Glossary / Abbreviations adds 10+ seconds to reload

[Lasikiewicz]

Following this

https://squidfunk.github.io/mkdocs-material/reference/tooltips/#adding-a-glossary

I have added an abbreviations file.

Now when making any changes to the site and saving in VScode the building documentation has gone from 3-4 seconds to 15-20+

Is this normal behaviour or have I configured something incorrectly?

For reference the abbreviations.md does contain 187 entries.

[kamilkrzyskow]

Hello @Lasikiewicz,
it all depends on the whole picture, because adding snippets pre-processes each markdown content to include the snippets, and then abbr also does a "double pass" to first extract the abbreviations from the bottom and then inject them in the contents. Also each Markdown page is a separate Markdown instance in the code, so I believe that for every file the abbreviations are being processed anew without cache, so the task length will increase the more data it has to process 🤔

So the bottleneck could be in either:

• the amount of your files in the docs
• snippets processing
• contents of your snippets*
• abbr processing

You could try to "disable" each part of the chain and see how much the performance improves:

• you're using git, right 👀? So remove some files temporarily
• instead of snippets, which scans the file, use a hook to append the glossary.md file in the on_page_markdown event
• decrease the snippets content length*
• disable the abbr extension and see if you gain a lot of time

You could try to remove the snippets from the picture and only use the glossary option of the abbr extension:

• https://python-markdown.github.io/extensions/abbreviations/#usage
  🤔 but adding 187 entries to the mkdocs.yml file seems like a lot

*In a different discussion one user claimed that adding empty lines between the entries increased the speed of snippets processing:

• Improve performance and authoring experience of mkdocs serve mkdocs/mkdocs#3695 (comment)

Side-note: Python and text processing isn't the best combination, so 10 seconds isn't as bad as it gets 😄

[Lasikiewicz]

Hello @kamilkrzyskow
Thank you for the in depth answer.

The fix was to add a space between each abbreviation in the list.

Here is an example

*[CE]: Consumer Electronics  

*[CI]: Carry In  

*[CMI]: Customer Monitoring Index  

*[CRRR]: Customer Repeat Repair Ratio

Saving documents now takes 3-5 seconds from the 15-20+

Thank you