Generate word_count and display in index.html (#81)

Added a script for calculating the total word count of translatable content (msgids) in a directory. Use script in generate.py/index.html to display Python Documentation current word count.

Author: StanFromIreland

generate.py

@@ -25,6 +25,7 @@
 import contribute
 from completion import branches_from_devguide, get_completion, TranslatorsData
 from repositories import Language, get_languages_and_repos
+from word_count import get_word_count
 
 generation_time = datetime.now(timezone.utc)
 
@@ -107,6 +108,7 @@ class LanguageProjectData:
         completion_progress=(completion_progress := list(get_completion_progress())),
         generation_time=generation_time,
         duration=(datetime.now(timezone.utc) - generation_time).seconds,
+        word_count=get_word_count(Path('clones', 'cpython', 'Doc', 'build', 'gettext')),
     )
 
     Path('index.html').write_text(output)

template.html.jinja

@@ -53,6 +53,7 @@ main | <a href="metadata.html" target="_self">meta</a>
 </tbody>
 </table>
 <p>* the number in parentheses shows change in the last 30 days, included in the total completion</p>
+<p>The Python documentation currently has a word count of {{ '{:,}'.format(word_count) }}.</p>
 <p>For more information about translations, see the <a href="https://devguide.python.org/documentation/translating/">Python Developer’s Guide</a>.</p>
 <p>Last updated at {{ generation_time.strftime('%A, %-d %B %Y, %-H:%M:%S %Z') }} (in {{ duration // 60 }}:{{ "{:02}".format(duration % 60) }} minutes).</p>
 </body>

word_count.py

@@ -0,0 +1,20 @@
+import os
+from polib import pofile
+
+
+def _count_words(pot) -> int:
+    pot = pofile(pot)
+    word_count = 0
+    for entry in pot:
+        word_count += len(entry.msgid.split())
+    return word_count
+
+
+def get_word_count(dir) -> int:
+    total_word_count = 0
+    for root, dirs, files in os.walk(dir):
+        for file in files:
+            if file.endswith('.pot'):
+                pot = os.path.join(root, file)
+                total_word_count += _count_words(pot)
+    return total_word_count