Improve header id generation

2 files changed, 16 insertions, 4 deletions

diff --git a/panfry/document.py b/panfry/document.py
index 7bc65c2..edf1cf8 100755
--- a/panfry/document.py
+++ b/panfry/document.py
@@ -42,18 +42,23 @@ class Document:
         '''
         re_link1 = re.compile(r'(\[(?P<s>[^\]^]+)\])\((?P<l>#.*)\)', re.S)
         re_link2 = re.compile(r'(\[(?P<s>[^\]^]+)\][^(])', re.S)
+        re_link3 = re.compile(r'(\[(?P<s>[^\]^]+)\])\[(?P<h>[^\]].*)\]', re.S)
         source = re_link1.sub(self._get_cross_link, source)
         source = re_link2.sub(self._get_cross_link, source)
+        source = re_link3.sub(self._get_cross_link, source)
 
         return source
 
     def _get_cross_link(self, m):
-        ref_txt = m.group('s')
+        if 'h' in m.groupdict():
+            heading = m.group('h')
+        else:
+            heading = m.group('s')
 
         for page in self.pages:
             for tocitem in page.toc:
-                if tocitem.heading == ref_txt.replace('\n', ' '):
-                    return '[%s](%s#%s)' % (ref_txt,
+                if tocitem.heading == heading.replace('\n', ' '):
+                    return '[%s](%s#%s)' % (heading,
                                             page.htmlfile,
                                             tocitem.header_id,
                                             )
diff --git a/panfry/page.py b/panfry/page.py
index 01d9fcb..42eda1e 100755
--- a/panfry/page.py
+++ b/panfry/page.py
@@ -35,7 +35,14 @@ class TOCItem:
 
     @property
     def header_id(self):
-        remove = re.compile('[+~!@#$%^&*\(\){}\[\];:"\',.<>?/\`]')
+        # - Remove all formatting, links, etc.
+        # - Remove all punctuation, except underscores, hyphens, and periods.
+        # - Replace all spaces and newlines with hyphens.
+        # - Convert all alphabetic characters to lowercase.
+        # - Remove everything up to the first letter (identifiers may not begin
+        #   with a number or punctuation mark).
+
+        remove = re.compile('^[^a-zA-Z]|[+~!@#$%^&*\(\){}\[\];:"\',<>?/\`]')
         header_id = self.heading.lower().replace('\n', ' ').replace(' ', '-')
         return remove.sub('', header_id)