Accept content with accents and change to TOML front matter

ghost2hugo.py

@@ -1,24 +1,33 @@
 #!/usr/bin/python3
+# -*- coding: utf-8 -*-
 
 import sqlite3
 import json
 from datetime import datetime
 
 
-conn = sqlite3.connect('ghost.db')
+# optional usage to remove accents on tag names
+# -------------
+import unicodedata
+def remove_accents(input_str):
+    nkfd_form = unicodedata.normalize('NFKD', unicode(input_str))
+    return u"".join([c for c in nkfd_form if not unicodedata.combining(c)])
+# -------------
+
+conn = sqlite3.connect('ghost-dev.db')
 conn.row_factory = sqlite3.Row
 
 c = conn.cursor()
 c2 = conn.cursor()
 
+l = []
 
 for i in c.execute('''
 SELECT id, title, meta_description as description, slug, markdown as text,
 status as draft, page, meta_title, image,
 DATE(published_at/1000, "unixepoch") as date,
 DATE(created_at/1000, "unixepoch") as date2
-FROM posts
-'''):
+FROM posts'''):
     g = {i.keys()[e]: tuple(i)[e] for e in range(len(i.keys()))}
     t = (i['id'],)
     g['tags'] = [e['name'] for e in c2.execute('''
@@ -33,22 +42,53 @@
         g['draft'] = True
     g.pop('date2')
 
+    # post description
+    if g['description'] == None:
+        g['description'] = ""
+
+    # post content    
     text = g.pop('text')
     text = text.replace("# ", "#")
     text = text.replace("#", "# ")
     text = text.replace("# # # ", "### ")
     text = text.replace("# # ", "## ")
     text = text.replace("\# ", "\#")
 
+    # post type
     if g['page'] == True:
         page = 'page'
     else:
         page = 'post'
     g['type'] = page
-    g.pop('page')
+    g.pop('page')    
+
+    with open('./content/%s.md' % (g['slug']), 'w') as post_file:
+        post_file.write('+++\n')
+        post_file.write('type = "%s"\n' % g['type'])
+        post_file.write('date = "%s"\n' % g['date'])
+        post_file.write('title = "%s"\n' % g['title'].encode('utf8'))
+        post_file.write('description = "%s"\n' % g['description'].encode('utf8'))
+        post_file.write('slug = "%s"\n' % g['slug'])
+
+        post_file.write('tags = [')
+
+        # encode each tag to accept accents or removed them
+        # and add a comma to separate each one
+        for i in xrange(0, len(g['tags'])):
+            if i < len(g['tags']) - 1 :
+                separator = ", "
+            else:
+                separator = ""
+
+            # encode string to keep accents etc. E.g. "Introdução e Avaliações"
+            tag = g['tags'][i].encode('utf8')
+
+            # uncomment if you like to remove accents. E.g. "Introducao e Avaliacoes"
+            tag = remove_accents(g['tags'][i])
+
+            post_file.write('"%s"' % tag+separator)
+
+        post_file.write(']\n')
 
-    f = open("./content/%s.md" % (g['title']), "w")
-    f.write(json.dumps(g, sort_keys=True, indent=4, separators=(',', ': ')))
-    f.write('\n\n\n')
-    f.write(text)
-    f.close()
+        post_file.write('+++\n\n')
+        post_file.write(text.encode('utf8'))