fix missing whitespace for links and names (#225)

* fix missing whitespace for links and names * remove comments in notebook * cleanup
ssciwr · Nov 22, 2024 · 8057153 · 8057153
1 parent dc6c904
commit 8057153
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 1 deletion.
diff --git a/ammico/notebooks/DemoNotebook_ammico.ipynb b/ammico/notebooks/DemoNotebook_ammico.ipynb
@@ -172,7 +172,7 @@
    "outputs": [],
    "source": [
     "image_dict = ammico.find_files(\n",
-    "    # path=\"/content/drive/MyDrive/misinformation-data/\",\n",
+    "    # path = \"/content/drive/MyDrive/misinformation-data/\",\n",
     "    path=str(data_path),\n",
     "    limit=15,\n",
     ")"

diff --git a/ammico/text.py b/ammico/text.py
@@ -237,6 +237,15 @@ def analyse_image(self) -> dict:
         if not self.subdict["text"]:
             print("No text found - skipping analysis.")
         else:
+            # make sure all full stops are followed by whitespace
+            # otherwise googletrans breaks
+            index_stop = self.subdict["text"].find(".")
+            if self.subdict["text"][index_stop + 1] != " ":
+                self.subdict["text"] = (
+                    self.subdict["text"][: index_stop + 1]
+                    + " "
+                    + self.subdict["text"][index_stop + 1 :]
+                )
             self.translate_text()
             self.remove_linebreaks()
             if self.analyse_text: