From 09c7b7abbdbf5b12f908d4a88fe5c0098094b2c7 Mon Sep 17 00:00:00 2001
From: Matt Singleton <matt@xcolour.net>
Date: Fri, 14 Jan 2022 10:30:10 -0600
Subject: switch to fsm implementation of gem2html

---
 fsm.py    | 86 +++++++++++----------------------------------------------------
 gemini.py | 81 +++++++++++------------------------------------------------
 2 files changed, 29 insertions(+), 138 deletions(-)

diff --git a/fsm.py b/fsm.py
index 40f9c2d..901e114 100644
--- a/fsm.py
+++ b/fsm.py
@@ -1,4 +1,6 @@
 import sys
+import urllib.parse
+
 
 class StackFSM(object):
     """
@@ -25,6 +27,7 @@ class StackFSM(object):
     def pop_state(self):
         return self._state_stack.pop()
 
+
 class Parser(object):
 
     def __init__(self, document, output=None):
@@ -38,29 +41,23 @@ class Parser(object):
 
     def parse(self):
         self._fsm.push_state(self.text_state)
-        while self._fsm._current_state() is not None:
+        while self._fsm._current_state() is not None and len(self._document) > self._offset:
             self._fsm.update()
 
     def text_state(self):
-        if len(self._document) <= self._offset:
-            self._fsm.pop_state()
-            return
         line = self._document[self._offset]
         if line.strip() == '':
             self._blanks += 1
         else:
             self._blanks = 0
         if line.strip() == '```':
-            self._fsm.pop_state()
             self._fsm.push_state(self.pre_state)
             self._output.write('<pre>\n')
             self._offset += 1
         elif line.startswith('* '):
-            self._fsm.pop_state()
             self._fsm.push_state(self.list_state)
             self._output.write('<ul>\n')
         elif line.startswith('=>'):
-            self._fsm.pop_state()
             self._fsm.push_state(self.link_state)
             self._output.write('<ul>\n')
         else:
@@ -80,95 +77,40 @@ class Parser(object):
             self._offset += 1
 
     def pre_state(self):
-        if len(self._document) < self._offset:
-            self.pop_state()
-            return
         line = self._document[self._offset]
         if line.strip() == '```':
             self._fsm.pop_state()
-            self._fsm.push_state(self.text_state)
             self._output.write('</pre>\n')
             self._offset += 1
-        elif line.startswith('* '):
-            self._fsm.pop_state()
-            self._fsm.push_state(self.list_state)
-            self._output.write('<ul>\n')
-        elif line.startswith('=>'):
-            self._fsm.pop_state()
-            self._fsm.push_state(self.link_state)
-            self._output.write('<ul>\n')
         else:
             self._output.write(line + '\n')
             self._offset += 1
 
     def list_state(self):
-        if len(self._document) < self._offset:
-            self.pop_state()
-            return
         line = self._document[self._offset]
         if line.startswith('* '):
             self._output.write('<li>{}</li>\n'.format(line[2:]))
             self._offset += 1
         else:
             self._fsm.pop_state()
-            self._fsm.push_state(self.text_state)
             self._output.write('</ul>\n')
 
     def link_state(self):
-        if len(self._document) < self._offset:
-            self.pop_state()
-            return
         line = self._document[self._offset]
         if line.startswith('=>'):
-            parts = line[2:].split(None, 2)
+            parts = line[2:].split(None, 1)
+            url = parts[0]
+            url_parts = urllib.parse.urlsplit(url)
+            if url_parts.scheme in ('gemini', ''):
+                external = ''
+            else:
+                external = ' class="external"'
             if len(parts) == 1:
-                self._output.write('<li><a href="{}">{}</a></li>\n'.format(parts[0], parts[0]))
+                text = url
             else:
-                self._output.write('<li><a href="{}">{}</a></li>\n'.format(parts[0], parts[1]))
+                text = parts[1]
+            self._output.write('<li{}><a href="{}">{}</a></li>\n'.format(external, url, text))
             self._offset += 1
         else:
             self._fsm.pop_state()
-            self._fsm.push_state(self.text_state)
             self._output.write('</ul>\n')
-
-document = """
-# h1
-hello
-hello
-
-## h2
-```
-code
-code
-```
-### h3
-hello
-
-
-hello
-
-### lists
-* hello
-* two
-* three
-
-text
-
-* one
-* two
-* three
-
-### links
-=>https://example.com hello
-=> https://example.com two
-=>  https://example.com three
-
-text
-
-=>https://example.com
-=> https://example.com
-=>  https://example.com
-"""
-
-p = Parser(document.split('\n'))
-p.parse()
diff --git a/gemini.py b/gemini.py
index 560b01f..cbc2eca 100644
--- a/gemini.py
+++ b/gemini.py
@@ -1,80 +1,26 @@
+import io
 import re
 import socket
 import ssl
 import urllib.parse
 
-import pyphen
-dic = pyphen.Pyphen(lang='en_US')
+import fsm
+
 
 def htmlescape(text: str) -> str:
     return text.replace('<', '&lt;').replace('>', '&gt;')
 
-def gem2html(gem: str) -> str:
-    html = []
-    html.append('<html>\n<body>\n<div id="root">')
-    state = 'text'
-    blanklines = 0
-    for line in gem.split('\n'):
-        if line.startswith('```'):
-            if state == 'pre':
-                newstate = 'text'
-                blanklines = 0
-            else:
-                newstate = 'pre'
-        elif state == 'pre':
-            newstate = 'pre'
-        elif line.startswith('=>'):
-            newstate = 'links'
-        elif line.startswith('* '):
-            newstate = 'list'
-        else:
-            newstate = 'text'
-
-        if state != 'pre':
-            if len(line.strip()) == 0:
-                blanklines += 1
-                if blanklines > 1:
-                    html.append('<br/>')
-                continue
-            blanklines = 0
-
-        if state != newstate:
-            if state in ('links', 'list'):
-                html.append('</ul>')
-            elif state == 'pre':
-                html.append('</pre>')
 
-            if newstate in ('links', 'list'):
-                html.append('<ul>')
-            elif newstate == 'pre':
-                html.append('<pre>')
-        state = newstate
+def gem2html(gem: str) -> str:
+    html = io.StringIO()
+    html.write('<html>\n<head>\n<style type="text/css">\n')
+    html.write(open('style.css').read())
+    html.write('</style>\n</head>\n<body>\n<div id="root">')
+    parser = fsm.Parser(gem.split('\n'), html)
+    parser.parse()
+    html.write('</div>\n</body>\n</html>')
+    return html.getvalue()
 
-        if line.startswith('```'):
-            pass
-        elif state == 'links':
-            tokens = line.split(None, 2)
-            url = tokens[1]
-            text = None if len(tokens) < 3 else tokens[2]
-            url_parts = urllib.parse.urlsplit(url)
-            external = ' class="external"' if url_parts.scheme not in ('gemini', '') else ''
-            html.append('<li{external}><a href="{url}">{text}</a></li>'.format(url=url, text=text or url, external=external))
-        elif state == 'list':
-            html.append('<li>{}</li>'.format(line[2:]))
-        elif state == 'pre':
-            html.append(line)
-        else:
-            if line.startswith('###'):
-                html.append('<p class="h3">{}</p>'.format(line[3:].lstrip()))
-            elif line.startswith('##'):
-                html.append('<p class="h2">{}</p>'.format(line[2:].lstrip()))
-            elif line.startswith('#'):
-                html.append('<p class="h1">{}</p>'.format(line[1:].lstrip()))
-            else:
-                ' '.join([dic.inserted(word, "\u00AD") for word in line.split()])
-                html.append('<p>{}</p>'.format(htmlescape(line)))
-    html.append('</div>\n</body>\n</html>')
-    return '\n'.join(html)
 
 def urljoin(base: str, url: str) -> str:
     if base is None:
@@ -83,6 +29,7 @@ def urljoin(base: str, url: str) -> str:
     url = re.sub('^gemini:', 'http:', url)
     return re.sub('^http:', 'gemini:', urllib.parse.urljoin(base, url))
 
+
 def get(url: str, follow_redirects: bool = True) -> dict:
     response = _get(url)
     if follow_redirects is True:
@@ -95,6 +42,7 @@ def get(url: str, follow_redirects: bool = True) -> dict:
             response = _get(response['meta'])
     return response
 
+
 def _parse_meta(meta: str) -> dict:
     mime, _, params_text = meta.lower().strip().partition(';')
     params = {}
@@ -105,6 +53,7 @@ def _parse_meta(meta: str) -> dict:
     params['mime'] = mime.strip()
     return params
 
+
 def _get(url: str) -> dict:
     url_parts = urllib.parse.urlsplit(url)
     if len(url_parts.path) == 0:
-- 
cgit v1.2.3