ENV['BOOK_NAME_FULL']}</title>
<meta charset="utf-8">
</head>
<body>

View File

Before

Width:  |  Height:  |  Size: 43 KiB

After

Width:  |  Height:  |  Size: 43 KiB