I’m building a site that’s almost entirely generated by content adapters (seriously, these are one of Hugo’s best features). One section has stopped building the pages, although the list page shows all the page titles I would expect to see, with correct links …that go to 404s. (Some of these pages have non-Latin characters in the slugs, which I know is a problem, but that’s for later.)
These pages *were* building correctly earlier, and I can’t figure out why they aren’t now (I’ve checked the draft status and tried running hugo server --cleanDestinationDir --printPathWarnings --logLevel debug, which didn’t give me any answers other than a taxonomy path collision, which I don’t think is related). There’s another section, built from a content generator using the same data, that is building just fine, although it’s using less complicated logic.
Repo is at https://github.com/Brown-University-Library/nq-proto.
My apologies, I thought it was a public repo already – fixed.
In themes/brown-hugo-starter/layouts/pseudonyms/page.html you’ve wrapped everything in a define block.
Other comments:
-
In hugo.toml you have disableKinds = ['taxonomy'] beneath the taxonomies table. It needs to be in the root table if you wish to disable generation of taxonomy pages.
-
In content/people/_content.gotmpl you have defined kind, draft, and description within the params map, but these are top-level front matter keys.
-
In content/pseudonyms/_content.gotmpl you have defined kind and draft within the params map, but these are top-level front matter keys.
Amazing what fresh eyes on your code will see. Thanks, it looks like that did the trick!
You can also remove --cleanDestinationDir from your GitHub Pages workflow. The public directory is not under source control; it is created each time the workflow is run.
And even if the public directory were under source control, that CLI flag is only effective when a static directory exists.
