Nested numbered lists in Markdown cause an error

[zstadler]

The following Markdown code:

1. Create a new context that combined the root context with `tagValueProducer`
2. Using `tagValueProducer`, find all layer's `ConfiguredFeature`s that match
3. For each layer feature
   1. Create a post-match context
   2. Execute `processFeature` for the feature using the post-match context

which seems to be well accepted by GitHub:

1. Create a new context that combined the root context with tagValueProducer
2. Using tagValueProducer, find all layer's ConfiguredFeatures that match
3. For each layer feature
   1. Create a post-match context
   2. Execute processFeature for the feature using the post-match context

Causes the following error:

[ERROR]     .../README.md
[ERROR]         @@ -41,8 +41,8 @@
[ERROR]          1.·Create·a·new·context·that·combined·the·root·context·with·`tagValueProducer`
[ERROR]          2.·Using·`tagValueProducer`,·find·all·layer's·`ConfiguredFeature`s·that·match
[ERROR]          3.·For·each·layer·feature
[ERROR]         -···1.·Create·a·post-match·context
[ERROR]         -···2.·Execute·`processFeature`·for·the·feature·using·the·post-match·context
[ERROR]         +1.·Create·a·post-match·context
[ERROR]         +2.·Execute·`processFeature`·for·the·feature·using·the·post-match·context

[nedtwigg]

we'd need to see your spotless configuration. I bet you've got an indentWithTabs rule, which defaults to 4 spaces I think, and the since there's a three-space indent above it's rounding down to 0. Just a guess.

[zstadler]

This is not "my" spotless configuration. I'm preparing a PR to the planetiler repository and the Spotless Maven plugin was run as part of the Maven checks.

There is no mention of an indentWithTabs string in that repository.

As far as I understand, and I'm not a Java/Maven expert, this is the section of the Maven configuration that's relevant to this issue:

          <markdown>
            <includes>
              <include>**/*.md</include>
            </includes>
            <excludes>
              <exclude>**/target/**</exclude>
              <exclude>planetiler-openmaptiles/**</exclude>
            </excludes>
            <flexmark/>
          </markdown>

I'd appreciate your help in changing the above to allow 3-space-char indentations in Markdown files.

[nedtwigg]

thanks to the configuration block you posted above, we can see that you are using the flexmark formatter. i went to that repo, searched indent, and it appears that it enforces a 4-space indent

https://github.com/vsch/flexmark-java/blob/bcfe84a3ab6d23d04adce3e5a0bae45c6b791d14/MarkdownProcessorsEmulation.md?plain=1#L75-L79

your options are to either indent with 4 spaces, or

• find the option in flexmark which turns it off (if it exists)
• pipe it through this https://github.com/diffplug/spotless/blob/main/plugin-maven/src/main/java/com/diffplug/spotless/maven/markdown/Flexmark.java#L25-L35
• and this https://github.com/diffplug/spotless/blob/main/lib/src/main/java/com/diffplug/spotless/markdown/FlexmarkStep.java

[zstadler]

Apparently, the solution to the problem is not within the above Markdown snippet.

While Github accepts

Given a source feature
1. Create a new context that combined the root context with `tagValueProducer`
2. Using `tagValueProducer`, find all layer's `ConfiguredFeature`s that match
3. For each layer feature
   1. Create a post-match context
   2. Execute `processFeature` for the feature using the post-match context

Flexmark requires an empty line before the nested list

Given a source feature

1. Create a new context that combined the root context with `tagValueProducer`
2. Using `tagValueProducer`, find all layer's `ConfiguredFeature`s that match
3. For each layer feature
   1. Create a post-match context
   2. Execute `processFeature` for the feature using the post-match context

I'm sorry for wasting your time.

[linarkou]

@zstadler had the same problem, you saved my time, thank you!