Implement json TOC

4 files changed, 75 insertions, 34 deletions

diff --git a/panfry/cli.py b/panfry/cli.py
index 6949474..650b724 100755
--- a/panfry/cli.py
+++ b/panfry/cli.py
@@ -107,6 +107,14 @@ Args:
         
         Example --pandoc-options="smart base-header-level=2"
 
+  json_toc:
+    flags: [--json-toc]
+    action: store_true
+    default: False
+    help: |
+        Generate toc.json file instead of toc integrated in
+        the html
+
   port:
     flags: [-P, --port]
     default: 8080
@@ -129,6 +137,7 @@ Subparsers:
       - pub_path
       - templates_path
       - simple_toc
+      - json_toc
       - assets
       - clean
       - css
diff --git a/panfry/document.py b/panfry/document.py
index 0ea619d..035bf56 100755
--- a/panfry/document.py
+++ b/panfry/document.py
@@ -5,7 +5,8 @@ import re
 import pandoc
 from tempfile import NamedTemporaryFile
 from panfry.util import *
-from panfry.page import TextFile, Page
+from panfry.page import TextFile, Page, TOCItem
+from panfry.toc import TOCItem
 
 
 class Document:
@@ -15,6 +16,8 @@ class Document:
         self.meta = self.get_meta(self.src_path)
         self.pages = self.get_pages(self.src_path)
         self.full_toc = True
+        self.standalone = True
+        self.json_toc = False
 
         self.css_file = 'css/style.css'
         self.assets_dir = 'assets'
@@ -24,6 +27,11 @@ class Document:
     def set_simple_toc(self):
             self.full_toc = False
 
+    @property
+    def set_json_toc(self):
+            self.json_toc = True
+            self.standalone = False
+
     def set_pandoc_options(self, options):
             self.pandoc_options = options
 
@@ -62,10 +70,10 @@ class Document:
         for page in self.pages:
             for tocitem in page.toc:
                 if tocitem.heading == heading.replace('\n', ' '):
-                    return '[%s](%s#%s)' % (heading,
-                                            page.htmlfile,
-                                            tocitem.header_id,
-                                            )
+                    return '[%s](%s%s)' % (heading,
+                                           page.htmlfile,
+                                           tocitem.header_id,
+                                           )
 
         return '[%s]' % ref_txt
 
@@ -143,7 +151,7 @@ class Document:
                 toc = [page.toc[0]]
 
             for tocitem in toc:
-                link = '%s#%s' % (page.htmlfile, tocitem.header_id)
+                link = '%s%s' % (page.htmlfile, tocitem.header_id)
                 links.append(dict(link=unicode(link, "utf8"),
                                   text=unicode(tocitem.heading, "utf8"),
                                   level=tocitem.level,
@@ -152,6 +160,37 @@ class Document:
         return links
 
     @property
+    def toc_json(self):
+        lastchild = top = toc = TOCItem(self.title, 1, 'title_block.html')
+
+        curlevel = 0
+        for page in self.pages:
+            heading = ''
+            for line in page.markdown.split('\n'):
+                line = line.strip()
+                if heading and re.match('[=]{2}', line):
+                    level = 1
+                elif heading and re.match('[-]{2}', line):
+                    level = 2
+                elif re.match('#+.+[A-z|0-9]', line):
+                    level = len(re.match('#+', line).group())
+                    heading = line.split(' ', 1)[1].strip()
+                else:
+                    heading = line.strip()
+                    continue
+
+                if curlevel < level:
+                    toc = lastchild
+                    curlevel = toc.level
+                while toc.parent and curlevel >= level:
+                    toc = toc.parent
+                    curlevel = toc.level
+
+                lastchild = toc.add_child(heading, level, page.htmlfile)
+
+        return top.json
+
+    @property
     def filename_base(self):
         filename = self.title
         filename = re.sub(r'\s+', ' ', filename)
@@ -231,24 +270,35 @@ class Document:
             doc.add_argument('toc')
             doc.add_argument('template=%s' % template_file.name)
             doc.add_argument('css=%s' % self.css_file)
+
             for option in self.pandoc_options:
                 if option == 'number-sections':
                     continue
                 doc.add_argument(option)
-            doc.markdown = '%s\n%s' % (self.header.source,
-                                       self.expand_int_links(page.markdown))
+
+            if self.standalone:
+                markdown = '%s\n' % self.title_block.source
+            else:
+                markdown = ''
+
+            markdown += self.expand_int_links(page.markdown)
+            doc.markdown = markdown
             content = doc.html5
             write_file(os.path.join(pub_path, page.htmlfile),
                        unicode(content, 'utf-8'))
 
         # If there is not explicit index.html, then link 'index.html'
         # to the toplevel page.
-        if not 'index.md' in self.pages:
+        if self.standalone and not 'index.md' in self.pages:
             src = os.path.join(self.pages[0].htmlfile)
             ref = os.path.join(pub_path, 'index.html')
             if os.path.exists(ref):
                 os.remove(ref)
             os.symlink(src, ref)
 
+        if self.json_toc:
+            write_file(os.path.join(pub_path, 'toc.json'),
+                       unicode(self.toc_json, 'utf-8'))
+
         ###### Copy any assets to publish directory
         self.publish_assets(pub_path)
diff --git a/panfry/main.py b/panfry/main.py
index 40c2b15..9572243 100755
--- a/panfry/main.py
+++ b/panfry/main.py
@@ -62,6 +62,9 @@ def gen(env):
     if env.simple_toc:
         document.set_simple_toc
 
+    if env.json_toc:
+        document.set_json_toc
+
     if env.clean and os.path.exists(env.pub_path):
         print("Cleanup, removing %s..." % env.pub_path)
         shutil.rmtree(env.pub_path)
diff --git a/panfry/page.py b/panfry/page.py
index 722241c..21b6323 100755
--- a/panfry/page.py
+++ b/panfry/page.py
@@ -2,6 +2,7 @@
 # -*- coding: utf-8 -*-
 import re
 import pandoc
+from panfry.toc import TOCItem
 
 
 class TextFile:
@@ -25,28 +26,6 @@ class TextFile:
         return self.lines[0]
 
 
-class TOCItem:
-    '''
-    Table of Contents Line Item
-    '''
-    def __init__(self, heading, level):
-        self.heading = heading
-        self.level = level
-
-    @property
-    def header_id(self):
-        # - 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)
-
-
 class Page(TextFile):
     '''
     Represents single page source file.
@@ -106,14 +85,14 @@ class Page(TextFile):
         for line in self.markdown.split('\n'):
             line = line.strip()
             if heading and re.match('[=]{2}', line):
-                toc.append(TOCItem(heading, 1))
+                toc.append(TOCItem(heading, 1, self.htmlfile))
                 continue
             if heading and re.match('[-]{2}', line):
-                toc.append(TOCItem(heading, 2))
+                toc.append(TOCItem(heading, 2, self.htmlfile))
                 continue
             if re.match('#+.+[A-z|0-9]', line):
                 level = len(re.match('#+', line).group())
-                toc.append(TOCItem(line.split(' ', 1)[1].strip(), level))
+                toc.append(TOCItem(line.split(' ', 1)[1].strip(), level, self.htmlfile))
                 continue
             heading = line.strip()