Merge pull request #940 from jeffrimko/master

AsciiDoc: Support both legacy and asciidoctor via CLI subprocess

asciidoc_reader/README.rst

@@ -1,30 +1,30 @@
 AsciiDoc Reader
 ###############
 
-This plugin allows you to use `AsciiDoc <http://www.methods.co.nz/asciidoc/>`_ 
-to write your posts. File extension should be ``.asc``, ``.adoc``, 
+This plugin allows you to use `AsciiDoc <http://www.methods.co.nz/asciidoc/>`_
+to write your posts. File extension should be ``.asc``, ``.adoc``,
 or ``.asciidoc``.
 
 Dependency
 ----------
 
-If you want to use AsciiDoc you need to install it from `source
-<http://www.methods.co.nz/asciidoc/INSTALL.html>`_ or use your operating
-system's package manager.
+There are two command line utilities commonly used to render AsciiDoc:
+``asciidoc`` and ``asciidoctor``. One of the two will need to be installed and
+on the PATH.
 
-**Note**: AsciiDoc does not work with Python 3, so you should be using Python 2.
+**Note**: The ``asciidoctor`` utility is recommended since the original
+``asciidoc`` is no longer maintained.
 
 Settings
 --------
 
 ========================================  =======================================================
 Setting name (followed by default value)  What does it do?
 ========================================  =======================================================
+``ASCIIDOC_CMD = asciidoc``               Selects which utility to use for rendering. Will
+                                          autodetect utility if not provided.
 ``ASCIIDOC_OPTIONS = []``                 A list of options to pass to AsciiDoc. See the `manpage
                                           <http://www.methods.co.nz/asciidoc/manpage.html>`_.
-``ASCIIDOC_BACKEND = 'html5'``            Backend format for output. See the `documentation 
-                                          <http://www.methods.co.nz/asciidoc/userguide.html#X5>`_
-                                          for possible values.
 ========================================  =======================================================
 
 Example file header
@@ -45,5 +45,5 @@ Following the `example <https://github.com/getpelican/pelican/blob/master/docs/c
   :summary: Short version for index and feeds
 
   == title level 2
-    
+
   and so on...

asciidoc_reader/asciidoc_reader.py

@@ -3,61 +3,81 @@
 AsciiDoc Reader
 ===============
 
-This plugin allows you to use AsciiDoc to write your posts. 
+This plugin allows you to use AsciiDoc to write your posts.
 File extension should be ``.asc``, ``.adoc``, or ``asciidoc``.
 """
 
 from pelican.readers import BaseReader
-from pelican.utils import pelican_open
 from pelican import signals
-import six
+import os
+import re
+import subprocess
 
-try:
-    # asciidocapi won't import on Py3
-    from .asciidocapi import AsciiDocAPI, AsciiDocError
-    # AsciiDocAPI class checks for asciidoc.py
-    AsciiDocAPI()
-except:
-    asciidoc_enabled = False
-else:
-    asciidoc_enabled = True
+def call(cmd):
+    """Calls a CLI command and returns the stdout as string."""
+    return subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True).communicate()[0].decode('utf-8')
 
+def default():
+    """Attempt to find the default AsciiDoc utility."""
+    for cmd in ALLOWED_CMDS:
+        if len(call(cmd + " --help")):
+            return cmd
 
+ALLOWED_CMDS = ["asciidoc", "asciidoctor"]
+
+ENABLED = None != default()
 
 class AsciiDocReader(BaseReader):
-    """Reader for AsciiDoc files"""
+    """Reader for AsciiDoc files."""
 
-    enabled = asciidoc_enabled
+    enabled = ENABLED
     file_extensions = ['asc', 'adoc', 'asciidoc']
-    default_options = ["--no-header-footer", "-a newline=\\n"]
-    default_backend = 'html5'
+    default_options = ['--no-header-footer']
 
     def read(self, source_path):
-        """Parse content and metadata of asciidoc files"""
-        from cStringIO import StringIO
-        with pelican_open(source_path) as source:
-            text = StringIO(source.encode('utf8'))
-        content = StringIO()
-        ad = AsciiDocAPI()
-
-        options = self.settings.get('ASCIIDOC_OPTIONS', [])
-        options = self.default_options + options
-        for o in options:
-            ad.options(*o.split())
+        """Parse content and metadata of AsciiDoc files."""
+        cmd = self._get_cmd()
+        content = ""
+        if cmd:
+            optlist = self.settings.get('ASCIIDOC_OPTIONS', []) + self.default_options
+            options = " ".join(optlist)
+            content = call("%s %s -o - %s" % (cmd, options, source_path))
+        metadata = self._read_metadata(source_path)
+        return content, metadata
 
-        backend = self.settings.get('ASCIIDOC_BACKEND', self.default_backend)
-        ad.execute(text, content, backend=backend)
-        content = content.getvalue().decode('utf8')
+    def _get_cmd(self):
+        """Returns the AsciiDoc utility command to use for rendering or None if
+        one cannot be found."""
+        if self.settings.get('ASCIIDOC_CMD') in ALLOWED_CMDS:
+            return self.settings.get('ASCIIDOC_CMD')
+        return default()
 
+    def _read_metadata(self, source_path):
+        """Parses the AsciiDoc file at the given `source_path` and returns found
+        metadata."""
         metadata = {}
-        for name, value in ad.asciidoc.document.attributes.items():
-            if value is None:
-                continue
-            name = name.lower()
-            metadata[name] = self.process_metadata(name, six.text_type(value))
-        if 'doctitle' in metadata:
-            metadata['title'] = metadata['doctitle']
-        return content, metadata
+        with open(source_path) as fi:
+            prev = ""
+            for line in fi.readlines():
+                # Parse for doc title.
+                if 'title' not in metadata.keys():
+                    title = ""
+                    if line.startswith("= "):
+                        title = line[2:].strip()
+                    elif line.count("=") == len(prev.strip()):
+                        title = prev.strip()
+                    if title:
+                        metadata['title'] = self.process_metadata('title', title)
+
+                # Parse for other metadata.
+                regexp = re.compile(r"^:[A-z]+:\s*[A-z0-9]")
+                if regexp.search(line):
+                    toks = line.split(":", 2)
+                    key = toks[1].strip().lower()
+                    val = toks[2].strip()
+                    metadata[key] = self.process_metadata(key, val)
+                prev = line
+        return metadata
 
 def add_reader(readers):
     for ext in AsciiDocReader.file_extensions: