turtlebasket/JobSpy
1
0
Fork 0
mirror of https://github.com/Bunsly/JobSpy.git synced 2026-03-10 23:44:30 -07:00
Code Issues Packages Projects Releases Wiki Activity

FEAT: Optional convertion to annual and know salary source (#170)

Browse Source
This commit is contained in:
Lluís Salord Quetglas
2024-07-18 04:05:33 +02:00
committed by GitHub
parent de70189fa2
commit 2a0cba8c7e
3 changed files with 30 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
src/jobspy/__init__.py
View File

@@ -10,7 +10,7 @@ from .scrapers.indeed import IndeedScraper
from .scrapers.ziprecruiter import ZipRecruiterScraper
from .scrapers.glassdoor import GlassdoorScraper
from .scrapers.linkedin import LinkedInScraper
from .scrapers import ScraperInput, Site, JobResponse, Country
from .scrapers import SalarySource, ScraperInput, Site, JobResponse, Country
from .scrapers.exceptions import (
LinkedInException,
IndeedException,
@@ -36,6 +36,7 @@ def scrape_jobs(
linkedin_company_ids: list[int] | None = None,
offset: int | None = 0,
hours_old: int = None,
enforce_annual_salary: bool = True,
verbose: int = 2,
**kwargs,
) -> pd.DataFrame:
@@ -165,7 +166,8 @@ def scrape_jobs(
job_data["min_amount"] = compensation_obj.get("min_amount")
job_data["max_amount"] = compensation_obj.get("max_amount")
job_data["currency"] = compensation_obj.get("currency", "USD")
if (
job_data["salary_source"] = SalarySource.DIRECT_DATA.value
if enforce_annual_salary and (
job_data["interval"]
and job_data["interval"] != "yearly"
and job_data["min_amount"]
@@ -180,7 +182,9 @@ def scrape_jobs(
job_data["min_amount"],
job_data["max_amount"],
job_data["currency"],
) = extract_salary(job_data["description"])
) = extract_salary(job_data["description"], enforce_annual_salary=enforce_annual_salary)
job_data["salary_source"] = SalarySource.DESCRIPTION.value
job_df = pd.DataFrame([job_data])
jobs_dfs.append(job_df)