[Plugin] DOCXImport

[patrik]

Ok, a short info in the plugin's output window would be nice for these sort of things.

Thanks for checking.

[DiapDealer]

Originally Posted by patrik

Ok, a short info in the plugin's output window would be nice for these sort of things.

That's part of mammoth. You can see the warnings for ignored elements in the Plugin Runner window (at least I do when I ran it on your document).

I think v0.2.4 of the plugin (or 0.2.5 when I get 0.2.4 fixed) will more clearly indicate that images were skipped.

[DiapDealer]

I have a potential solution for converting wmf/emf images to png when importing with this plugin. Unfortunately, it's not possible to so in a completely self-contained (not to mention cross-platform) manner, so there are external prerequisites for it working. I honestly think encountering these types of Windows images is going to fairly rare, so I'm not going look for, or work on, a more comprehensive solution.

If you have LibreOffice and ImageMagick installed and on your PATH (or manually configured in the plugin's json file) it should work. I've tested it on Linux and Windows, and I assume it will work on Macs (but the configuration may have to be tweaked).

Anyone with those programs installed (or is willing to install them) who is willing to do a little beta-testing?

[cmcclaflinprcvi]

I'm not sure if I'm in the right place, apologies if I'm not. My organization remediates text based documents to create epub files for students with disabilities. We're now using DOCXImport to import word docs into Sigil.

When an image that does not have alt-text is imported there is no alt attribute in the resulting <img> tag in Sigil. Unfortunately when many screen readers (text to speech engines used by people with low or no vision) encounter an image with no alt-text they will read out the file name.

This issue can be worked around by adding alt="" to the image tag. We would like to propose that DOCXImport be modified to add this empty alt attribute to images from word that do not contain alt-text in order to improve the accessibility of the files produced.

Here is a simple 'citation' backing up this logic: https://davidwalsh.name/accessibilit...alt-attributes

Thank you to everyone who works on this project!

[KevinH]

In the interim, you may want to check out the Access Aide Accessibility Helper plugin for Sigil. It will allow easy access and changes to image alt tags.

Originally Posted by cmcclaflinprcvi

I'm not sure if I'm in the right place, apologies if I'm not. My organization remediates text based documents to create epub files for students with disabilities. We're now using DOCXImport to import word docs into Sigil.

When an image that does not have alt-text is imported there is no alt attribute in the resulting <img> tag in Sigil. Unfortunately when many screen readers (text to speech engines used by people with low or no vision) encounter an image with no alt-text they will read out the file name.

This issue can be worked around by adding alt="" to the image tag. We would like to propose that DOCXImport be modified to add this empty alt attribute to images from word that do not contain alt-text in order to improve the accessibility of the files produced.

Here is a simple 'citation' backing up this logic: https://davidwalsh.name/accessibilit...alt-attributes

Thank you to everyone who works on this project!

[DiapDealer]

Originally Posted by cmcclaflinprcvi

This issue can be worked around by adding alt="" to the image tag. We would like to propose that DOCXImport be modified to add this empty alt attribute to images from word that do not contain alt-text in order to improve the accessibility of the files produced.

The underlying Mammoth Python code does much of the heavy lifting for this plugin. I've merely created a Sigil plugin "wrapper" around it. I've made a few tweaks to the Mammoth codebase, but I hesitate to get my version of it too out of sync with the original. It makes it too hard to update to newer versions of Mammoth as they come out.

I'll certainly look into it, but it depends on how much rewiring of someone else's code I would need to do (and how hard it would make the upgrade process). It may actually be simpler to create a new plugin that would add empty alt attributes to img tags without any. Or if Kevin's Accessibility Helper plugin can be used to fix the tags, that would work too.

You could also try reporting this upstream at Mammoth. Perhaps he'd be willing to modify his code to add an empty alt attribute to img tags generated from images without alt-text/descriptions. That would be the least complicated way for the change to make it into this plugin

Originally Posted by cmcclaflinprcvi

Thank you to everyone who works on this project!

You mean me, myself, and I?

[DiapDealer]

Actually ... never mind my last post. I just realized there's a quick and painless way I can make this happen without changing Mammoth at all. I can take care of it in my custom ImageWriter routine.

[DiapDealer]

Originally Posted by cmcclaflinprcvi

We would like to propose that DOCXImport be modified to add this empty alt attribute to images from word that do not contain alt-text in order to improve the accessibility of the files produced.

Give this test version a whirl. If everything is OK, I'll release a new version with this change included,

[Tex2002ans]

Originally Posted by DiapDealer

You mean me, myself, and I?

Yeah, all 3 of those amazing people.

[DiapDealer]

Originally Posted by Tex2002ans

Yeah, all 3 of those amazing people.

To be fair, I should include the guy who maintains Mammoth. So make it an even 4.

[cmcclaflinprcvi]

Originally Posted by DiapDealer

Give this test version a whirl. If everything is OK, I'll release a new version with this change included,

We tested images with and without alt text, both cases behaved exactly as expected. When there was alt text in word it still came through just fine, and when there was not alt="" was now added. Thank you so much!

For what it's worth, I thought I'd let you know a little bit about the positive impact your plugin is having. Our organization helps remediate inaccessible print documents into accessible digital versions for students with perceptual disabilities. Thanks to your plugin we've now got an easy way to create epub files for students with access needs across the province of British Columbia. Thank you so much for creating and maintaining this plugin, it's so appreciated.

[DiapDealer]

Originally Posted by cmcclaflinprcvi

We tested images with and without alt text, both cases behaved exactly as expected. When there was alt text in word it still came through just fine, and when there was not alt="" was now added. Thank you so much!

You're welcome. Thanks for reporting back. I'll get the first post updated with the new version in the near future.

Originally Posted by cmcclaflinprcvi

For what it's worth, I thought I'd let you know a little bit about the positive impact your plugin is having. Our organization helps remediate inaccessible print documents into accessible digital versions for students with perceptual disabilities. Thanks to your plugin we've now got an easy way to create epub files for students with access needs across the province of British Columbia. Thank you so much for creating and maintaining this plugin, it's so appreciated.

Happy it's doing some good. That's always nice to hear! Keep in mind that this plugin is one of several ways to get docx documents into an epub. The one that sticks out most in my mind is Toxaris' Word add-in. It's very thorough and configurable. Do yourself a favor and check it out. You should also check out the plugin KevinH mentioned upthread that strives to make sure epubs meet ACE requirements.

[DiapDealer]

Updated first post to v0.2.5.

- Add empty alt attributes with images that have no alt_text/description in the DOCX

[JeffPalley]

Hi,

I have the bad feeling that I don't belong here … I just followed a Google search for "DOCX to EPUB converters" and it suggested this site. If there is some other place for people who know nothing, please let me know.

Anyway, once I got here, I downloaded and installed the DOCXImport plugin for Sigil. Problem is that I have no idea how to get it to work. If I open it as a plugin, it just pops up a window that tells me "Launcher: missing or incorrect paths passed in."

I've tried the various (but obviously insufficient) ways I know of to ask Sigil to open the SAMPLE.DOCX file … pretty much just File>Open and File>Add. But the best that happens is that Sigil imports it as an attached DOCX file. I assume there's some obvious way I'm missing, since DiapDealer (among others) was Immediately enthusiastic about this plugin. If someone would explain to me how to use it (5th grade level), I'd be very grateful. Or if I'm in the wrong place, please tell me where to go and I won't bother you again.
Thanks, -Jeff

[DiapDealer]

You using an official Sigil install? If you can install a Sigil plugin, you should have a recent enough version of Sigil to run the plugin.

"Launcher" suggests that you might be on a Mac, but I have no idea why you'd see such a message from Sigil.

Originally Posted by JeffPalley

If I open it as a plugin, it just pops up a window that tells me "Launcher: missing or incorrect paths passed in."

"Open it as plugin" has me concerned. Do you mean from within Sigil? As in (from the Sigil menu): Plugins->Input->DOCXImport? That's how you launch the plugin. That will in turn launch a dialog to let you specify the input document and the various plugin settings. In Sigil, plugins are separate beasts altogether. They don't alter Sigil's GUI or standard feature set. I think you're expecting the plugin to alter Sigil's File->Open to work with docx documents and that's just not how Sigil plugins work.

Even if you get it working, you may be disappointed. This probably isn't the turn-key DOCX to EPUB conversion solution you're hoping for.