added basic Markdown parsing and some support for Obsidian constructs

src/dragonglass/dragonglass.py

@@ -4,6 +4,7 @@ import argparse
 import sys
 from pathlib import Path
 from tree import SourceNode
+from mparse import create_markdown_parser
 
 # The command line parser
 parser = argparse.ArgumentParser(prog='dragonglass')
@@ -28,11 +29,19 @@ def main():
     for node in nodes:
         node.load_metadata(source_dir)
 
+    mdparse = create_markdown_parser()
+    for node in nodes:
+        node.parse_markdown(source_dir, mdparse)
+
     # TEMP
     for node in nodes:
         print(node)
         if node.metadata:
             print(f"Metadata: {node.metadata}")
+        if node.text:
+            print("----- BEGIN TEXT -----")
+            print(node.text)
+            print("------ END TEXT ------")
 
     return 0
 

src/dragonglass/mparse.py

@@ -0,0 +1,54 @@
+#/usr/bin/env python3
+
+import markdown
+import xml.etree.ElementTree as etree
+from markdown.extensions import Extension
+from markdown.inlinepatterns import InlineProcessor, SimpleTagInlineProcessor
+from markdown.preprocessors import Preprocessor
+
+
+class MetaStripper(Extension):
+
+    class MetaStripperProc(Preprocessor):
+        def run(self, lines):
+            if lines[0] == '---':
+                lines.pop(0)
+                while lines[0] != '---':
+                    lines.pop(0)
+                lines.pop(0)
+            return lines
+
+    def extendMarkdown(self, md):
+        md.preprocessors.register(MetaStripper.MetaStripperProc(md), 'metastripper', 0)
+
+
+class ObsidianLinks(Extension):
+
+    class ObsidianLinksProc(InlineProcessor):
+        def __init__(self, pattern, md):
+            super(ObsidianLinks.ObsidianLinksProc, self).__init__(pattern, md)
+
+        def handleMatch(self, m, data):
+            dstr = m.group(0)[2:-2]
+            el = etree.Element('a')
+            el.set('href', dstr)
+            el.text = dstr
+            return el, m.start(0), m.end(0)
+
+    def extendMarkdown(self, md):
+        OBSLINK_PATTERN = r'\[\[(.*?)\]\]'
+        md.inlinePatterns.register(ObsidianLinks.ObsidianLinksProc(OBSLINK_PATTERN, md), 'obsidian_links', 0)
+
+
+class Strikeout(Extension):
+    def extendMarkdown(self, md):
+        md.inlinePatterns.register(SimpleTagInlineProcessor(r'()~~(.*?)~~', 'del'), 'strikeout', 0)
+
+
+class Highlight(Extension):
+    def extendMarkdown(self, md):
+        md.inlinePatterns.register(SimpleTagInlineProcessor(r'()\=\=(.*?)\=\=', 'ins'), 'highlight', 0)
+
+
+def create_markdown_parser():
+    return markdown.Markdown(extensions=[MetaStripper(), ObsidianLinks(), Strikeout(), Highlight()])

src/dragonglass/tree.py

@@ -16,6 +16,7 @@ class SourceNode:
         self._is_dir = is_dir
         self._is_md = path.match(MARKDOWN_PAT)
         self.metadata = None
+        self.text = None
 
     def __str__(self):
         return f"SourceNode({self._path}, {self._is_dir}) [is_md={self._is_md}]"
@@ -52,3 +53,9 @@ class SourceNode:
                         metalines.append(l)
                         l = f.readline()
                     self.metadata = yaml.full_load(''.join(metalines))
+
+    def parse_markdown(self, source_dir, markdown_parser):
+        if self._is_md:
+            markdown_parser.reset()
+            with open(source_dir / self._path, "r", encoding="utf-8") as f:
+                self.text = markdown_parser.convert(f.read())