Parse Resumes with Python

import os
import requests

API_URL = "https://api.resumeparser.cc/v1/parse"
API_KEY = os.environ["RESUME_PARSER_API_KEY"]  # rp_live_...
MAX_BYTES = 5 * 1024 * 1024  # 5 MB


def parse_resume(file_path: str) -> dict:
    if os.path.getsize(file_path) > MAX_BYTES:
        raise ValueError("file exceeds 5 MB")

    with open(file_path, "rb") as f:
        response = requests.post(
            API_URL,
            headers={"Authorization": f"Bearer {API_KEY}"},
            files={"file": (os.path.basename(file_path), f, "application/pdf")},
            timeout=90,  # parsing can take up to 60 seconds
        )

    # A proxy error can return HTML instead of JSON
    if not response.headers.get("content-type", "").startswith("application/json"):
        response.raise_for_status()

    body = response.json()
    if response.status_code == 422:
        # No resume data found, usually a scanned PDF without a text layer. Not charged.
        raise ValueError(f"could not parse {file_path}: {body['error']['message']}")
    if not response.ok:
        error = body["error"]
        raise RuntimeError(f"{response.status_code} {error['code']}: {error['message']}")

    return body


result = parse_resume("resume.pdf")
data = result["data"]

# Every field is optional: it is left out when the CV does not have it
print(data.get("firstName"), data.get("lastName"))
print(data.get("emailAddress"))

for experience in data.get("workExperiences", []):
    print(experience["title"], "at", experience["organization"])
    print(experience["startMonth"], experience["startYear"], "ongoing:", experience["isOngoing"])

for education in data.get("educations", []):
    print(education["degree"], "at", education["school"])

print(result["usage"])  # billedAs, freeRemaining, creditBalance

import os
import time
import json
from pathlib import Path

import requests

API_URL = "https://api.resumeparser.cc/v1/parse"
API_KEY = os.environ["RESUME_PARSER_API_KEY"]

# 30 requests per minute per key: one request every 2 seconds, one at a time
SECONDS_BETWEEN_REQUESTS = 2

output_dir = Path("parsed")
output_dir.mkdir(exist_ok=True)

for pdf_path in sorted(Path("resumes").glob("*.pdf")):
    with open(pdf_path, "rb") as f:
        response = requests.post(
            API_URL,
            headers={"Authorization": f"Bearer {API_KEY}"},
            files={"file": (pdf_path.name, f, "application/pdf")},
            timeout=60,
        )

    body = response.json()
    if response.ok:
        (output_dir / f"{pdf_path.stem}.json").write_text(json.dumps(body["data"], indent=2))
        print(pdf_path.name, "ok, free remaining:", body["usage"]["freeRemaining"])
    else:
        print(pdf_path.name, "failed:", body["error"]["code"])

    time.sleep(SECONDS_BETWEEN_REQUESTS)

import os
import time

import requests

API_URL = "https://api.resumeparser.cc/v1/parse"
API_KEY = os.environ["RESUME_PARSER_API_KEY"]

# 429 rate_limited and 503 busy are worth retrying.
# 400 invalid_file, 401 invalid_api_key, 402 insufficient_credits and
# 413 file_too_large will fail again with the same request.
RETRYABLE_STATUSES = {429, 503}


def parse_resume_with_retry(file_path: str, max_attempts: int = 4) -> dict:
    for attempt in range(1, max_attempts + 1):
        with open(file_path, "rb") as f:
            response = requests.post(
                API_URL,
                headers={"Authorization": f"Bearer {API_KEY}"},
                files={"file": (os.path.basename(file_path), f, "application/pdf")},
                timeout=60,
            )

        if response.ok:
            return response.json()

        error = response.json()["error"]
        if response.status_code not in RETRYABLE_STATUSES or attempt == max_attempts:
            raise RuntimeError(f"{response.status_code} {error['code']}: {error['message']}")

        time.sleep(10 * attempt)

    raise RuntimeError("unreachable")