Remove duplicate extract_package_and_build_text call

[kenodegard]

Description

Looking to see just how much of a speedup we get by removing this redundant line.

Extracted from:

• Miscellaneous render speedups #5225
• Add cached _split_line_selector to avoid redundant parsing in select_lines #5237
• Speed up rendering by avoiding costly/redundant function calls #5248

Checklist - did you ...

• Add a file to the news directory (using the template) for the next release's release notes?
• Add / update necessary tests?
• Add / update outdated documentation?

[codspeed-hq]

CodSpeed Performance Report

Merging #5253 will improve performances by 26.85%

_{Comparing kenodegard:remove-duplicate (8c64506) with main (6982cbd)}

Summary

⚡ 1 improvements
✅ 2 untouched benchmarks

Benchmarks breakdown

	Benchmark	main	kenodegard:remove-duplicate	Change
⚡	test_pin_subpackage_benchmark	118.7 s	93.5 s	+26.85%

[kenodegard]

FYI @mbargull

[mbargull]

Thanks, I wondered too and wanted to do the same, just didn't get to it yet :).
I've pushed the other commit I was interested in (somewhat) isolation.

[mbargull]

Benchmark	main	kenodegard:remove-duplicate	Change	modified functions
test_pin_subpackage_benchmark	122.9 s	97 s	+26.64%	build_id
test_pin_subpackage_benchmark	122.9 s	90.9 s	+35.22%	build_id get_output_dicts_from_metadata
test_pin_subpackage_benchmark	122.9 s	83.8 s	+46.52%	build_id get_output_dicts_from_metadata _filter_recipe_text
test_pin_subpackage_benchmark	122.8 s	73.5 s	+67.07%	build_id get_output_dicts_from_metadata _filter_recipe_text _split_line_selector
test_pin_subpackage_benchmark	122.8 s	71.4 s	+72.06%	build_id get_output_dicts_from_metadata _filter_recipe_text _split_line_selector select_lines/selector_cache
test_pin_subpackage_benchmark	122.8 s	73.7 s	+66.55%	build_id get_output_dicts_from_metadata _filter_recipe_text select_lines/namespace hash
test_pin_subpackage_benchmark	122.8 s	71.4 s	+72.07%	build_id get_output_dicts_from_metadata _filter_recipe_text _split_line_selector select_lines/selector_cache regex shortcut
test_pin_subpackage_benchmark	122.8 s	71.3 s	+72.16%	build_id get_output_dicts_from_metadata _filter_recipe_text _split_line_selector select_lines/selector_cache regex shortcut get_used_loop_vars+get_vars
test_pin_subpackage_benchmark	122.8 s	40.4 s	×3	build_id get_output_dicts_from_metadata _filter_recipe_text _split_line_selector select_lines/selector_cache regex shortcut get_used_loop_vars+get_vars Config.copy

[mbargull]

test_pin_subpackage_benchmark	122.9 s	97 s	+26.64%	build_id
test_pin_subpackage_benchmark	122.9 s	90.9 s	+35.22%	build_id get_output_dicts_from_metadata
test_pin_subpackage_benchmark	122.9 s	83.8 s	+46.52%	build_id get_output_dicts_from_metadata _filter_recipe_text

It's nice to see that these innocuous changes would yield considerable improvements already.

test_pin_subpackage_benchmark	122.8 s	73.5 s	+67.07%	build_id get_output_dicts_from_metadata _filter_recipe_text _split_line_selector
test_pin_subpackage_benchmark	122.8 s	71.4 s	+72.06%	build_id get_output_dicts_from_metadata _filter_recipe_text _split_line_selector select_lines/selector_cache
test_pin_subpackage_benchmark	122.8 s	73.7 s	+66.55%	build_id get_output_dicts_from_metadata _filter_recipe_text select_lines/namespace hash

As hoped for in #5237 (comment) , your approach offers at least a much improvements as my more complicated bolted-on version!
It's also nice to see what part contributes the most improvement in the benchmark, i.e., as expected, factoring out _split_line_selector has the most impact.

test_pin_subpackage_benchmark	122.8 s	71.4 s	+72.07%	build_id get_output_dicts_from_metadata _filter_recipe_text _split_line_selector select_lines/selector_cache regex shortcut

This was just to confirm that any impact (if at all) of these micro "only-re.match/re.search-if-needed" optimizations would not be reflected in the current benchmarks.
The test cases carry no script files and the meta.yaml (or lines therein) might be too small for sel_pat.match(line) to matter; plus, due to the memoization the latter might not make much of difference anymore anyway (which would be the desired outcome, of course!).

[kenodegard]

@mbargull thanks for doing this! should we split each of the commits into separate PRs now?

[beckermr]

What can I do to get these changes moving towards being included in the next conda-build release?

[kenodegard]

All of these improvements have been split off into separate PRs for easier review. Reverting this PR to the first commit.