I referred to here for the basic method. [How to add Sphinx-based documentation to Dash.app (Japanese documentation is OK) --Programmer March Chapter 2](http://takuan-osho.hatenablog.com/entry/2013/02/20 / how-to-add-sphinx-based-documentation-in-dash-app)
So, as mentioned on this site, I couldn't index Classes or Modules just by simply converting.
Below are the steps for Mac OS X 10.9.2.
I didn't have pip, so I installed it with easy_install. You need to install command line tools. If you get an error while building the dependent package lxml,
sudo ARCHFLAGS=-Wno-error=unused-command-line-argument-hard-error-in-future easy_install doc2dash
It was okay if I set the flag and execute it. Probably the same with pip. Reference: Error installing python module lxml on osx mavericks --Stack Overflow
Download a set of html from python-doc-ja --Python document Japanese translation project --Google Project Hosting and convert it with doc2dash.
doc2dash -n Python2.7ja -i _static/py.png -I index.html -d pathto/output pathto/python-doc-2.7ja1-html
When I tried feeding this to Dash, the index was not created after all.
So I decided to convert the same version of the English document and see the difference.
Download a set of html from Download — Python v2.7.6 documentation and convert it with doc2dash. It took longer than the Japanese document, and the index was completed normally. Looking inside the docset, the size of docSet.dsidx
was clearly different, so I copied this to the Japanese docset. Then I got a list of Classes and Modules and the links worked fine. I haven't validated all the indexes, so it may be inconsistent somewhere.
It may be possible to extract from the genuine Dash docset, but it seems that the conversion method is different from doc2dash, and the index file type is different.
I think you can do the same with Python 3 series, but I haven't tried it because there seems to be no place to download Japanese documents all at once.
Dash - Documentation Browser, Snippet Manager - Kapeli doc2dash 1.2.0 : Python Package Index
I made the 3.6.1 version by forcibly replacing the html of the English version of docset. You can use the full index of the English version as it is. It's a bit complicated script because you need to embed the anchor in the html. https://gist.github.com/hetima/26f21023c417e68be1d4db2d1e388816
Recommended Posts