# HG changeset patch
# User Paul Boddie <paul@boddie.org.uk>
# Date 1315759816 -7200
# Node ID a2e8615152cfdd14e8749e4fba980e55d1a1221f
# Parent  5e3955c2714ba2964097db2142d5982887df5e7a
Introduced detection of certain syntactic features that should end sections,
such as horizontal rules and Include macro invocations, as well as headings.

diff -r 5e3955c2714b -r a2e8615152cf MoinContentSupport.py
--- a/MoinContentSupport.py	Sat Feb 26 21:45:56 2011 +0100
+++ b/MoinContentSupport.py	Sun Sep 11 18:50:16 2011 +0200
@@ -16,30 +16,40 @@
 # Regular expressions.
 # NOTE: These overlap with ImprovedMoinSearch and EventAggregator.
 
-heading_regexp = re.compile(r"^(?P<level>=+)\s*(?P<heading>.*?)\s*(?P=level)$", re.UNICODE | re.MULTILINE)
+heading_regexp_str = r"^(?P<level>=+)\s*(?P<heading>.*?)\s*(?P=level)$"
+hrule_regexp_str = r"^----$"
+include_regexp_str = r"^\s*<<Include.*?>>$"
+section_regexp_str = "(" + heading_regexp_str + "|" + hrule_regexp_str + "|" + include_regexp_str + ")"
+section_regexp = re.compile(section_regexp_str, re.UNICODE | re.MULTILINE)
 
 category_membership_str = ur"^\s*(?:(Category\S+)(?:\s+(Category\S+))*)\s*$"
 category_membership_regexp = re.compile(category_membership_str, re.MULTILINE | re.UNICODE)
 category_declarations_regexp = re.compile("^----$\s*" + category_membership_str, re.MULTILINE | re.UNICODE)
 
-def getHeadingDetails(body, min_level=None, max_level=None):
+def getSectionDetails(body, min_level=None, max_level=None):
 
     """
-    Return heading details from the given 'body' for headings with the given
+    Return section details from the given 'body' for headings with the given
     'min_level' to 'max_level' range. Specifying None or omitting 'max_level' or
     'min_level' removes the appropriate constraint on the range.
 
-    A list of tuples containing the heading, the level, and the span (the start
-    offset and the end offset as a tuple) is returned.
+    A list of tuples is returned for each section divider. For headings, the
+    heading text and level number are returned as the first and second elements
+    of each tuple, whereas other section dividers (typically indicating the end
+    of a section) employ None as the first and second elements. The span (start
+    offset and end offset of the divider) is provided as a tuple in the third
+    element of each result tuple.
     """
 
     headings = []
 
-    for match in heading_regexp.finditer(body):
-        level = len(match.group("level"))
+    for match in section_regexp.finditer(body):
+        level = match.group("level")
+        level = level and len(match.group("level"))
 
         if (min_level is None or min_level <= level) and \
-            (max_level is None or level <= max_level):
+            (max_level is None or level <= max_level) or \
+            level is None:
 
             headings.append((match.group("heading"), level, match.span()))
 
diff -r 5e3955c2714b -r a2e8615152cf actions/SectionBreakout.py
--- a/actions/SectionBreakout.py	Sat Feb 26 21:45:56 2011 +0100
+++ b/actions/SectionBreakout.py	Sun Sep 11 18:50:16 2011 +0200
@@ -33,7 +33,7 @@
         # Acquire heading details from the page.
 
         body = page.get_raw_body()
-        heading_details = getHeadingDetails(body, level, level)
+        heading_details = getSectionDetails(body, level, level)
 
         d = {
             "buttons_html"                  : buttons_html,
@@ -55,7 +55,8 @@
       <td>''' % d
 
         for heading, level, span in heading_details:
-            html += "%s<br />" % heading
+            if heading is not None:
+                html += "%s<br />" % escape(heading)
 
         html += '''
       </td>
@@ -122,25 +123,36 @@
         regions = []
         current_region_start = None
 
-        for heading, found_level, (start, end) in getHeadingDetails(page_body):
+        for heading, found_level, (start, end) in getSectionDetails(page_body):
 
-            # Upon finding a suitable heading, begin a new region to be broken
-            # out.
+            # Where a heading is provided, consider starting a section.
+
+            if found_level is not None:
 
-            if current_region_start is None and found_level >= level:
-                current_region_start = heading, start
+                # Upon finding a suitable heading, begin a new region to be broken
+                # out.
 
-            # Upon finding a higher-level heading, end any open region.
+                if current_region_start is None and found_level >= level:
+                    current_region_start = heading, start
+
+                # Upon finding a same-level or higher-level heading, end any open
+                # region.
 
-            elif current_region_start is not None and found_level <= level:
-                regions.append(current_region_start + (start,))
+                elif current_region_start is not None and found_level <= level:
+                    regions.append(current_region_start + (start,))
 
-                # For headings at the requested level, open a new region.
+                    # For headings at the requested level, open a new region.
 
-                if found_level == level:
-                    current_region_start = heading, start
-                else:
-                    current_region_start = None
+                    if found_level == level:
+                        current_region_start = heading, start
+                    else:
+                        current_region_start = None
+
+            # Where no heading is provided, end the section.
+
+            elif current_region_start is not None:
+                regions.append(current_region_start + (start,))
+                current_region_start = None
 
         # End any open region.