Spaces:
Runtime error
Runtime error
| from fastapi import FastAPI, HTTPException | |
| from models import CVExtracted, EvaModul, JobAndCV, ClassificationResult, InsertedLink, EvalResult | |
| import os | |
| from io import BytesIO | |
| # import extractor | |
| import extractor_llm | |
| from datetime import datetime | |
| from pypdf import PdfReader | |
| import requests | |
| import classificator | |
| import evaluator | |
| import extractor_helper | |
| os.environ['TRANSFORMERS_CACHE'] = '/transformers_cache' | |
| os.environ['HF_HOME'] = '/transformers_cache' | |
| app = FastAPI() | |
| def getall(): | |
| return {"hello":"world"} | |
| # @app.post("/ext", response_model=CVExtracted) | |
| # async def extract(text: InsertedText): | |
| # dictresult = extractor.predict(text.text) | |
| # return CVExtracted(**dictresult) | |
| async def classify(body:JobAndCV): | |
| mininmal_start = 0 | |
| maximal_end = 0 | |
| positions = [] | |
| userMajors = [] | |
| yoe = 0 | |
| if len(body.cv.experiences) > 0: | |
| mininmal_start = datetime.strptime(body.cv.experiences[0]['start'], "%Y-%m-%d").date() if body.cv.experiences[0].get('start') != None else datetime.today().date() | |
| maximal_end = datetime.strptime(body.cv.experiences[0]['end'], "%Y-%m-%d").date() if body.cv.experiences[0].get('end') != None else datetime.today().date() | |
| for exp in body.cv.experiences: | |
| positions.append(exp['position']) | |
| if exp.get('end') == None: | |
| exp['end'] = datetime.today().strftime("%Y-%m-%d") | |
| if datetime.strptime(exp['start'], "%Y-%m-%d").date() < mininmal_start: | |
| mininmal_start = datetime.strptime(exp['start'], "%Y-%m-%d").date() | |
| if datetime.strptime(exp['end'], "%Y-%m-%d").date() > maximal_end: | |
| maximal_end = datetime.strptime(exp['end'], "%Y-%m-%d").date() | |
| yoe = (maximal_end - mininmal_start).days//365 | |
| for edu in body.cv.educations: | |
| userMajors.append(edu['major']) | |
| cv = { | |
| "experiences": str(body.cv.experiences), | |
| "positions": str(positions), | |
| "userMajors": str(userMajors), | |
| "skills": str(body.cv.skills), | |
| "yoe": yoe, | |
| "location": str(body.cv.location) | |
| } | |
| job = { | |
| "jobDesc": body.job.jobDesc, | |
| "role": body.job.role, | |
| "majors": str(body.job.majors), | |
| "skills": str(body.job.skills), | |
| "minYoE": body.job.minYoE, | |
| "location": str(body.job.location) | |
| } | |
| weight = body.weight.dict() | |
| results = classificator.predict(cv, job, weight) | |
| return ClassificationResult(**results) | |
| async def extract(link: InsertedLink): | |
| response = requests.get(link.link) | |
| if response.status_code == 200: | |
| # Open the PDF from bytes in memory | |
| pdf_reader = PdfReader(BytesIO(response.content)) | |
| number_of_pages = len(pdf_reader.pages) | |
| # Optionally, read text from the first page | |
| page = pdf_reader.pages[0] | |
| text = page.extract_text() | |
| for i in range(1, number_of_pages): | |
| text+= '\n' + pdf_reader.pages[i].extract_text() | |
| else: | |
| #return error, make 500 because file server error | |
| raise HTTPException(status_code=response.status_code, detail="File server error") | |
| dictresult = extractor_llm.predict(text) | |
| return dictresult | |
| async def eval(eva: EvaModul): | |
| transcript = extractor_helper.extract_technical(eva.competences, eva.transcript) | |
| return evaluator.evaluate_interview(competences=eva.competences, transcript=transcript, lang=eva.lang) |