Excluding included files from spelling check

[kondziu]

Let's say a document includes a file via include or csv-table:

.. include:: data_dump.preformatted.log

.. csv-table::
   :file: data.csv
   :header-rows: 1

Then, the contents are subject to a spellcheck. Some of these should be excluded from spellcheck though, if they include data or program output. But they are apparently always included, even if I list them in spelling_exclude_patterns.

Is there a way to exclude them?

Listing all the words individually in the enclosing document (e.g. with the spelling directive) technically works, but it is impractical because (a) I don't want to exclude checking the spelling of these words from the enclosing document, and (b) it's a big list.

[dhellmann]

A literal include directive should cause the content to be ignored, but I haven't actually tested that.

There was a similar question about todo directives in #203 where the directive itself is injecting an invalid word into the output. In the cases you present, we may need to do something different to allow the content to be ignored. The builder traverses the parsed document (https://github.com/sphinx-contrib/spelling/blob/master/sphinxcontrib/spelling/builder.py#L217) and looks for specific node types to include. Maybe it needs a way to configure node types to ignore as well, so that they are not traversed.

[kondziu]

A literal include directive should cause the content to be ignored, but I haven't actually tested that.

I confirm. I have literalinclude directives, and they are not spellchecked.

Maybe it needs a way to configure node types to ignore as well, so that they are not traversed.

I am not sure what a good solution would be, but from my point of view it'd be easiest to be able to turn the spellchecker off for some section of the content, a la:

.. spelling-stop::

Lorem ipsum dolor sit amet, consectetuer adipiscing elit. Maecenas lorem. Nunc dapibus tortor vel mi dapibus sollicitudin. Nunc tincidunt ante vitae massa. Aliquam erat volutpat. Etiam dictum tincidunt diam. 

.. include:: file.log

Cras pede libero, dapibus nec, pretium sit amet, tempor quis. Sed vel lectus. Donec odio tempus molestie, porttitor ut, iaculis quis, sem. Morbi leo mi, nonummy eget tristique non, rhoncus non leo.

.. spelling-resume::

Another solution that would work for me would be a directive that skips spelling for entire nodes along the lines of :spelling:ignore:

.. spelling-ignore::

    Lorem ipsum dolor sit amet, consectetuer adipiscing elit. Maecenas lorem. Nunc dapibus tortor vel mi dapibus sollicitudin. Nunc tincidunt ante vitae massa. Aliquam erat volutpat. Etiam dictum tincidunt diam. 

.. spelling-ignore::

    .. include:: file.log

.. spelling-ignore::

    Cras pede libero, dapibus nec, pretium sit amet, tempor quis. Sed vel lectus. Donec odio tempus molestie, porttitor ut, iaculis quis, sem. Morbi leo mi, nonummy eget tristique non, rhoncus non leo.

[kondziu]

Sorry for making a mess, I must have misclicked.

[dhellmann]

I did some experimenting in #205, and I think you can use the "only" directive for what you want. Something like

.. only:: html

   .. include:: file.log

[LecrisUT]

Another usecase is for bibliography. In that case another solution would be to ignore all spellchecking in that given directive.

[kondziu]

I did some experimenting in #205, and I think you can use the "only" directive for what you want. Something like

.. only:: html

   .. include:: file.log

Sorry for the geological speed of reply.

This almost works, except there are a bunch of other builders involved. The more general solution probably is:

.. only:: not spelling

   .. include:: file.log

(At least it solves my immediate problem.)