From 22f6db0a6bf02fa7a8473e87be529b14d86a3ba1 Mon Sep 17 00:00:00 2001
From: Debanjum Singh Solanky <debanjum@gmail.com>
Date: Sat, 22 Jun 2024 16:00:06 +0530
Subject: [PATCH] Upgrade RapidOCR and enable for Python 3.12. Fix PDF OCR test

---
 pyproject.toml                |  2 +-
 tests/test_docx_to_entries.py |  2 --
 tests/test_pdf_to_entries.py  | 14 +++++++++++---
 3 files changed, 12 insertions(+), 6 deletions(-)

diff --git a/pyproject.toml b/pyproject.toml
index de4a05f0..19e7a076 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -73,7 +73,7 @@ dependencies = [
     "psycopg2-binary == 2.9.9",
     "lxml == 4.9.3",
     "tzdata == 2023.3",
-    "rapidocr-onnxruntime == 1.3.11; python_version<'3.12'",
+    "rapidocr-onnxruntime == 1.3.22",
     "openai-whisper >= 20231117",
     "django-phonenumber-field == 7.3.0",
     "phonenumbers == 8.13.27",
diff --git a/tests/test_docx_to_entries.py b/tests/test_docx_to_entries.py
index 089c7fec..dce0109b 100644
--- a/tests/test_docx_to_entries.py
+++ b/tests/test_docx_to_entries.py
@@ -1,5 +1,3 @@
-import os
-
 from khoj.processor.content.docx.docx_to_entries import DocxToEntries
 
 
diff --git a/tests/test_pdf_to_entries.py b/tests/test_pdf_to_entries.py
index 31ccb387..3a25b05d 100644
--- a/tests/test_pdf_to_entries.py
+++ b/tests/test_pdf_to_entries.py
@@ -1,4 +1,5 @@
 import os
+import re
 
 from khoj.processor.content.pdf.pdf_to_entries import PdfToEntries
 from khoj.utils.fs_syncer import get_pdf_files
@@ -38,16 +39,23 @@ def test_multi_page_pdf_to_jsonl():
 
 def test_ocr_page_pdf_to_jsonl():
     "Convert multiple pages from single PDF file to jsonl."
-    # Act
+    # Arrange
+    expected_str = "playing on a strip of marsh"
+    expected_str_with_variable_spaces = re.compile(expected_str.replace(" ", r"\s*"), re.IGNORECASE)
+
     # Extract Entries from specified Pdf files
     with open("tests/data/pdf/ocr_samples.pdf", "rb") as f:
         pdf_bytes = f.read()
-
     data = {"tests/data/pdf/ocr_samples.pdf": pdf_bytes}
+
+    # Act
     entries = PdfToEntries.extract_pdf_entries(pdf_files=data)
+    raw_entry = entries[1][0].raw
+
+    # Assert
     assert len(entries) == 2
     assert len(entries[1]) == 1
-    assert "playing on a strip of marsh" in entries[1][0].raw
+    assert re.search(expected_str_with_variable_spaces, raw_entry) is not None
 
 
 def test_get_pdf_files(tmp_path):