fix(ansi): handle CRLF line endings in Text.from_ansi

[Alm0stSurely]

Problem

Fixes #4090.

When Text.from_ansi is given input with CRLF (\r\n) line endings, all text content is lost and only empty lines remain.

Reproduction:

from rich.text import Text
Text.from_ansi("Hello\r\nWorld\r\n").plain
# Actual: '\n\n'
# Expected: 'Hello\nWorld\n'

Analysis

The root cause is an interaction between line splitting and carriage-return handling:

1. AnsiDecoder.decode() splits on \n using re.split(r"(?<=\n)", ...), keeping the \n attached to the preceding line.
2. After stripping \n, a line that originally ended with \r\n becomes ...\r.
3. decode_line() then calls line.rsplit("\r", 1)[-1] to simulate terminal carriage-return behavior (where \r moves the cursor to the start of the line, effectively overwriting prior content).
4. For a line ending in \r (from CRLF), rsplit("\r", 1)[-1] returns the empty string after the final \r, discarding all content.

Solution

Normalize \r\n to \n in decode() before splitting. This treats CRLF as a plain line-ending sequence rather than a terminal carriage-return that overwrites content.

The fix is a single line:

terminal_text = terminal_text.replace("\r\n", "\n")

This preserves the existing behavior of standalone \r (used in progress bars and terminal updates) while correctly handling CRLF line endings common on Windows and in network protocols.

Benchmarks / Verification

• All 957 existing tests pass (25 skipped).
• Added test_decode_crlf covering:
  • Simple CRLF: "Hello\r\nWorld\r\n"
  • Leading CRLF: "\r\nHello\r\n"
  • Consecutive CRLF: "Hello\r\n\r\nWorld"
  • Mixed line endings: "Hello\nWorld\r\n" and "Hello\r\nWorld\n"

Notes

• This is a minimal, localized change with no API surface changes.
• Standalone \r (without following \n) continues to work as before, since the replace only targets the \r\n sequence.

[willmcgugan]

Your PR has been closed due to a AI policy violation.

Please read the following before submitting further PRs.

https://github.com/Textualize/textual/blob/main/AI_POLICY.md