lucasvanhaaren / pelican-data-files Goto Github PK

Hey there,
title says it all, maybe I overlooked something, but right now Pelican does not pick up changes in data files.

Before I can do something about this, I need a pointer to Pelican's internals / plugin API where to change that 🕺

Cheers,
S1SYPHOS

It would be convenient to also support sub-directories similar to jekyl (https://jekyllrb.com/docs/datafiles/#subfolders).

For example: Files in the content/data/subdir/ directory would be accessible via the DATA_SUBDIR_ prefix variable in Jinja templates.

Followed all instructions that you have laid out. I am using Pelican 4.7.1 and the error persists. I even created a fresh virtual environment just in case with the same result.

Check if THEME is filled in pelicanconf.py.
If true, THEME value become the default for fetch param.

It avoids mentioning the theme when there is a great chance that it is the one we want.

Provide samples for both themes and standalone websites usages.

ℹ️ For now, orbelican was like a sample but I think it's better to have sample in the same repo to update both at time.

TOML (Tom's Obvious Minimal Language) is a "config file format for humans". It is readable, unambiguous, and has a standardised format. It is also "designed to map unambiguously to a hash table". As such, it is a good choice for creating configuration or data files. Python 3.11 will release a TOML parser in the standard library.

Add some contribution help (CONTRIBUTING.md, developer docs, ...) in order to follow open-source best practices and allow for new comers to be onboard the project smoothly and ease contributing.

• apply community standards
• detailed installation instructions
• detailed pelican generation instructions
• how to build a theme with
• how to contribute in
• how it works under the hood

Hey there,
for my next project I'm thinking about using your plugin - going through its code, I noticed line 75 in generators.py: TODO check for duplicates (eg: profile.json and profile.yaml). I want to contribute to your plugin, and this seems an easy target 😀 In case of a duplicate, would you .. 1. log error and exit or 2. prefer one over the other - and if so, preferring by modified time seems reasonable, no?

What do you think?

Cheers
S1SYPHOS

Refactor tools/fetch script to use pathlib instead of os.path.
Because of the benefits and the readability.

Bootstrap CI-CD with features:

• automated testing (load files from all supported formats)
• automated package release (build and push package to PyPi ⚠️ BUT NEED MANUAL ACTION)

Some features are broken in tasks.py because of the wrong path :

PKG_NAME = "data-files" must become PKG_NAME = "data_files"

And the fact that version tagging is not automated, block twine tasks while uploading new version of package :

• add a bump_version task before twin upload
• OR add automated testing via github workflow and review version tagging

Add YAML file support for pelican-data-files.
Use pyyaml implementation.

This add a repository banner