Working parser/converter

The parser now works as expected with no known bugs.

As expected from the spec each line of ordinary text is treated as a separate
paragraph.

.gitignore

@@ -0,0 +1,11 @@
+.idea/.name
+.idea/gemtext2html.iml
+.idea/misc.xml
+.idea/modules.xml
+.idea/vcs.xml
+.idea/inspectionProfiles/profiles_settings.xml
+.idea/inspectionProfiles/Project_Default.xml
+.vscode/launch.json
+.vscode/settings.json
+.vscode/.ropeproject/config.py
+.vscode/.ropeproject/objectdb

.idea/.gitignore

@@ -0,0 +1,5 @@
+# Default ignored files
+/shelf/
+/workspace.xml
+/.idea/
+/.vscode/

src/convert.py

@@ -24,81 +24,94 @@ class Tag:
 class GemParser:
     def __init__(self):
         self.mstate = State.INITIAL
-        self.toggle = False
+        self.recurse = False
 
     def parse_line(self, line):
         rstring = str()
+        is_text = self.is_text(line)
+        starts_with_pre = line.startswith(Tag.PRE)
+        starts_with_ul = line.startswith(Tag.UNORDERED)
+        starts_with_quote = line.startswith(Tag.QUOTE)
+        if line == '\n' and self.mstate != State.PRETEXT:
+            return rstring
+        # This is blockquote end tag
+        if self.mstate == State.QUOTE and not starts_with_quote:
+            rstring += self.get_end_tag()
+            self.recurse = True
+        # This is paragraph end tag
+        if self.mstate == State.TEXT and not is_text:
+            rstring += self.get_end_tag()
+            self.recurse = True
+        # This is ul end tag
+        if self.mstate == State.UNORDERED and not starts_with_ul:
+            rstring += self.get_end_tag()
+            self.recurse = True
         # This is a pre start tag
-        if line.startswith(Tag.PRE) and not self.toggle:
+        if starts_with_pre and self.mstate != State.PRETEXT:
             self.mstate = State.PRETEXT
-            self.toggle = True
+            rstring += self.get_start_tag().format(' aria-label="{}"'.format(
-            rstring = self.get_start_tag().format(' aria-label="{}"'.format(
                 line[4:].rstrip('\n')))
+            return rstring
+        # This is in pre formatted text
+        elif self.mstate == State.PRETEXT and not starts_with_pre:
+            rstring += line
+            return rstring
         # This is a pre end tag
-        elif line.startswith(Tag.PRE):
+        elif starts_with_pre and self.mstate == State.PRETEXT:
-            rstring = self.get_end_tag()
+            rstring += self.get_end_tag()
             self.mstate = State.INITIAL
-            self.toggle = False
+            return rstring
-        # This is in pre formatted text
-        elif self.mstate == State.PRETEXT:
-            rstring = line
         else:
+            # This is paragraph start tag
+            if is_text and self.mstate != State.TEXT:
+                self.mstate = State.TEXT
+                rstring += self.get_start_tag()
+                rstring += line
+                return rstring
+            # This is in paragraph
+            if is_text and self.mstate == State.TEXT:
+                rstring += line
+                return rstring
+            # This is a heading
+            if line.startswith(Tag.HEADING):
+                self.mstate = State.HEADING
+                rstring += self.parse_heading(line)
+                self.mstate = State.INITIAL
+                return rstring
+            # This is a link
+            if line.startswith(Tag.LINK):
+                self.mstate = State.LINK
+                rstring += self.parse_link(line)
+                self.mstate = State.INITIAL
+                return rstring
             # This is ul start tag
-            if line.startswith(Tag.UNORDERED) and not self.toggle:
+            if starts_with_ul and not self.mstate == State.UNORDERED:
                 self.mstate = State.UNORDERED
-                self.toggle = True
+                rstring += self.get_start_tag()
-                rstring = self.get_start_tag()
                 rstring += '<li>{}</li>\n'.format(line[1:].strip())
+                return rstring
             # This is in middle of unordered list
-            elif line.startswith(Tag.UNORDERED):
+            if starts_with_ul and self.mstate == State.UNORDERED:
                 rstring += '<li>{}</li>\n'.format(line[1:].strip())
-            # This is ul end tag
+                return rstring
-            elif self.mstate == State.UNORDERED and not line.startswith(
-                    Tag.UNORDERED):
-                rstring += self.get_end_tag()
-                self.mstate = State.INITIAL
-                self.toggle = False
-                # Recurse so we don't miss whats next
-                rstring += self.parse_line(line)
             # This is blockquote start tag
-            if line.startswith(Tag.QUOTE) and not self.toggle:
+            if starts_with_quote and self.mstate != State.QUOTE:
                 self.mstate = State.QUOTE
-                self.toggle = True
+                rstring += self.get_start_tag()
-                rstring = self.get_start_tag()
                 rstring += line[1:]
+                return rstring
             # This is in quote
-            elif line.startswith(Tag.QUOTE):
+            if starts_with_quote:
                 rstring += line[1:]
-            # FIXME: If a list is placed directly after a quote there will ba a list item in the quote...
+                return rstring
-            # This is blockquote end tag
+        if self.recurse:
-            elif self.mstate == State.QUOTE and not line.startswith(Tag.QUOTE):
+            self.recurse = False
-                rstring += self.get_end_tag()
-                self.mstate = State.INITIAL
-                self.toggle = False
-                # Recurse so we don't miss whats next
-                rstring += self.parse_line(line)
-            # This is paragraph start tag
-            if self.is_text(line) and not self.toggle:
-                self.mstate = State.TEXT
-                self.toggle = True
-                rstring = self.get_start_tag()
-                rstring += line
-            # This is in paragraph
-            elif self.is_text(line):
-                rstring += line
-            # This is paragraph end tag
-            elif self.mstate == State.TEXT:
-                rstring += self.get_end_tag()
             self.mstate = State.INITIAL
-                self.toggle = False
             # Recurse so we don't miss whats next
-                rstring += self.parse_line(line)
+            return "{}{}".format(rstring, self.parse_line(line))
-            # This is a link
+        else:
-            if line.startswith(Tag.LINK):
+            print("We should never be here, the line is: {}".format(line))
-                self.mstate = State.LINK
+            sys.exit(1)
-                rstring = self.parse_link(line)
-                self.mstate = State.INITIAL
-        return rstring
 
     def parse_link(self, line):
         linearr = line[2:].strip().split(' ')
@@ -146,13 +159,20 @@ class GemParser:
         with open(filename) as gemtext:
             mline: str = gemtext.readline()
             while mline:
-                rdocument += self.parse_line(mline)
+                rline = self.parse_line(mline)
+                if rline is not None and rline != str() and rline != '\n':
+                    rdocument += rline
                 mline = gemtext.readline()
-            if self.mstate != State.INITIAL:
+        rdocument += '{}</body>\n</html>\n'.format(self.get_end_tag())
-                rdocument += self.get_end_tag()
-        rdocument += '</body>\n</html>\n'
         return rdocument
 
+    def parse_heading(self, line, level=0):
+        if line.startswith(Tag.HEADING):
+            return self.parse_heading(line[1:], level + 1)
+        else:
+            return "{}{}{}".format(self.get_start_tag().format(level), line.strip().rstrip('\n'),
+                                   self.get_end_tag().format(level))
+
 
 if __name__ == '__main__':
     mparser: GemParser = GemParser()