TextArea: make sure to use \n as line separator

[UlrichGoebel]

Hi,

Short version:

How to make sure, that TextArea uses \n as line separator, as linux normaly does, regardless of the text loeded and even if no text is loadet?

Long version:

I read in the doc that the TextArea Widget decides from the given content which line separator to use. As I understand, the line separator is one of \n, \r, \n\r, \r\n, and this behaviour should make shure, that the text follows the same convention after as before beeing edited.

I am building a UI for existing data in a database. Other UIs may have used different line separators, and I don't have control over that. I know that I have to solve this at database level. Nevertheless I want to make sure, that TextArea will use \n as line seperator.

Should I scan the text before loading it into TextArea and replace \r, \n\r, \r\n by \n? Or can I manipulate the TextArea.document.newline attribute? Or should I scan/replace after the text has beeing edited?

And which line separator TextArea will use if no text is initially loadet?

Any hint is welcome.

Regards
Ulrich

[TomJGooding]

It sounds like you already have a pretty good understanding of the TextArea widget from the docs.

When content is loaded into the TextArea, the lines of the document are stored separate from the newline character, which is based on the "first occurrence of a line separator". Here's the underlying code (though note only intended for internal use from the leading underscore) :

textual/src/textual/document/_document.py

Lines 203 to 210 in 4adaf03

	class Document(DocumentBase):
	"""A document which can be opened in a TextArea."""
	def __init__(self, text: str) -> None:
	self._newline = _detect_newline_style(text)
	"""The type of newline used in the text."""
	self._lines: list[str] = text.splitlines(keepends=False)
	"""The lines of the document, excluding newline characters.

You can't change the TextArea.document.newline since this is only a property, so I think the best method is to replace any unwanted line separators (at least the first occurance!) before loading into the TextArea.

Since the TextArea doesn't support mixed line endings as all will be converted to this newline, you shouldn't need to worry about converting after this content has been edited.

[UlrichGoebel]

Thank's, that makes it clear.