fix: escape Rich markup in node strings and use rich_format for catal...

[Br1an67]

Fixes #5199

Description

This PR fixes two issues with Rich logging integration that were causing unexpected behavior in console and file logs:

1. Console log issue: Brackets around node inputs/outputs (e.g., [companies]) were being swallowed by RichHandler because they were interpreted as Rich markup tags.

2. File log issue: Rich markup tags like [dark_orange]X_train[/dark_orange] were leaking into file logs because the markup was being added directly to the log message, which was then sent to all handlers including non-Rich file handlers.

Development notes

Changes made:

1. kedro/pipeline/node.py: Escape brackets in Node.__str__ using \[ and \] so that RichHandler renders them as literal brackets instead of interpreting them as markup.

2. kedro/io/data_catalog.py:

   • Use the rich_format extra attribute instead of manually adding markup to log messages
   • Remove the _use_rich_markup attribute and unused imports (_format_rich, _has_rich_handler)
   • The rich_format mechanism is handled by Kedro's custom RichHandler, which formats only the specified arguments with colors, while non-Rich handlers receive the unformatted message

3. Tests updated:

   • tests/pipeline/test_node.py: Updated to expect escaped brackets in node string representation
   • tests/framework/project/test_logging.py: Updated test to reflect the new approach where markup is never in the message itself

How it works:

• When Node.__str__ returns identity(\[in\]) -> \[out\], RichHandler with markup=True renders this as identity([in]) -> [out] (correct display)
• When DataCatalog logs use extra={"rich_format": ["dark_orange"]}, only RichHandler applies the color formatting, while file handlers receive the plain dataset name

Checklist

• Read the contributing guidelines
• Signed off each commit with a Developer Certificate of Origin (DCO)
• Opened this PR as a 'Draft Pull Request' if it is work-in-progress
• Updated the documentation to reflect the code changes
• Added a description of this change in the RELEASE.md file
• Added tests to cover my changes
• Checked if this change will affect Kedro-Viz, and if so, communicated that with the Viz team

 kedro/io/data_catalog.py                | 11 ++++-------
 kedro/pipeline/node.py                  |  2 +-
 tests/framework/project/test_logging.py |  5 +++--
 tests/pipeline/test_node.py             | 22 +++++++++++-----------
 4 files changed, 19 insertions(+), 21 deletions(-)

[ElenaKhaustova]

Thank you @Br1an67 for your contribution!

I made a first pass, and the DataCatalog fix looks good to me.

However, the Node fix has a fundamental design flaw (see the comment), and some tests need to be updated as now they are causing CI failures: for example, tests/pipeline/test_pipeline.py test_full should fail as not updated for escaped brackets.

[ElenaKhaustova]

I think this might leak Rich syntax into all non-Rich consumers.

Node.__str__ is a general-purpose method used in many non-Rich contexts:

• File log handlers — Node.run() at line 531 logs "Running node: %s", str(self). Non-Rich file handlers will write Running node: identity(\[in]) -> \[out] (with literal backslashes) to log files.

• Pipeline.describe(names_only=False) at pipeline.py:510 — returns a plain string that includes str(node). Users and non-Rich log consumers will see escaped brackets.

• Any user code calling str(node) — debugging, notebooks, error messages, etc.

Here we should probably follow the same approach as with DataCatalog: keep __str__ returning [in] and handle Rich escaping at the log call site.

[noklam]

#4496

Is this a regression from the original fix? Or is this a new issue?