Generate sitemap

example/example_site.toml

@@ -5,6 +5,12 @@
 # name of the folder that the site will be generated in
 name = "Notion Test Site"
 
+# Sitemap configuration
+# Remove if you don't want to generate sitemap.xml
+protocol = "https"
+domain = "example.com"
+remove_html_extension = true
+
 # the notion.so page to being parsing from. This page will become the index.html
 # of the generated site, and loconotation will parse all sub-pages present on the page
 page = "https://www.notion.so/Loconotion-Example-Page-03c403f4fdc94cc1b315b9469a8950ef"

loconotion/modules/notionparser.py

@@ -754,6 +754,21 @@ def parse_subpages(self, subpages):
                 if sub_page not in self.processed_pages.keys():
                     self.parse_page(sub_page)
 
+    def export_sitemap(self, protocol:str, domain:str, processed_pages:set, remove_html_extension:bool):
+        # Open file in dist/sitemap.xml to write sitemap
+        with open(self.dist_folder / "sitemap.xml", "w") as f:
+            # Write XML header
+            f.write('<?xml version="1.0" encoding="UTF-8"?>\r')
+            # Write sitemap index opening tag
+            f.write('<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">\r')
+            # Write the sitemap from domain and processed pages
+            for page in processed_pages:
+                if remove_html_extension:
+                    page = page.replace(".html", "")
+                f.write(f'\t<url><loc>{protocol}://{domain}/{page}</loc></url>\r')
+            # Write sitemap index closing tag
+            f.write("</urlset>")
+
     def load(self, url):
         self.driver.get(url)
         WebDriverWait(self.driver, 60).until(notion_page_loaded())
@@ -762,6 +777,12 @@ def run(self):
         start_time = time.time()
         self.processed_pages = {}
         self.parse_page(self.starting_url)
+        if self.config.get("domain",None):
+            self.export_sitemap(
+                self.config.get("protocol", "https"),
+                self.config.get("domain"),
+                set(self.processed_pages.values()),
+                self.config.get("remove_html_extension", False))
         elapsed_time = time.time() - start_time
         formatted_time = "{:02d}:{:02d}:{:02d}".format(
             int(elapsed_time // 3600),