# HG changeset patch
# User Paul Boddie <paul@boddie.org.uk>
# Date 1422886699 -3600
# Node ID 11e412862d4515a5305c8e6689827645bfd8f720
# Parent  d54fd34e39c0bb0f89369c2031cff1841b52a1bd
Fixed handling of links in comment pages, which are represented by subpages in
Moin and therefore need deeper relative links.

diff -r d54fd34e39c0 -r 11e412862d45 TO_DO.txt
--- a/TO_DO.txt	Sat Jan 10 15:11:34 2015 +0100
+++ b/TO_DO.txt	Mon Feb 02 15:18:19 2015 +0100
@@ -1,3 +1,13 @@
+Recent Issues
+=============
+
+Linking in pages with slash characters in the actual titles. (Difficult to
+handle, especially if the pages eventually get renamed.)
+
+(Linking from comment subpages.)
+
+
+
 Enhancements
 ============
 
diff -r d54fd34e39c0 -r 11e412862d45 convert.py
--- a/convert.py	Sat Jan 10 15:11:34 2015 +0100
+++ b/convert.py	Mon Feb 02 15:18:19 2015 +0100
@@ -164,17 +164,19 @@
             if not body:
                 body = "## Empty page."
 
+            is_comment_page = content.get("content:class") == "Comment"
+
             # NOTE: Very simple technique employed for guessing the format.
 
             if no_translate:
-                fn = write
+                fn = notranslate
             elif body.startswith("<"):
                 fn = xmltranslate
             else:
                 fn = translate
 
             try:
-                fn(join(versions_dir, content["content"]), body)
+                fn(join(versions_dir, content["content"]), body, is_comment_page)
             except:
                 err = codecs.getwriter("utf-8")(sys.stderr)
                 print >>err, "Error parsing", content["content"]
@@ -210,7 +212,12 @@
 
         "Record properties in the current content dictionary."
 
-        self.content[attributes[-1]["name"]] = text.strip()
+        property_name = attributes[-1]["name"]
+        self.content[property_name] = text.strip()
+
+        property_class = attributes[-1].get("class")
+        if property_class:
+            self.content["%s:%s" % (property_name, "class")] = property_class.strip()
 
     def handle_id(self, name, elements, attributes, all_text, text):
 
@@ -272,7 +279,7 @@
     finally:
         f.close()
 
-def translate(filename, body, fn=None):
+def translate(filename, body, is_comment_page, fn=None):
 
     """
     Write to the file with the given 'filename' a translation of the given
@@ -284,12 +291,15 @@
     out = codecs.open(filename, "w", encoding="utf-8")
     try:
         print >>out, "#pragma page-filename", filename
-        fn(body, out)
+        fn(body, out, is_comment_page)
     finally:
         out.close()
 
-def xmltranslate(filename, body):
-    translate(filename, body, xmlparser.parse)
+def xmltranslate(filename, body, is_comment_page):
+    translate(filename, body, is_comment_page, xmlparser.parse)
+
+def notranslate(filename, body, is_comment_page):
+    write(filename, body)
 
 def sort_comments(pages_dir, pageid):
 
diff -r d54fd34e39c0 -r 11e412862d45 wikiparser.py
--- a/wikiparser.py	Sat Jan 10 15:11:34 2015 +0100
+++ b/wikiparser.py	Mon Feb 02 15:18:19 2015 +0100
@@ -3,7 +3,7 @@
 """
 Confluence Wiki syntax parsing.
 
-Copyright (C) 2012, 2013 Paul Boddie <paul@boddie.org.uk>
+Copyright (C) 2012, 2013, 2015 Paul Boddie <paul@boddie.org.uk>
 
 This software is free software; you can redistribute it and/or
 modify it under the terms of the GNU General Public License as
@@ -425,7 +425,8 @@
 
     "A parser for Confluence markup."
 
-    def __init__(self):
+    def __init__(self, is_comment_page=False):
+        self.is_comment_page = is_comment_page
         self.max_level = self.level = 0
         self.in_heading = False
         self.held_anchors = []
@@ -495,6 +496,8 @@
 
             else:
                 prefix = "../"
+                if self.is_comment_page:
+                    prefix += "../"
 
                 # Make the link tidier by making a target if none was given.
 
@@ -804,11 +807,11 @@
     def forbids_macros(self):
         return self.in_heading or self.macro
 
-def parse(s, out):
+def parse(s, out, is_comment_page=False):
 
     "Parse the content in the string 's', writing a translation to 'out'."
 
-    parser = ConfluenceParser()
+    parser = ConfluenceParser(is_comment_page)
     out.write(parser.parse_text(s, top=True))
 
 if __name__ == "__main__":
diff -r d54fd34e39c0 -r 11e412862d45 xmlparser.py
--- a/xmlparser.py	Sat Jan 10 15:11:34 2015 +0100
+++ b/xmlparser.py	Mon Feb 02 15:18:19 2015 +0100
@@ -139,9 +139,10 @@
 
     "Handle content from Confluence 4 page revisions."
 
-    def __init__(self, out):
+    def __init__(self, out, is_comment_page=False):
         Parser.__init__(self)
         self.out = out
+        self.is_comment_page = is_comment_page
 
         # Link target and label information.
 
@@ -302,6 +303,8 @@
                     prefix = link_target_prefixes.get(attrname)
                     if prefix:
                         target_details.insert(0, prefix)
+                        if self.is_comment_page:
+                            target_details.insert(0, prefix)
 
             # Make a link based on the details.
 
@@ -511,7 +514,7 @@
     def normalise(self, text, name):
         return normalise_regexp.sub(self.get_replacement(name), text)
 
-def parse(s, out):
+def parse(s, out, is_comment_page=False):
 
     "Parse the content in the string 's', writing a translation to 'out'."
 
@@ -530,7 +533,7 @@
 
     f = StringIO(s.encode("utf-8"))
     try:
-        parser = ConfluenceXMLParser(out)
+        parser = ConfluenceXMLParser(out, is_comment_page)
         parser.parse(f)
     finally:
         f.close()