These *.mkd files are now the master files for the book (i.e. I will not run the conversion any more).
To produce the PDF of the book, you will need to install LaTeX on your system. For Debian-derived (Ubuntu, Mint, etc.) Linux:
sudo apt-get install texlive-full
sudo apt-get install pandoc
For Macintosh,
https://www.tug.org/mactex/
https://www.tug.org/mactex/mactextras.html
To produce the book run
bash book.sh
The output bash book.sh is in the file x.pdf and x.epub.
Note that the build scripts require Python 2
There is a continuous build server that I have on Digital Ocean that builds the PDF and epub versions of the book every hour or so.
http://do1.dr-chuck.com/pythonlearn/EN_us/pythonlearn.pdf
http://do1.dr-chuck.com/pythonlearn/EN_us/pythonlearn.epub
Just check in changes and these files will update.
This server also rebuilds any translations every hour or so.
In addition to the official book.sh1, there are other build scripts that make
alternate versions of the book. If you make changes to the content, you
should run all these scripts and check it all into github so it ends up online.
-
phpbook.shwill make an html/php verion of the book that is an extension of this site in../html3- this is then checked into github. -
htmlbook.shwill make an html verion of the book, with interactive examples embedded in trinkets and then make a zip file. -
zipbook.shwill make two html versions of the book with Trinket branding, one with interactive examples (that require an internet connection to work) and one with syntax highlighted code blocks for completely offline viewing. A zip containing these is at/book3/zips/pfe.zipif you'd just like to download it. -
trinketbook.shwill make the nunjucks template that we use to host the book at books.trinket.io. This is likely not of use to you unless you're looking for an example of how to get the book source into your own templating language. If you'd like to see the output of this script it's inbooks/trinket/pfe. Also updates../trinket3.
If you'd like to make your own build script, you can use these as examples. If your build script might have use to others, consider contributing it in a pull request. Note that each build script plays nicely with the others and the represent parallel workflows. Please don't alter any of the python scripts that are used by another script if you intend on contributing a new script.
The book.sh script will generate the x.mobi file is KindleGen is in the path:
https://www.amazon.com/gp/feature.html?docId=1000765211
For Linux:
curl -O http://kindlegen.s3.amazonaws.com/kindlegen_linux_2.6_i386_v2_9.tar.gz
tar xfv kindlegen_linux_2.6_i386_v2_9.tar.gz
cp kindlegen /usr/local/bin
Just take the x.pdf and x.mobi files and copy them into the createspace
folder, adding a date in the filename as version and then upload them to
createspace ins kindle publishing.
If you are building the file for Chinese, touch the file
touch .chinese
So it runs LaTeX in a way to produce chinese documents.
Install the 'Noto Serif CJK SC' Font. Download from here and unzip:
https://noto-website.storage.googleapis.com/pkgs/NotoSerifCJK.ttc.zip
You should get a file like NotoSerifCJK.ttc.
On Mac copy the file to ~/Library/Fonts and rebuild font cache:
sudo atsutil databases -remove
On Linux put them in /usr/share/fonts:
[ -d /usr/share/fonts/opentype ] || sudo mkdir /usr/share/fonts/opentype
[ -d /usr/share/fonts/opentype/noto ] || sudo mkdir /usr/share/fonts/opentype/noto
sudo mv NotoSerifCJK.ttc /usr/share/fonts/opentype/noto
sudo fc-cache -f -v
Strange Note: If you are running on a small memory linux system, you
may encounter the error "I can't write on file test.pdf'" - turns out this is xelatex` running out of RAM - this fixed it.
dd if=/dev/zero of=/var/512mb.swap bs=1M count=512
mkswap /var/512mb.swap
swapon /var/512mb.swap
Ref: https://tex.stackexchange.com/questions/16801/xelatex-i-cant-write-on-file-test-pdf
If you want to contribute, feel free to fork the py4e repository and send me pull requests.
We can also use the issue tracker to coordinate if that helps.
/Chuck