Work around Pandoc bug #7398

[etc/mutt.git] / .mutt / markdown2html
diff --git a/.mutt/markdown2html b/.mutt/markdown2html

index 35b4ea2900f9a5764a3d0429b9679ffd65ba13a1..60e6bea16e75c20e977d2b6eba02bc47d2e615e1 100755 (executable)
--- a/.mutt/markdown2html
+++ b/.mutt/markdown2html
@@ -107,7 +107,6 @@ HTML_DOCUMENT = '''<!DOCTYPE html>
  <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
  <meta charset="utf-8"/>
  <meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes"/>
-<title>HTML E-Mail</title>
  </head><body class="email">
  {htmlbody}
  </body></html>'''
@@ -134,6 +133,13 @@ def _preprocess_markdown(mdwn):
      # text-mode HTML2text converters, and so it's left commented for now.
      #ret = re.sub(r'\n>', r'  \n>[>]{.quotechar}', ret, flags=re.MULTILINE)
  
+    # With the autolink_bare_uris extension, we do not need to put links into
+    # angle brackets to have them converted, so let's conserve the brackets
+    # when used around email addresses. Note that this needs a postprocessing
+    # hack because the pandoc autolink converted includes the ambersand
+    # (https://github.com/jgm/pandoc/issues/7398).
+    ret = re.sub(r'<([^@]+@.+\.[^>]+)>', r'&lt;\g<1> -PANDOC_BUG_7398-&gt;', ret)
+
      return ret
  
  
@@ -215,7 +221,7 @@ def _reformat_quotes(html):
  
  def _convert_with_pandoc(mdwn, inputfmt='markdown', outputfmt='html5',
                           ext_enabled=None, ext_disabled=None,
-                         standalone=True, title="HTML E-Mail"):
+                         standalone=True, selfcontained=True, title=None):
      '''
      Invoke pandoc to do the actual conversion of Markdown to HTML5.
      '''
@@ -258,6 +264,8 @@ def _convert_with_pandoc(mdwn, inputfmt='markdown', outputfmt='html5',
      args = []
      if standalone:
          args.append('--standalone')
+    if selfcontained:
+        args.append('--self-contained')
      if title:
          args.append(f'--metadata=pagetitle:"{title}"')
  
@@ -276,6 +284,10 @@ def _postprocess_html(html):
      '''
      Postprocess the generated and styled HTML.
      '''
+
+    # Preprocessing leaves a sentinel to work around
+    # https://github.com/jgm/pandoc/issues/7398, and so we need to remove it:
+    html = html.replace('</a> -PANDOC_BUG_7398-&gt;', '</a>&gt;')
      return html
  
  
@@ -295,7 +307,7 @@ def convert_markdown_to_html(mdwn):
      if body:
          body = _preprocess_markdown(body)
          body = _identify_quotes_for_later(body)
-        html = _convert_with_pandoc(body, standalone=False)
+        html = _convert_with_pandoc(body, standalone=True, selfcontained=True)
          html = _reformat_quotes(html)
  
      if sig: