parsers: add gemtext parser

mdiocre/parsers/gem.py

@@ -0,0 +1,136 @@
+import re
+import random
+import xml.etree.ElementTree as ET
+from . import BaseParser, sub_func
+
+class GemParser(BaseParser):
+	'''
+	Gemtext parser. Comments are parsed the same way as Zim does
+	'''
+
+	RE_COMMENTS = re.compile(r'\[mdiocre:(.+?)\]')
+
+	def convert_markup(self, markup):
+		# read line by line
+		MARKUP = markup.split('\n')
+		P_ADD_RE         = re.compile(r'^(?!#|\*|>|```|=>)(.+)$')
+		HEADINGS_RE      = re.compile(r'^(#{1,6})\s+(.+)$') # spec says mandatory space chara
+		LINKS_RE         = re.compile(r'^=>\s*(\w+://[^\s]+)(\s+(.+$))?')
+		PREFORMATTED_RE  = re.compile(r'^```(.+)$')
+		ULIST_RE         = re.compile(r'^\*\s+(.+)$')
+		LINKS_INSIDE     = re.compile(r'\(=>\s*(\w+://[^\s]+)(\s+(.+?))?\)') 
+		BLOCKQUOTE_RE    = re.compile(r'^>\s+(.+)$')
+
+		output_markup = []
+		lists_mode = {
+			'ul': False,
+			'ol': False
+		}
+		pre_mode = False
+		cur_code_hash = ''
+		code_alts = {}
+
+		def make_code_hash():
+			new_key = '__code__'
+			for i in random.randbytes(32):
+				new_key += hex(i)[2:]
+			return new_key
+
+		for i in range(len(MARKUP)):
+			line = MARKUP[i]
+
+			if not pre_mode: # standard text
+				# add paragraph tags
+				if re.search(P_ADD_RE, line):
+					line = '<p>' + line + '</p>'
+
+				# add header tags
+				line = re.sub(HEADINGS_RE,
+					lambda a: "<h%d>%s</h%d>" % (len(a.group(1)), a.group(2), len(a.group(1))),
+					line
+				)
+
+				# add blockquotes
+				line = re.sub(BLOCKQUOTE_RE,
+					lambda a: "<blockquote>%s</blockquote>" % (a.group(1)),
+					line
+				)
+
+				is_list_or_link_set = (bool(re.match(ULIST_RE, line)) or bool(re.match(LINKS_RE, line)))
+
+				# lists
+				if re.match(ULIST_RE, line):
+					if not lists_mode['ul']:
+						lists_mode['ul'] = True
+						output_markup.append('<ul>')
+					line = re.sub(ULIST_RE,
+						lambda a: "<li>%s</li>" % (a.group(1)),
+						line
+					)
+
+				if not is_list_or_link_set:
+					if lists_mode['ul']:
+						lists_mode['ul'] = False
+						output_markup.append('</ul>')
+
+				# links
+				def gen_link(result):
+					link_str = '<a href="%s">%s</a>'
+					if result.group(3):
+						return link_str % (result.group(1), result.group(3))
+					# no unique txt
+					return link_str % (result.group(1), result.group(1))
+
+				# links on its own is treated like lists
+				if re.match(LINKS_RE, line):
+					if not lists_mode['ul']:
+						lists_mode['ul'] = True
+						output_markup.append('<ul class="gemtext-links">')
+					line = "<li>%s</li>" % (re.sub(LINKS_RE, gen_link, line))
+
+				line = re.sub(LINKS_INSIDE, gen_link, line)
+
+				if re.match(PREFORMATTED_RE, line):
+					cur_code_hash = make_code_hash()
+					line = '<pre alt="%s"><code>' % (cur_code_hash)
+					pre_mode = True
+
+				if False: # comment out this line to literally render blank lines as <br>
+					if len(line.strip()) == 0:
+						line = '<br>'
+			else: # preformatted text
+				code_end = re.match(PREFORMATTED_RE, line)
+				if code_end:
+					line = '</code></pre>'
+					pre_mode = False
+					if code_end.group(1):
+						code_alts[cur_code_hash] = code_end.group(1)
+
+			output_markup.append(line)
+		output_str = '\n'.join(output_markup)
+
+		# resolve code alt text
+		for key, value in code_alts.items():
+			output_str = output_str.replace(key, value)
+
+		return output_str
+
+	def to_variables(self, html, v, ignore_content=False):
+		def trf_sub_func(match):
+			return sub_func(match, v)
+
+		# do substitution...
+		gmitxt = re.sub(self.RE_COMMENTS, trf_sub_func, html)
+
+		html = self.convert_markup(gmitxt)
+		# escape all text
+		etr = ET.fromstring("<_doc_>%s</_doc_>" % html)
+		html = '\n'.join(
+			ET.tostring(etr, encoding='unicode', method='html')\
+			.split('\n')[1:-1]
+		)
+
+		if not ignore_content:
+			v.variables["content"] = html
+
+		return v

mdiocre/wizard.py

@@ -24,7 +24,9 @@ class Wizard():
 	              'rst'  : 'rst',
 	              'html' : 'html',
 	              'htm'  : 'html',
-	              'zimtxt'  : 'zim'
+	              'zimtxt'  : 'zim',
+	              'gem'  : 'gem',
+	              'gmi'  : 'gem',
 	              }
 
 	def __init__(self):