Editorial Workflows

#coding: utf-8

# Markdown auto-linker - auto-links words in a post based on what's already 
# been linked or on a common reference document.
#
# So if, like me, you keep linking to the same URLs time and again, all you
# need to do is set up a reference document with:
#
# [Editorial][edapp]
# [edapp]: http://omz-software.com/editorial/
#
# And the word "Editorial" will be replaced accordingly, and link references
# added to your new document if necessary

import workflow, console, editor, re

console.show_activity("Parsing files...")

def extract_references(buffer):
	"""Naive URL and reference link extractor"""
	buffer = unicode(buffer)
	urls = re.findall("\n\[(\w+)\]\:\s+(.+)",  buffer, re.MULTILINE | re.IGNORECASE | re.UNICODE)
	refs = re.findall("\[([\w']+)\]\[(\w+)\]", buffer, re.MULTILINE | re.IGNORECASE | re.UNICODE)
	return urls, refs

# Load common link references if available
references = workflow.get_variable('References')

# Get a sane selection range and buffer
range = editor.get_selection()

if not range:
	buffer = editor.get_text()
	range = (0,len(buffer))
else:
	buffer = editor.get_selected_text()

# Pad out the selection
selection = unicode(u"\n%s\n" % buffer)

# Extract references
urls, refs = extract_references(references)
post_urls, post_refs = extract_references(editor.get_text())

# Merge them
all_refs = dict(refs)
all_refs.update(dict(post_refs))

# Characters we'll allow around references
surround = "[\s\"\'\-\:\;\,\.\_\*]+"

# now find all words that link to those references and replace them
for word in all_refs:
	raw_ref = all_refs[word]
	ref     = re.escape(raw_ref)
	selection = re.sub(r"(%(surround)s)%(word)s(%(surround)s)" % locals(), 
	                   r"\1[%(word)s][%(ref)s]\2" % locals(), selection)

editor.replace_text(range[0],range[1],selection.strip())

# Now that we've replaced things, let's see what references we need to add
post_urls, post_refs = extract_references(editor.get_text())
post_urls = dict(post_urls)

# This reverses the tuples and sets up a nice dictionary
used_refs = dict(map(lambda x: x[::-1], post_refs))

# Build a dictionary with all the URLs we know of
all_urls = dict(urls)
all_urls.update(post_urls)

# Now add each new reference at the end of the document
output = []
for ref in used_refs:
	if ref not in post_urls:
		try:
			end = len(editor.get_text())
			url = all_urls[ref]
			editor.replace_text(end, end,"\n[%(ref)s]: %(url)s" % locals())
		except KeyError:
			word = used_refs[ref]
			output.append("[%(word)s][%(ref)s]" % locals())
			pass

workflow.set_output("\n".join(output))