From dd0bf0a5ef3065ff978c52df899cabefa07624cc Mon Sep 17 00:00:00 2001
From: Holger Schurig <holgerschurig@gmail.com>
Date: Thu, 1 Jul 2010 14:09:48 +0200
Subject: [PATCH] link.py: removed a TODO item: finally correcly search for
 wrong links

---
 in/plugins/read_html.md     |  2 +-
 in/plugins/read_markdown.md |  2 +-
 in/plugins/read_rst.md      |  2 +-
 plugins/link.py             | 28 +++++++++++++++++++++++-----
 4 files changed, 26 insertions(+), 8 deletions(-)

diff --git a/in/plugins/read_html.md b/in/plugins/read_html.md
index 9ae4540..d2e6345 100644
--- a/in/plugins/read_html.md
+++ b/in/plugins/read_html.md
@@ -21,4 +21,4 @@ A sample "`test.html`" document looks like this:
 	<!-- to be continued -->
 
 You'll find more about "`title:`", "`parent:`" and "`ctime:`" in the
-[[page format|pageformat.html]] description.
+[[page format|pageformat]] description.
diff --git a/in/plugins/read_markdown.md b/in/plugins/read_markdown.md
index c9682a9..cc4206b 100644
--- a/in/plugins/read_markdown.md
+++ b/in/plugins/read_markdown.md
@@ -27,7 +27,7 @@ A sample "`test.md`" document looks like this:
 	Don't send me spam, *ever*!
 
 You'll find more about "`title:`", "`parent:`" and "`ctime:`" in the
-[[page format|pageformat.html]] description.
+[[page format|pageformat]] description.
 
 = Modifications =
 
diff --git a/in/plugins/read_rst.md b/in/plugins/read_rst.md
index 45543d8..d229335 100644
--- a/in/plugins/read_rst.md
+++ b/in/plugins/read_rst.md
@@ -27,4 +27,4 @@ standard Python module "`docutils`" to convert RST into HTML. A sample
 	Don't send me spam, *ever*!
 
 You'll find more about "`title:`", "`parent:`" and "`ctime:`" in the
-[[page format|pageformat.html]] description.
\ No newline at end of file
+[[page format|pageformat]] description.
diff --git a/plugins/link.py b/plugins/link.py
index f7ecb2a..8a3404a 100644
--- a/plugins/link.py
+++ b/plugins/link.py
@@ -1,6 +1,8 @@
 # -*- coding: iso-8859-1 -*-
 from webber import *
-import os, re
+import os, re, urlparse
+
+_file_links = {}
 
 # To understand this beast, read /usr/share/doc/python2.5-doc/html/lib/module-re.html :-)
 
@@ -19,6 +21,7 @@ reLink = re.compile(r'''
 	\]\]				# end of link
 	''', re.VERBOSE)
 
+
 def do_link(m):
 	"""Used in re.sub() to substitute link with HTML"""
 	text = m.group(1) or ""
@@ -41,11 +44,13 @@ def do_link(m):
 				break
 	if not text:
 		text = link
-	# validate link
-	# TODO: validating local files still not working
-	if not link.startswith("http:") and not link.endswith(".html") and not link.endswith(".png"):
+	# validate local files
+	components = urlparse.urlparse(link)
+	if components.scheme in ("", "file"):
 		file = get_current_file()
-		warning("%s: unknown link to '%s'" % (file.rel_path, link) )
+		fname = os.path.join(file.direc, components.path)
+		fname = os.path.normpath(fname)
+		_file_links[fname] = file.rel_path
 	return '<a href="%s%s">%s</a>' % (link, anchor, text)
 
 
@@ -65,6 +70,7 @@ def test_link():
 		else:
 			print "No link:", s
 
+
 def test_sub():
 	for s in (
 			'Before [[!macro]] after',
@@ -80,3 +86,15 @@ def test_sub():
 @set_hook("linkify")
 def linkify(params):
 	params.file.contents = reLink.sub(do_link, params.file.contents)
+
+
+@set_hook("finish")
+def check_links(params):
+	"""Checks all links that are stored in _file_links to warn if the
+	file doesn't exist"""
+
+	for s in _file_links:
+		#print "check:", s, cfg.out_dir
+		out_file = os.path.join(cfg.out_dir, s)
+		if not os.path.exists(out_file):
+	    		warning("%s: invalid link to '%s'" % (_file_links[s], s))
-- 
2.39.2