add masking

wragge · Apr 16, 2024 · 7301c36 · 7301c36
1 parent 88cb726
commit 7301c36
Show file tree

Hide file tree

Showing 7 changed files with 202 additions and 43 deletions.
diff --git a/00_articles.ipynb b/00_articles.ipynb
diff --git a/README.md b/README.md
@@ -1,5 +1,5 @@
-trove-newspaper-images
-================
+# trove-newspaper-images
+
 
 <!-- WARNING: THIS FILE WAS AUTOGENERATED! DO NOT EDIT! -->
 
@@ -54,6 +54,13 @@ specify a directory for the downloaded images. For example:
 trove_newspaper_images.download 107024751 --output_dir images
 ```
 
+Add the `--masked` parameter to try and remove content from neighbouring
+articles.
+
+``` shell
+trove_newspaper_images.download 107024751 --masked
+```
+
 ------------------------------------------------------------------------
 
 Created by [Tim Sherratt](https://timsherratt.org)

diff --git a/index.ipynb b/index.ipynb
@@ -80,6 +80,12 @@
     "\n",
     "```shell\n",
     "trove_newspaper_images.download 107024751 --output_dir images\n",
+    "```\n",
+    "\n",
+    "Add the `--masked` parameter to try and remove content from neighbouring articles.\n",
+    "\n",
+    "```shell\n",
+    "trove_newspaper_images.download 107024751 --masked\n",
     "```"
    ]
   },
@@ -95,7 +101,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
+   "display_name": "python3",
    "language": "python",
    "name": "python3"
   }

diff --git a/settings.ini b/settings.ini
@@ -5,7 +5,7 @@
 ### Python library ###
 repo = trove_newspaper_images
 lib_name = %(repo)s
-version = 0.2.1
+version = 0.3.1
 min_python = 3.8
 license = mit
 
@@ -15,7 +15,7 @@ lib_path = trove_newspaper_images
 nbs_path = .
 recursive = False
 tst_flags = notest
-put_version_in_init = False
+put_version_in_init = True
 
 ### Docs ###
 branch = master

diff --git a/trove_newspaper_images/__init__.py b/trove_newspaper_images/__init__.py
@@ -1 +1 @@
-__version__ = "0.2.0"
+__version__ = "0.3.1"
diff --git a/trove_newspaper_images/_modidx.py b/trove_newspaper_images/_modidx.py
@@ -12,4 +12,4 @@
                                                  'trove_newspaper_images.articles.get_box': ( 'articles.html#get_box',
                                                                                               'trove_newspaper_images/articles.py'),
                                                  'trove_newspaper_images.articles.main': ( 'articles.html#main',
-                                                                                           'trove_newspaper_images/articles.py')}}}
+                                                                                           'trove_newspaper_images/articles.py')}}}
diff --git a/trove_newspaper_images/articles.py b/trove_newspaper_images/articles.py
@@ -29,21 +29,30 @@ def get_box(zones):
     page_id = zones[0]["data-page-id"]
 
     # Loop through zones to find the outer boundaries of the article
+    z_boxes = []
     for zone in zones:
-        if int(zone["data-y"]) < top:
-            top = int(zone["data-y"])
-        if int(zone["data-x"]) < left:
-            left = int(zone["data-x"])
-        if (int(zone["data-x"]) + int(zone["data-w"])) > right:
-            right = int(zone["data-x"]) + int(zone["data-w"])
-        if (int(zone["data-y"]) + int(zone["data-h"])) > bottom:
-            bottom = int(zone["data-y"]) + int(zone["data-h"])
+        z_left = int(zone["data-x"])
+        z_top = int(zone["data-y"])
+        z_right = int(zone["data-x"]) + int(zone["data-w"])
+        z_bottom = int(zone["data-y"]) + int(zone["data-h"])
+        z_boxes.append(
+            {"left": z_left, "top": z_top, "right": z_right, "bottom": z_bottom}
+        )
+        if z_top < top:
+            top = z_top
+        if z_left < left:
+            left = z_left
+        if z_right > right:
+            right = z_right
+        if z_bottom > bottom:
+            bottom = z_bottom
     return {
         "page_id": page_id,
         "left": left,
         "top": top,
         "right": right,
         "bottom": bottom,
+        "zones": z_boxes,
     }
 
 
@@ -102,7 +111,7 @@ def get_article_boxes(article_id):
     return boxes
 
 
-def download_images(article_id, output_dir="", size=None):
+def download_images(article_id, output_dir="", size=None, masked=False):
     """
     Extract an image of a newspaper article from the page image(s), download and save it, and return the image filename(s).
 
@@ -111,6 +120,7 @@ def download_images(article_id, output_dir="", size=None):
     * article_id -- identifier for a Trove newspaper article
     * output_dir -- a directory to save images in (will be created if it doesn't exist)
     * size -- maximum dimensions of image
+    * masked -- `True` or `False`, remove content that isn't part of the article
 
     Returns:
 
@@ -122,7 +132,7 @@ def download_images(article_id, output_dir="", size=None):
         output_path.mkdir(exist_ok=True, parents=True)
     else:
         output_path = ""
-    
+
     # Get position of article on the page(s)
     boxes = get_article_boxes(article_id)
     for box in boxes:
@@ -134,24 +144,49 @@ def download_images(article_id, output_dir="", size=None):
             page_url = f'https://trove.nla.gov.au/ndp/imageservice/nla.news-page{box["page_id"]}/level7'
 
             # Download the page image
-            response = requests.get(page_url)
+            response = requests.get(page_url, stream=True)
 
             # Open download as an image for editing
             img = Image.open(BytesIO(response.content))
 
             # Use coordinates of the bounding box to crop article
             points = (box["left"], box["top"], box["right"], box["bottom"])
 
-            # Crop image to article box
-            cropped = img.crop(points)
+            if masked:
+
+                # Create a new empty image the same size as the original
+                new_img = Image.new("RGB", img.size, "#fdfdfd")
+
+                # Process each zone separately
+                for zone in box["zones"]:
+
+                    # Get zone coords
+                    z_points = (
+                        zone["left"],
+                        zone["top"],
+                        zone["right"],
+                        zone["bottom"],
+                    )
+
+                    # Crop the zone from the original image
+                    zone_crop = img.crop(z_points)
+
+                    # Paste the zone into the new image
+                    new_img.paste(zone_crop, z_points)
+
+                # Crop the new image to the article box
+                cropped = new_img.crop(points)
+
+            else:
+
+                # Crop image to article box
+                cropped = img.crop(points)
 
             # Resize if necessary
             if size:
                 cropped.thumbnail((size, size), Image.LANCZOS)
 
             # Save cropped image
-
-
             cropped.save(cropped_file)
         images.append(cropped_file.name)
     # print(f'Downloaded: {images}')
@@ -166,5 +201,6 @@ def main():
     parser.add_argument("article_id", help="article identifier")
     parser.add_argument("--output_dir", help="directory to save images")
     parser.add_argument("--size", help="maximum image dimensions")
+    parser.add_argument('--masked', action="store_true", help="mask image")
     args = parser.parse_args()
-    download_images(args.article_id, args.output_dir, args.size)
+    download_images(args.article_id, args.output_dir, args.size, args.masked)