Fix indentation for raw string newlines #625
Conversation
|
Codecov ReportAll modified and coverable lines are covered by tests ✅
❗ Your organization needs to install the Codecov GitHub app to enable full functionality. Additional details and impacted files@@ Coverage Diff @@
## master #625 +/- ##
=======================================
Coverage 77.36% 77.36%
=======================================
Files 22 22
Lines 7885 7887 +2
=======================================
+ Hits 6100 6102 +2
Misses 1368 1368
Partials 417 417 |
| @@ -28,8 +28,6 @@ const ( | |||
| type Encoder struct { | |||
| writer io.Writer | |||
| opts []EncodeOption | |||
| indent int | |||
There was a problem hiding this comment.
The Encoder previously had two fields, indent and indentNum, but indentNum was not being used anywhere. Since the name indentNum is used elsewhere, I migrated the functionality of the indent field to the indentNum field and removed the indent field.
| if e.indentSequence { | ||
| e.column -= e.indentNum | ||
| } | ||
| }() |
There was a problem hiding this comment.
Minor refactoring to ensure the subtraction 🙏
| @@ -816,11 +816,12 @@ func (n *StringNode) String() string { | |||
| // It works mostly, but inconsistencies occur if line break characters are mixed. | |||
| header := token.LiteralBlockHeader(n.Value) | |||
| space := strings.Repeat(" ", n.Token.Position.Column-1) | |||
| indent := strings.Repeat(" ", n.Token.Position.IndentNum) | |||
There was a problem hiding this comment.
The token's IndentNum represents the number of spaces from the beginning of the line. In other words, this is not a value that represents "how many spaces are used for indentation".
Therefore, the behavior is different between the case where the Go value is directly converted to a string using the encoder, and the case where the string is created using the ast obtained by parsing.
However, when encoding the value of Go directly, the context literal is missing, so the AST that is always created becomes a StringNode, whereas when parsing, it becomes a LiteralNode. For this reason, treating the value of IndentNum in this way works well for StringNode.
This behavior may cause confusion, so it may be necessary to refactor it in the future.
|
LGTM 👍 |
Closes #292
Before submitting your PR, please confirm the following.