mirror of https://github.com/Bunsly/JobSpy
add long scrape example (#81)
parent
a7ad616567
commit
89a5264391
|
@ -2,12 +2,11 @@ from jobspy import scrape_jobs
|
|||
import pandas as pd
|
||||
|
||||
jobs: pd.DataFrame = scrape_jobs(
|
||||
site_name=["indeed", "linkedin", "zip_recruiter"],
|
||||
site_name=["indeed", "linkedin", "zip_recruiter", "glassdoor"],
|
||||
search_term="software engineer",
|
||||
location="Dallas, TX",
|
||||
results_wanted=50, # be wary the higher it is, the more likey you'll get blocked (rotating proxy should work tho)
|
||||
results_wanted=25, # be wary the higher it is, the more likey you'll get blocked (rotating proxy can help tho)
|
||||
country_indeed="USA",
|
||||
offset=25 # start jobs from an offset (use if search failed and want to continue)
|
||||
# proxy="http://jobspy:5a4vpWtj8EeJ2hoYzk@ca.smartproxy.com:20001",
|
||||
)
|
||||
|
|
@ -0,0 +1,77 @@
|
|||
from jobspy import scrape_jobs
|
||||
import pandas as pd
|
||||
import os
|
||||
import time
|
||||
|
||||
# creates csv a new filename if the jobs.csv already exists.
|
||||
csv_filename = "jobs.csv"
|
||||
counter = 1
|
||||
while os.path.exists(csv_filename):
|
||||
csv_filename = f"jobs_{counter}.csv"
|
||||
counter += 1
|
||||
|
||||
# results wanted and offset
|
||||
results_wanted = 1000
|
||||
offset = 0
|
||||
|
||||
all_jobs = []
|
||||
|
||||
# max retries
|
||||
max_retries = 3
|
||||
|
||||
# nuumber of results at each iteration
|
||||
results_in_each_iteration = 30
|
||||
|
||||
while len(all_jobs) < results_wanted:
|
||||
retry_count = 0
|
||||
while retry_count < max_retries:
|
||||
print("Doing from", offset, "to", offset + results_in_each_iteration, "jobs")
|
||||
try:
|
||||
jobs = scrape_jobs(
|
||||
site_name=["indeed"],
|
||||
search_term="software engineer",
|
||||
# New York, NY
|
||||
# Dallas, TX
|
||||
|
||||
# Los Angeles, CA
|
||||
location="Los Angeles, CA",
|
||||
results_wanted=min(results_in_each_iteration, results_wanted - len(all_jobs)),
|
||||
country_indeed="USA",
|
||||
offset=offset,
|
||||
# proxy="http://jobspy:5a4vpWtj8EeJ2hoYzk@ca.smartproxy.com:20001",
|
||||
)
|
||||
|
||||
# Add the scraped jobs to the list
|
||||
all_jobs.extend(jobs.to_dict('records'))
|
||||
|
||||
# Increment the offset for the next page of results
|
||||
offset += results_in_each_iteration
|
||||
|
||||
# Add a delay to avoid rate limiting (you can adjust the delay time as needed)
|
||||
print(f"Scraped {len(all_jobs)} jobs")
|
||||
print("Sleeping secs", 100 * (retry_count + 1))
|
||||
time.sleep(100 * (retry_count + 1)) # Sleep for 2 seconds between requests
|
||||
|
||||
break # Break out of the retry loop if successful
|
||||
except Exception as e:
|
||||
print(f"Error: {e}")
|
||||
retry_count += 1
|
||||
print("Sleeping secs before retry", 100 * (retry_count + 1))
|
||||
time.sleep(100 * (retry_count + 1))
|
||||
if retry_count >= max_retries:
|
||||
print("Max retries reached. Exiting.")
|
||||
break
|
||||
|
||||
# DataFrame from the collected job data
|
||||
jobs_df = pd.DataFrame(all_jobs)
|
||||
|
||||
# Formatting
|
||||
pd.set_option("display.max_columns", None)
|
||||
pd.set_option("display.max_rows", None)
|
||||
pd.set_option("display.width", None)
|
||||
pd.set_option("display.max_colwidth", 50)
|
||||
|
||||
print(jobs_df)
|
||||
|
||||
jobs_df.to_csv(csv_filename, index=False)
|
||||
print(f"Outputted to {csv_filename}")
|
Loading…
Reference in New Issue