Remove unused test.

JSv4 · Oct 14, 2024 · 7bb640e · 7bb640e
1 parent 7adf4eb
commit 7bb640e
Show file tree

Hide file tree

Showing 3 changed files with 5 additions and 46 deletions.
diff --git a/opencontractserver/documents/signals.py b/opencontractserver/documents/signals.py
@@ -7,7 +7,9 @@
 from opencontractserver.tasks.doc_tasks import (
     extract_pdf_thumbnail,
     nlm_ingest_pdf,
-    set_doc_lock_state, extract_txt_thumbnail, ingest_txt,
+    set_doc_lock_state,
+    extract_txt_thumbnail,
+    ingest_txt,
 )
 from opencontractserver.tasks.embeddings_task import calculate_embedding_for_doc_text
 

diff --git a/opencontractserver/tasks/doc_tasks.py b/opencontractserver/tasks/doc_tasks.py
@@ -4,8 +4,6 @@
 import io
 import json
 import logging
-import pathlib
-import uuid
 from typing import Any
 
 import requests
@@ -39,7 +37,8 @@
 from opencontractserver.utils.etl import build_document_export, pawls_bbox_to_funsd_box
 from opencontractserver.utils.files import (
     check_if_pdf_needs_ocr,
-    split_pdf_into_images, create_text_thumbnail,
+    split_pdf_into_images,
+    create_text_thumbnail,
 )
 from opencontractserver.utils.permissioning import set_permissions_for_obj_to_user
 from opencontractserver.utils.text import __consolidate_common_equivalent_chars

diff --git a/opencontractserver/tests/test_doc_parser.py b/opencontractserver/tests/test_doc_parser.py
@@ -1,15 +1,10 @@
 #  Copyright (C) 2022  John Scrudato
-import io
 import logging
-import pathlib
-import uuid
 
 from django.contrib.auth import get_user_model
 from django.core.files.base import ContentFile
-from django.core.files.storage import default_storage
 from django.db import transaction
 from django.test import TestCase
-from PyPDF2 import PdfReader, PdfWriter
 
 from opencontractserver.annotations.models import Annotation, AnnotationLabel
 from opencontractserver.corpuses.models import Corpus
@@ -18,7 +13,6 @@
     burn_doc_annotations,
     convert_doc_to_funsd,
     extract_pdf_thumbnail,
-    process_pdf_page,
     set_doc_lock_state,
 )
 from opencontractserver.tests.fixtures import (
@@ -66,42 +60,6 @@ def test_pdf_thumbnail_extraction(self):
         # TODO - expand test to actually check results
         extract_pdf_thumbnail.s(doc_id=self.doc.id).apply().get()
 
-    def test_process_pdf_page(self):
-        page_bytes_stream = io.BytesIO()
-        doc_path = self.doc.pdf_file.name
-        doc_file = default_storage.open(doc_path, mode="rb")
-        pdf = PdfReader(doc_file)
-        page = pdf.pages[0]
-        pdf_writer = PdfWriter()
-        pdf_writer.add_page(page)
-        pdf_writer.write(page_bytes_stream)
-
-        pdf_fragment_folder_path = pathlib.Path(
-            f"/tmp/user_{self.user.id}/pdf_fragments"
-        )
-        pdf_fragment_folder_path.mkdir(parents=True, exist_ok=True)
-        pdf_fragment_path = pdf_fragment_folder_path / f"{uuid.uuid4()}.pdf"
-        with pdf_fragment_path.open("wb") as f:
-            f.write(page_bytes_stream.getvalue())
-
-        page_path = pdf_fragment_path.resolve().__str__()
-
-        result = (
-            process_pdf_page.si(
-                total_page_count=23,
-                page_num=0,
-                page_path=page_path,
-                user_id=self.user.id,
-            )
-            .apply()
-            .get()
-        )
-
-        self.assertEqual(len(result), 3)
-        self.assertEqual(result[0], 0)  # page number
-        self.assertIsInstance(result[1], str)  # pawls fragment path
-        self.assertEqual(result[2], pdf_fragment_path.__str__())  # page path
-
     def test_set_doc_lock_state(self):
         set_doc_lock_state.apply(kwargs={"locked": True, "doc_id": self.doc.id}).get()