mirror of https://github.com/wallabag/wallabag.git
53 lines
1.6 KiB
ReStructuredText
53 lines
1.6 KiB
ReStructuredText
Write config files
|
|
==================
|
|
|
|
wallabag can use specific site config files to parse website articles.
|
|
These files are stored in the
|
|
```inc/3rdparty/site_config/standard`` <https://github.com/wallabag/wallabag/tree/master/inc/3rdparty/site_config/standard>`__
|
|
folder.
|
|
|
|
The format used for these files is
|
|
`XPath <http://www.w3.org/TR/xpath20/>`__. Look at some examples in the
|
|
folder.
|
|
|
|
Automatic config files generation
|
|
---------------------------------
|
|
|
|
Fivefilters has created a `very useful
|
|
tool <http://siteconfig.fivefilters.org/>`__ to create config files. You
|
|
just type in the adress of the article to work on with, and you select
|
|
the area containing the content you want.
|
|
|
|
.. figure:: https://lut.im/RNaO7gGe/l9vRnO1b
|
|
:alt: siteconfig
|
|
|
|
siteconfig
|
|
| You should confirm this area by trying with other articles.
|
|
| When you got the right area, just click on *Download Full-Text RSS
|
|
site config* to download your file.
|
|
|
|
Manual config file generation
|
|
-----------------------------
|
|
|
|
If Fivefilters tool doesn't work correctly, take a look at the source
|
|
(Ctrl + U on Firefox and Chromium). Search for your content and get the
|
|
``class`` or the ``id`` attribute of the area containing what you want.
|
|
|
|
Once you've got the id or class, you can write for example one or
|
|
another of these lines:
|
|
|
|
::
|
|
|
|
body: //div[@class='myclass']
|
|
body: //div[@id='myid']
|
|
|
|
Then, test you file. If you got the right content but you want to strip
|
|
unnecessary parts, do:
|
|
|
|
::
|
|
|
|
strip: //div[@class='hidden']
|
|
|
|
You can look at other options for siteconfig files
|
|
`here <http://help.fivefilters.org/customer/portal/articles/223153-site-patterns>`__.
|