Generate anchors for single-page mode automatically

docs/tools/build.py

@@ -54,6 +54,7 @@ markdown.extensions.ClickHouseMarkdown = ClickHouseMarkdown
 
 def build_for_lang(lang, args):
     logging.info('Building %s docs' % lang)
+    os.environ['SINGLE_PAGE'] = '0'
 
     config_path = os.path.join(args.docs_dir, 'toc_%s.yml' % lang)
 
@@ -139,6 +140,7 @@ def build_for_lang(lang, args):
 
 def build_single_page_version(lang, args, cfg):
     logging.info('Building single page version for ' + lang)
+    os.environ['SINGLE_PAGE'] = '1'
 
     with autoremoved_file(os.path.join(args.docs_dir, lang, 'single.md')) as single_md:
         concatenate(lang, args.docs_dir, single_md)

docs/tools/concatenate.py

@@ -37,34 +37,21 @@ def concatenate(lang, docs_path, single_page_file):
     first_file = True
 
     for path in files_to_concatenate:
-
-        single_page_file.write('\n\n')
-
         with open(os.path.join(lang_path, path)) as f:
+            tmp_path = path.replace('.md', '/')
+            prefixes = ['', '../', '../../', '../../../']
+            parts = tmp_path.split('/')
+            single_page_file.write('<a name="%s/"></a>\n' % parts[-2])
+            single_page_file.write('\n\n')
 
-            # function is passed into re.sub() to process links
-            def link_proc(matchObj):
-                text, link = matchObj.group().strip('[)').split('](', 1)
-                if link.startswith('http:') or link.startswith('https:') or '.jpeg' in link or '.jpg' in link or '.png' in link or '.gif' in link:
-                    return '[' + text + '](' + link + ')'
-                else:
-                    sharp_pos = link.find('#')
-                    if sharp_pos > -1:
-                        return '[' + text + '](' + link[sharp_pos:] + ')'
-                    else:
-                        return '[' + text + '](#' + link.replace('../', '').replace('/index.md', '').replace('.md', '') + ')'
+            for part in parts[0:-2]:
+                for prefix in prefixes:
+                    single_page_file.write('<a name="%s"></a>\n' % (prefix + tmp_path))
+                tmp_path = tmp_path.replace(part, '..')
                 
             for l in f:
-                # Processing links in a string
-                l = re.sub(r'\[.+?\]\(.+?\)', link_proc, l)
-
-                # Correcting headers levels
-                if not first_file:
-                    if l.startswith('#'):
-                        l = '#' + l
-                else:
-                    first_file = False
-
+                if l.startswith('#'):
+                    l = '#' + l
                 single_page_file.write(l)
 
     single_page_file.flush()

docs/tools/mdx_clickhouse.py

@@ -2,31 +2,41 @@
 # -*- coding: utf-8 -*-
 from __future__ import unicode_literals
 
+import os
+
 import markdown.inlinepatterns
 import markdown.extensions
 import markdown.util
 
 import slugify as slugify_impl
 
-class NofollowMixin(object):
+class ClickHouseLinkMixin(object):
+
     def handleMatch(self, m):
+        single_page = (os.environ.get('SINGLE_PAGE') == '1')
         try:
-            el = super(NofollowMixin, self).handleMatch(m)
+            el = super(ClickHouseLinkMixin, self).handleMatch(m)
         except IndexError:
             return
 
         if el is not None:
             href = el.get('href') or ''
-            if href.startswith('http') and not href.startswith('https://clickhouse.yandex'):
+            is_external = href.startswith('http:') or href.startswith('https:')
+            if is_external and not href.startswith('https://clickhouse.yandex'):
                 el.set('rel', 'external nofollow')
+            elif single_page:
+                if '#' in href:
+                    el.set('href', '#' + href.split('#', 1)[1])
+                else:
+                    el.set('href', '#' + href.replace('.md', '/'))
         return el
 
 
-class NofollowAutolinkPattern(NofollowMixin, markdown.inlinepatterns.AutolinkPattern):
+class ClickHouseAutolinkPattern(ClickHouseLinkMixin, markdown.inlinepatterns.AutolinkPattern):
     pass
 
 
-class NofollowLinkPattern(NofollowMixin, markdown.inlinepatterns.LinkPattern):
+class ClickHouseLinkPattern(ClickHouseLinkMixin, markdown.inlinepatterns.LinkPattern):
     pass
 
 
@@ -41,9 +51,8 @@ class ClickHouseMarkdown(markdown.extensions.Extension):
 
     def extendMarkdown(self, md, md_globals):
         md.preprocessors['clickhouse'] = ClickHousePreprocessor()
-        md.inlinePatterns['link'] = NofollowLinkPattern(markdown.inlinepatterns.LINK_RE, md)
-        md.inlinePatterns['autolink'] = NofollowAutolinkPattern(markdown.inlinepatterns.AUTOLINK_RE, md)
-
+        md.inlinePatterns['link'] = ClickHouseLinkPattern(markdown.inlinepatterns.LINK_RE, md)
+        md.inlinePatterns['autolink'] = ClickHouseAutolinkPattern(markdown.inlinepatterns.AUTOLINK_RE, md)
 
 def makeExtension(**kwargs):
     return ClickHouseMarkdown(**kwargs)