[Wiki] Some excellent shitcode for document TOCs

5 files changed, 96 insertions, 5 deletions

diff --git a/pysite/rst/__init__.py b/pysite/rst/__init__.py
index e0fc973e..0c069615 100644
--- a/pysite/rst/__init__.py
+++ b/pysite/rst/__init__.py
@@ -1,15 +1,83 @@
 # coding=utf-8
+import re
+
 from docutils.core import publish_parts
 from docutils.parsers.rst.roles import register_canonical_role
 
 from pysite.rst.roles import icon_role, page_role, url_for_role
 
+RST_TEMPLATE = """.. contents::
+
+{0}"""
+
+CONTENTS_REGEX = re.compile(r"""<div class=\"contents topic\" id=\"contents\">(.*?)</div>""", re.DOTALL)
+HREF_REGEX = re.compile(r"""<a class=\"reference internal\" href=\"(.*?)\".*?>(.*?)</a>""")
+
 
 def render(rst: str):
-    return publish_parts(
+    rst = RST_TEMPLATE.format(rst)
+    html = publish_parts(
         source=rst, writer_name="html5", settings_overrides={"halt_level": 2, "syntax_highlight": "short"}
     )["html_body"]
 
+    data = {
+        "html": html,
+        "headers": []
+    }
+
+    match = CONTENTS_REGEX.search(html)  # Find the contents HTML
+
+    if match:
+        data["html"] = html.replace(match.group(0), "")  # Remove the contents from the document HTML
+
+        depth = 0
+        headers = []
+        current_header = {}
+
+        group = match.group(1)
+
+        # Sanitize the output so we can more easily parse it
+        group = group.replace("<li>", "<li>\n")
+        group = group.replace("</li>", "\n</li>")
+        group = group.replace("<p>", "<p>\n")
+        group = group.replace("</p>", "\n</p>")
+
+        for line in group.split("\n"):
+            line = line.strip()  # Remove excess whitespace
+
+            if not line:  # Nothing to process
+                continue
+
+            if line.startswith("<li>") and depth <= 2:
+                #  We've found a header, or the start of a header group
+                depth += 1
+            elif line.startswith("</li>") and depth >= 0:
+                # That's the end of a header or header group
+
+                if depth == 1:
+                    # We just dealt with an entire header group, so store it
+                    headers.append(current_header.copy())  # Store a copy, since we're clearing the dict
+                    current_header.clear()
+
+                depth -= 1
+            elif line.startswith("<a") and depth <= 2:
+                # We've found an actual URL
+                match = HREF_REGEX.match(line)  # Parse the line for the ID and header title
+
+                if depth == 1:  # Top-level header, so just store it in the current header
+                    current_header["id"] = match.group(1)
+                    current_header["title"] = match.group(2)
+                else:  # Second-level (or deeper) header, should be stored in a list of sub-headers under the current
+                    sub_headers = current_header.get("sub_headers", [])
+                    sub_headers.append({
+                        "id": match.group(1),
+                        "title": match.group(2)
+                    })
+                    current_header["sub_headers"] = sub_headers
+
+        data["headers"] = headers
+    return data
+
 
 register_canonical_role("icon", icon_role)
 register_canonical_role("page", page_role)
diff --git a/pysite/views/wiki/edit.py b/pysite/views/wiki/edit.py
index a111f9ce..a2e98e20 100644
--- a/pysite/views/wiki/edit.py
+++ b/pysite/views/wiki/edit.py
@@ -35,11 +35,14 @@ class EditView(RouteView, DBMixin):
     @csrf
     def post(self, page):
         rst = request.form["rst"]
+        rendered = render(rst)
+
         obj = {
             "slug": page,
             "title": request.form["title"],
             "rst": rst,
-            "html": render(rst)
+            "html": rendered["html"],
+            "headers": rendered["headers"]
         }
 
         self.db.insert(
diff --git a/pysite/views/wiki/render.py b/pysite/views/wiki/render.py
index aa365c4a..9d3e8cc3 100644
--- a/pysite/views/wiki/render.py
+++ b/pysite/views/wiki/render.py
@@ -30,7 +30,7 @@ class RenderView(APIView):
 
         data = data[0]["data"]
         try:
-            html = render(data)
+            html = render(data)["html"]
 
             return jsonify({"data": html})
         except SystemMessage as e:
@@ -51,7 +51,7 @@ class RenderView(APIView):
                 if match:
                     data["error_lines"].append(
                         {
-                            "row": int(match.group(1)) - 1,
+                            "row": int(match.group(1)) - 3,
                             "column": 0,
                             "type": "error",
                             "text": match.group(2)
diff --git a/pysite/views/ws/rst.py b/pysite/views/ws/rst.py
index 19c4129b..24bdb3ca 100644
--- a/pysite/views/ws/rst.py
+++ b/pysite/views/ws/rst.py
@@ -23,7 +23,7 @@ class RSTWebsocket(WS):
         self.log.debug(f"RST | Message: {message}")
 
         try:
-            data = render(message)
+            data = render(message)["html"]
         except Exception as e:
             self.log.exception("Parsing error")
             data = str(e)
diff --git a/templates/wiki/base.html b/templates/wiki/base.html
index e400be02..60f031b9 100644
--- a/templates/wiki/base.html
+++ b/templates/wiki/base.html
@@ -33,6 +33,26 @@
             <div class="uk-flex uk-flex-row uk-flex-1">
                 <div class="uk-card uk-card-body uk-flex-left uk-flex uk-card-primary">
                     <ul class="uk-nav-default uk-nav-parent-icon" uk-nav>
+                        {% if data is defined %}
+                            {% if "headers" in data %}
+                                <li class="uk-nav-header">Contents</li>
+                                {% for header in data.headers %}
+                                    {%  if "sub_headers" in header %}
+                                        <li class="uk-parent">
+                                            <a href="{{ header.id }}">{{ header.title }}</a>
+                                            <ul class="uk-nav-sub">
+                                                {% for sub in header.sub_headers %}
+                                                    <li><a href="{{ sub.id }}"><i class="uk-icon fas fa-diamond fa-fw" style="font-size: 0.5rem; vertical-align: 1px"></i> &nbsp;{{ sub.title }}</a></li>
+                                                {% endfor %}
+                                            </ul>
+                                        </li>
+                                    {% else %}
+                                        <li><a href="{{ header.id }}">{{ header.title }}</a></li>
+                                    {% endif %}
+                                {% endfor %}
+                            {% endif %}
+                            <li class="uk-nav-divider"></li>
+                        {% endif %}
                         <li><a href="{{ url_for("wiki.page", page="home") }}">
                             <i class="uk-icon fas fa-fw fa-home"></i> &nbsp;Home
                         </a></li>