Merging changes from upstream

cdp_backend/utils/file_utils.py

@@ -9,10 +9,7 @@
 from typing import Optional, Tuple, Union
 
 import fsspec
-import imageio
-import numpy as np
 from fsspec.core import url_to_fs
-from PIL import Image
 
 ###############################################################################
 
@@ -207,6 +204,8 @@ def get_static_thumbnail(
  The name of the thumbnail file:
  Always session_content_hash + "-static-thumbnail.png"
  """
+ import imageio
+ from PIL import Image
 
  reader = imageio.get_reader(video_path)
  png_path = ""
@@ -221,15 +220,9 @@ def get_static_thumbnail(
  reader = imageio.get_reader(video_path)
  image = reader.get_data(0)
 
- if image.shape[0] > MAX_THUMBNAIL_HEIGHT or image.shape[1] > MAX_THUMBNAIL_WIDTH:
- height_ratio = MAX_THUMBNAIL_HEIGHT / image.shape[0]
- width_ratio = MAX_THUMBNAIL_WIDTH / image.shape[1]
-
- if height_ratio > width_ratio:
- final_ratio = height_ratio
- else:
- final_ratio = width_ratio
+ final_ratio = find_proper_resize_ratio(image.shape[0], image.shape[1])
 
+ if final_ratio < 1:
  image = Image.fromarray(image).resize(
  (
  math.floor(image.shape[1] * final_ratio),
@@ -264,6 +257,10 @@ def get_hover_thumbnail(
  The name of the thumbnail file:
  Always session_content_hash + "-hover-thumbnail.png"
  """
+ import imageio
+ import numpy as np
+ from PIL import Image
+
  reader = imageio.get_reader(video_path)
  gif_path = ""
  if reader.get_length() > 1:
@@ -274,27 +271,18 @@ def get_hover_thumbnail(
  count += 1
  step_size = math.floor(count / num_frames)
 
- final_ratio = 1
  height = image.shape[0]
  width = image.shape[1]
-
- if height > MAX_THUMBNAIL_HEIGHT or width > MAX_THUMBNAIL_WIDTH:
- height_ratio = MAX_THUMBNAIL_HEIGHT / height
- width_ratio = MAX_THUMBNAIL_WIDTH / width
-
- if height_ratio > width_ratio:
- final_ratio = height_ratio
- else:
- final_ratio = width_ratio
+ final_ratio = find_proper_resize_ratio(height, width)
 
  with imageio.get_writer(gif_path, mode="I") as writer:
  for i in range(0, num_frames):
- if final_ratio == 1:
- image = Image.fromarray(reader.get_data(i * step_size))
- else:
+ if final_ratio < 1:
  image = Image.fromarray(reader.get_data(i * step_size)).resize(
  (math.floor(width * final_ratio), math.floor(height * final_ratio))
  )
+ else:
+ image = Image.fromarray(reader.get_data(i * step_size))
 
  final_image = np.asarray(image, dtype="int32")
  writer.append_data(final_image)

setup.py

@@ -21,12 +21,14 @@
  "imageio-ffmpeg~=0.4.4",
  "nltk~=3.6",
  "pandas~=1.2",
+ "pdftotext~=2.2.0",
  "prefect~=0.14.0",
  "pulumi~=3.3",
  "pulumi-google-native~=0.7.0",
  "pulumi-gcp~=5.7",
  "rapidfuzz~=1.4",
  "spacy~=3.0",
+ "textract~=1.6.4",
  "truecase~=0.0.12",
  "webvtt-py~=0.4.6",
 ]