improved detection/exclusion of spurious matplotlib plain text output

Author: jjallaire

news/changelog-1.4.md

@@ -109,6 +109,7 @@
 - ([#6344](https://github.com/quarto-dev/quarto-cli/issues/6344)): Somewhat improve the error message in case of YAML parsing errors in metadata of Python code cells.
 - ([#6367](https://github.com/quarto-dev/quarto-cli/issues/6367)): Fix bug with nested code cells in the generation of Jupyter notebook from .qmd files.
 - ([#6393](https://github.com/quarto-dev/quarto-cli/pull/6393)): Search `JULIA_HOME` for Julia-specific Python installations.
+- Improved detection/exclusion of spurious matplotlib plain text output
 
 ## OJS engine
 

src/core/jupyter/jupyter.ts

@@ -1617,6 +1617,7 @@ function isDiscadableTextExecuteResult(output: JupyterOutput) {
       if (textPlain && textPlain.length) {
         return [
           "[<matplotlib",
+          "<matplotlib",
           "<seaborn.",
           "<ggplot:",
         ].some((startsWith) => textPlain[0].startsWith(startsWith));