fix: pdf parsing, pass as str instead of posix path

[erichare]

This Pull Request updates our PDF parser to fix an issue with PosixPath's needing to be converted to str

[Cristhianzl]

lgtm

[ogabrielluiz]

Thanks, @erichare !

LGTM

[codspeed-hq]

CodSpeed Performance Report

Merging #4448 will create unknown performance changes

_{Comparing bugfix-pdf-parsing (187e381) with main (9827463)}

Summary

🆕 15 new benchmarks

Benchmarks breakdown

	Benchmark	main	bugfix-pdf-parsing	Change
🆕	test_create_starter_projects	N/A	401.9 ms	N/A
🆕	test_get_and_cache_all_types_dict	N/A	1 ms	N/A
🆕	test_initialize_services	N/A	139.7 ms	N/A
🆕	test_initialize_super_user	N/A	142.6 ms	N/A
🆕	test_load_flows	N/A	1.3 ms	N/A
🆕	test_setup_llm_caching	N/A	1.6 ms	N/A
🆕	test_build_flow	N/A	691.4 ms	N/A
🆕	test_build_flow_from_request_data	N/A	696.6 ms	N/A
🆕	test_get_all	N/A	515.4 ms	N/A
🆕	test_invalid_run_with_input_type_chat	N/A	16.6 ms	N/A
🆕	test_starter_projects	N/A	1.2 s	N/A
🆕	test_successful_run_with_input_type_any	N/A	494.7 ms	N/A
🆕	test_successful_run_with_input_type_text	N/A	450.3 ms	N/A
🆕	test_successful_run_with_output_type_any	N/A	452.1 ms	N/A
🆕	test_successful_run_with_output_type_debug	N/A	407.1 ms	N/A