* develop:
  Fix streamlit#1230: OSError not caught when writing cache to disk (streamlit#8)
  Improve caching code (streamlit#2)
  Remove incorrect background of code inside error messages (streamlit#4)
  Fix line number in error message (streamlit#3)
  Update README.md
  Implement Caching object to support running code blocks only once (streamlit#1)
  Update README.md
  Add font licenses to NOTICE (streamlit#1267)
  Clean up repo for open source release. Remove unecessary folders. (streamlit#1266)
  Fix st.cache (streamlit#1273)
  Close streamlit#1092 (streamlit#1270)
  Change license header in variables.scss so it's compatible with scss-to-json. (streamlit#1268)
  Sidebar (streamlit#1224)
  Add Apache license and update all license headers (streamlit#1265)
  Handling objects with empty docs in st.help (streamlit#1261)
  Enhance Mochawesome report with screenshots (streamlit#1263)
  Add LICENSES file (streamlit#1262)

* Streamlit pink -> red70, normal text -> gray85

* Update font sizes

* Change fonts to Source Sans/Serif/Code Pro

* Add font files to Streamlit

* Update test snapshots

* Introduce fastparquet on the Python side and parquet-wasm on the JS side to bypass the Arrow serialization for DataFrame

* Patch DataEditor not to use PyArrow

* Add setTimeout() so the import of parquet-wasm to work in @stlite/mountable

* Fix comments

* Fix incompatibilities with some column types

* Change logic to handle lists from fastparquet

* Move the decoding above the string parsing

---------

Co-authored-by: lukasmasuch <[email protected]>

…ions

Addresses three suggestions from PR #12807 review:

1. Simplify test cases (suggestion #2)
   - Reduced from 9 parameterized tests to 3 focused tests
   - Removed fragment and expander contexts (regression only in containers)
   - Eliminated parameterization that made test app hard to read
   - Simple, clear structure with named containers

2. Add deterministic width assertions (suggestion #3)
   - stretch: Container width=600, expect ~580-600px (allows padding)
   - content: Figure 6.4in @ 100 DPI = 640px, expect ~620-640px
   - pixel: Explicit width=500, expect ~480-500px
   - More stable than arbitrary >200px threshold

3. Clean up unnecessary comments (suggestion #4)
   - Removed comments that just describe next line
   - Kept meaningful comments explaining test logic and expectations
   - Improved readability

Test structure:
- 3 regression tests in containers only
- Specific figure size (6.4in x 4.8in at 100 DPI)
- Known container width for stretch test
- Deterministic assertions with reasonable margins

Note: Suggestion #1 (use_container_width check) addressed with explanation
in PR comment - current logic is correct for the new widthConfig system.

…ions

Addresses three suggestions from PR #12807 review:

1. Simplify test cases (suggestion #2)
   - Reduced from 9 parameterized tests to 3 focused tests
   - Removed fragment and expander contexts (regression only in containers)
   - Eliminated parameterization that made test app hard to read
   - Simple, clear structure with named containers

2. Add deterministic width assertions (suggestion #3)
   - stretch: Container width=600, expect ~580-600px (allows padding)
   - content: Figure 6.4in @ 100 DPI = 640px, expect ~620-640px
   - pixel: Explicit width=500, expect ~480-500px
   - More stable than arbitrary >200px threshold

3. Clean up unnecessary comments (suggestion #4)
   - Removed comments that just describe next line
   - Kept meaningful comments explaining test logic and expectations
   - Improved readability

Test structure:
- 3 regression tests in containers only
- Specific figure size (6.4in x 4.8in at 100 DPI)
- Known container width for stretch test
- Deterministic assertions with reasonable margins

Note: Suggestion #1 (use_container_width check) addressed with explanation
in PR comment - current logic is correct for the new widthConfig system.

Removed the strings prop from ChatAudioRecorder test that was passing
a prop that doesn't exist in the component interface. This prop was
removed when we fixed the over-engineering in comment #3, but the test
wasn't updated. Test still passes as it searches for hard-coded labels.

Resolves PR #12834 comment #14