docs, tools: add local link checker

This commit adds tools/check_links.py script, that checks that all local
links from documentation files are valid; fixes some of the invalid links
that we had; and makes travis buildbot check them as well.

1 files changed, 66 insertions, 0 deletions

diff --git a/tools/check_links.py b/tools/check_links.py
new file mode 100755
index 000000000..14a8517c1
--- /dev/null
+++ b/tools/check_links.py
@@ -0,0 +1,66 @@
+#!/usr/bin/python
+
+import os
+import re
+import sys
+
+link_re = re.compile('\[' + '[^\[\]]+' + '\]' + '\(' + '([^\(\)]+)' + '\)')
+
+if len(sys.argv) < 3:
+	print 'Usage: <root_dir> <doc_files>...'
+	sys.exit(1)
+
+root = sys.argv[1]
+docs = sys.argv[2:]
+
+links = []
+
+for doc in docs:
+	with open(doc) as f:
+		data = f.read()
+		r = link_re.findall(data)
+		for link in r:
+			links += [(doc, link)]
+
+def filter_link((doc, link)):
+	if link.startswith('http'):
+		return False
+	if link.startswith('#'):
+		return False
+	return True
+
+links = filter(filter_link, links)
+
+def fix_link((doc, link)):
+	link = link.split('#')[0]
+	link = link.split('?')[0]
+	return (doc, link)
+
+links = map(fix_link, links)
+
+errors = []
+
+def check_link((doc, link)):
+	path = os.path.dirname(doc)
+	full_link = None
+	if link[0] == '/':
+		link = link[1:]
+		full_link = os.path.join(root, link)
+	else:
+		full_link = os.path.join(root, path, link)
+	if not os.path.exists(full_link):
+		return False
+	return True
+
+for link in links:
+	if not check_link(link):
+		errors += [link]
+
+if len(errors) == 0:
+	print '%d links checked: OK' % (len(links),)
+	sys.exit(0)
+
+for (doc, link) in errors:
+	print 'File %s linked from %s not found' % (link, doc)
+
+sys.exit(2)