Drop support for Ruby 2.3 and older #31
Conversation
|
I'd also look for |
|
@ekohl fixed it |
Ruby 2.3 is dead since a long time. This PR drops support for Ruby 2.3 and older.
There was a problem hiding this comment.
I'm a bit unsure about %i[]. Personally I don't like multiple ways of writing the same thing and I don't write enough Ruby to remember %i[arrow_alignment hard_tabs] should be read as [:arrow_alignment, :hard_tabs]. Maybe this is my Python mind, but I like it when there's one way to write things.
| @@ -68,10 +68,10 @@ def insert(index, token) | |||
| token.next_token = current_token.next_token | |||
| token.prev_token = current_token | |||
|
|
|||
| current_token.next_token.prev_token = token unless current_token.next_token.nil? | |||
| current_token.next_token&.prev_token = token | |||
There was a problem hiding this comment.
I thought the safe navigator was Ruby 2.5. Am I messing up versions in my head?
There was a problem hiding this comment.
I did the safe autofix with ruby 2.4 as target version
| @@ -103,7 +103,7 @@ def fix(problem) | |||
| problem[:token].prev_code_token.prev_token.value = ' ' * problem[:newline_indent] | |||
|
|
|||
| end_param_idx = tokens.index(problem[:token].prev_code_token) | |||
| start_param_idx = tokens.index(problem[:token].prev_token_of([:INDENT, :NEWLINE])) | |||
| start_param_idx = tokens.index(problem[:token].prev_token_of(%i[INDENT NEWLINE])) | |||
| param_length = tokens[start_param_idx..end_param_idx].map { |r| r.to_manifest.length }.reduce(0) { |sum, x| sum + x } + 1 | |||
There was a problem hiding this comment.
This is completely unrelated, but I'm just looking at this and this is a bit hard to read. I really wonder if this isn't the same as:
| param_length = tokens[start_param_idx..end_param_idx].map { |r| r.to_manifest.length }.reduce(0) { |sum, x| sum + x } + 1 | |
| param_length = tokens[start_param_idx..end_param_idx].sum { |r| r.to_manifest.length } + 1 |
There was a problem hiding this comment.
looks ok to me, but I don't know enough about ruby to ensure this works. could be handled with a test.
There was a problem hiding this comment.
I checked and Enumable#sum was introduced in Ruby 2.4 so my suggestion is only valid after this PR's changes.
Co-authored-by: Ewoud Kohl van Wijngaarden <ewoud@kohlvanwijngaarden.nl>
|
@bastelfreak Appreciate this PR has been sat for a while now but I wonder if it's worth updating it to bump the min Puppet version to 2.5? |
|
See #50 for an updated version of this PR. |
This commit fixes #31. It appears that the `resource_indexes` method cannot differentiate between a resource declaration and a defaults declaration. If we compare the two, you can see that a defaults declaration type starts with a capital letter whereas a resource declaration does not. ```puppet Service { 'foo': ensure => running, } ``` ```puppet service { 'foo': ensure => running, } ``` When the `resource_indexes` method runs, it initially selects tokens by the presence of a colon then works to the right to decide if there are any violations. That means we are actually only starting to collect information about the declaration after the name.. so we never know the actual type of the declaration. ```puppet Service { 'foo': ↑ (here) ensure => running, } ``` https://github.com/puppetlabs/puppet-lint/blob/main/lib/puppet-lint/data.rb#L167 In contrast, `defaults_indexes` will skip any tokens that do not have a type of `:CLASSREF` because we know that a defaults declaration will always start with a capital letter (See [puppet-lint.com](http://puppet-lint.com/developer/tokens/) for more information on token types). This commit fixes the above by ensuring that `resource_indexes` checks the type of resource that it is analysing and skips the itteration if it detects a `:CLASSREF`.
No description provided.