Makes prefetcher none cache default

[ankitaluthra1]

No description provided.

[gemini-code-assist]

Code Review

This pull request enables the adaptive concurrent prefetcher by default, updating the default concurrency from 1 to 4 and dynamically setting the default cache type to "none" when prefetching is active (or "readahead" when disabled). The documentation and test suites have been updated to reflect these new defaults. Feedback on the changes suggests handling potential string values for the prefetching kwarg to avoid incorrect boolean evaluation, updating ZonalFile to align with the new dynamic cache defaults, and wrapping the environment variable integer parsing in a try-except block to prevent import-time crashes.

[ankitaluthra1]

/gemini review

[gemini-code-assist]

Code Review

This pull request enables adaptive concurrent prefetching by default, changing the default concurrency from 1 to 4 and setting the default cache type to "none" when prefetching is active. It also updates the documentation and test suite to reflect these new defaults. Regarding the feedback, a discrepancy was identified in how use_experimental_adaptive_prefetching is resolved in GCSFile.__init__ compared to ZonalFile.__init__. Specifically, passing None explicitly in kwargs would evaluate to False and disable prefetching while keeping cache_type="none", resulting in no caching or prefetching at all. A code suggestion has been provided to resolve this consistently.

[gemini-code-assist]

There is a discrepancy in how use_experimental_adaptive_prefetching is resolved when explicitly passed as None in kwargs. In ZonalFile.__init__, an explicit None falls back to the environment variable default. However, in GCSFile.__init__, checking "use_experimental_adaptive_prefetching" in kwargs evaluates to True even if the value is None, which then evaluates bool(None) to False. This disables the prefetcher while keeping cache_type="none", leading to no caching or prefetching at all. Using kwargs.get() to handle None consistently resolves this issue.

Suggested change

	if "use_experimental_adaptive_prefetching" in kwargs:
	val = kwargs["use_experimental_adaptive_prefetching"]
	use_prefetch_reader = (
	val.lower() in ("true", "1") if isinstance(val, str) else bool(val)
	)
	else:
	use_prefetch_reader = os.environ.get(
	"USE_EXPERIMENTAL_ADAPTIVE_PREFETCHING", "true"
	).lower() in (
	"true",
	"1",
	)
	val = kwargs.get("use_experimental_adaptive_prefetching")
	if val is not None:
	use_prefetch_reader = (
	val.lower() in ("true", "1") if isinstance(val, str) else bool(val)
	)
	else:
	use_prefetch_reader = os.environ.get(
	"USE_EXPERIMENTAL_ADAPTIVE_PREFETCHING", "true"
	).lower() in (
	"true",
	"1",
	)

[codecov]

Codecov Report

❌ Patch coverage is 82.14286% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.98%. Comparing base (370ca0b) to head (d2f33f2).

Files with missing lines	Patch %	Lines
gcsfs/prefetcher.py	72.72%	3 Missing ⚠️
gcsfs/zb_hns_utils.py	50.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #992      +/-   ##
==========================================
+ Coverage   89.68%   89.98%   +0.29%     
==========================================
  Files          16       16              
  Lines        3579     3604      +25     
==========================================
+ Hits         3210     3243      +33     
+ Misses        369      361       -8     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.