Merge pull request #26 from HappyHackingSpace/feat/pdf-extraction-improvement

feat: switch to pdfplumber for improved resume text extraction

Author: baranylcn

levelup/app.py

@@ -1,11 +1,10 @@
-import io
 import json
 import re
 from typing import Any, cast
 
 import google.generativeai as genai
 import pandas as pd  # type: ignore[import-untyped]
-import pypdf
+import pdfplumber
 import streamlit as st
 
 from levelup import config
@@ -23,11 +22,10 @@
 
 def extract_text_from_pdf(uploaded_file: Any) -> str | None:
     try:
-        pdf_reader = pypdf.PdfReader(io.BytesIO(uploaded_file.read()))
-        parts: list[str] = []
-        for page in pdf_reader.pages:
-            parts.append(page.extract_text() or "")
-        return "\n".join(parts).strip()
+        with pdfplumber.open(uploaded_file) as pdf:
+            parts = [page.extract_text() or "" for page in pdf.pages]
+        text = "\n".join(parts).strip()
+        return text
     except Exception as e:
         st.error(f"PDF reading error: {e}")
         return None

pyproject.toml

@@ -23,14 +23,14 @@ dependencies = [
     "metatron>=0.5",
     "psycopg2-binary>=2.9.11",
     "pydantic>=2.12.3",
-    "pypdf>=6.1.3",
     "sentry-sdk>=2.43.0",
     "slowapi>=0.1.9",
     "sqlalchemy>=2.0.44",
     "sqlalchemy-utils>=0.42.0",
     "starlette>=0.49.1",
     "streamlit>=1.51.0",
-    "pytest-mock==3.15.1"
+    "pytest-mock==3.15.1",
+    "pdfplumber==0.11.8"
 ]
 
 [dependency-groups]
@@ -166,7 +166,6 @@ module = [
     "google.generativeai.*",
     "streamlit.*",
     "metatron.*",
-    "pypdf.*",
     "chromedriver_autoinstaller.*",
     "selenium.*",
     "trafilatura.*",

tests/test_resume.py

@@ -1,6 +1,6 @@
-import io
 import os
 import sys
+from io import BytesIO
 from unittest.mock import MagicMock
 
 from pytest_mock import MockerFixture
@@ -22,15 +22,22 @@
 
 
 def test_extract_text_from_pdf(mocker: MockerFixture) -> None:
-    """Test extract_text_from_pdf function."""
-    mock_reader = MagicMock()
     mock_page = MagicMock()
     mock_page.extract_text.return_value = "test"
-    mock_reader.pages = [mock_page]
 
-    mocker.patch("levelup.app.pypdf.PdfReader", return_value=mock_reader)
+    mock_pdf = MagicMock()
+    mock_pdf.pages = [mock_page]
 
-    assert extract_text_from_pdf(io.BytesIO(b"test")) == "test"
+    mock_open = mocker.patch("levelup.app.pdfplumber.open")
+    mock_open.return_value.__enter__.return_value = mock_pdf
+    mock_open.return_value.__exit__.return_value = False
+
+    fake_file = BytesIO(b"%PDF-1.4 fake content")
+
+    text = extract_text_from_pdf(fake_file)
+
+    assert text == "test"
+    mock_open.assert_called_once_with(fake_file)
 
 
 def test_analyzecv_pdf_withllm_success(mocker: MockerFixture) -> None: