Tool to add related posts

related.txt

@@ -0,0 +1,29 @@
+2008-12-14-commu
+2008-12-20-commus-new-release
+2009-07-10-commu2009071
+
+2008-12-15-netviz
+2010-04-12-social-networks-local-visualization
+
+2008-12-19-wet-folding-first-attempt
+2009-01-25-two-more-attempts
+
+2008-12-27-how-to-take-lecture-notes-with-latex
+2008-12-29-latex-class-for-lecture-notes
+2014-02-17-suboptimal-latex-1-intro
+2014-02-24-suboptimal-latex-2-spacing
+2014-03-03-suboptimal-latex-3-mathematical-environments
+2014-03-10-suboptimal-latex-4-mathematics
+2014-05-25-suboptimal-latex-5-miscellanea
+
+2012-10-01-ioi-2012-my-experience
+2012-10-09-ioi-translation-systems-a-survey-of-possible-approaches
+2012-10-15-ioi-translation-systems-an-ideal-workflow
+2012-10-24-ioi-translation-systems-an-actual-solution
+
+2009-03-22-my-two-cents-on-the-daylight-saving-time
+2014-01-06-how-much-is-time-wrong-around-the-world
+2014-06-02-emotional-time
+
+2012-02-29-pydepgraph-a-dependencies-analyzer-for-python
+2013-10-25-give-python-a-bit-of-type-safety-with-pydoc-checker

series.txt

@@ -0,0 +1,12 @@
+2012-10-09-ioi-translation-systems-a-survey-of-possible-approaches
+2012-10-15-ioi-translation-systems-an-ideal-workflow
+2012-10-24-ioi-translation-systems-an-actual-solution
+
+2014-02-17-suboptimal-latex-1-intro
+2014-02-24-suboptimal-latex-2-spacing
+2014-03-03-suboptimal-latex-3-mathematical-environments
+2014-03-10-suboptimal-latex-4-mathematics
+2014-05-25-suboptimal-latex-5-miscellanea
+
+2014-07-07-if-it-looks-like-a-words-then-it-is-a-word
+2014-08-25-if-it-looks-like-a-word-it-is-a-word-one-third-of-the-time

setup_related_posts.py

@@ -0,0 +1,118 @@
+#!/usr/bin/env python
+
+import re
+import sys
+
+from glob import glob
+
+
+SEP = "\n<!-- DO NOT EDIT BELOW THIS LINE -->\n* * *"
+
+
+def load(filename):
+   """Load a file containing groups of posts to be cross-linked.
+
+   The format is: a post per line, group separated by an empty line.
+
+   """
+   groups = open(filename).read().split("\n\n")
+   groups = [set(group.strip().split("\n")) for group in groups]
+   return groups
+
+
+def check(groups, posts):
+   """Check that each referenced post exist."""
+   for group in groups:
+      for post in group:
+         if post not in posts:
+            print >> sys.stderr, "Post %s not found" % post
+            return False
+   return True
+
+
+def find_contents_and_titles(posts):
+   """Extract the non-auto-generated content and the title of each post."""
+   contents = {}
+   titles = {}
+   for post in posts:
+      contents[post] = open("./_posts/%s.md" % post).read().split(SEP)[0]
+      title_match = re.search(r'title: (.*)', contents[post])
+      if title_match is None:
+         print >> sys.stderr, "Cannot find title for post %s" % post
+         titles[post] = ""
+      else:
+         titles[post] = title_match.group(1).strip("'")
+   return contents, titles
+
+def build_links(series, related, titles):
+   """Build and return the cross-link part of the post."""
+   if series == set() and related == set():
+      return ""
+
+   # Keep only related posts that are not part of the series.
+   related.difference_update(series)
+
+   links = [SEP]
+   if series != set():
+      links.append("\n\n### Part of this series\n\n")
+      for i, post in enumerate(sorted(series)):
+         links.append("1. [%s][%d]\n" % (titles[post], i + 1000))
+      links.append("\n")
+      for i, post in enumerate(sorted(series)):
+         links.append(" [%d]: {%% post_url %s %%}\n" % (i + 1000, post))
+
+   if related != set():
+      links.append("\n\n### See also\n\n")
+      for i, post in enumerate(sorted(related)):
+         links.append("1. [%s][%d]\n" % (titles[post], i + 2000))
+      links.append("\n")
+      for i, post in enumerate(sorted(related)):
+         links.append(" [%d]: {%% post_url %s %%}\n" % (i + 2000, post))
+
+   return "".join(links)
+
+
+def run():
+   """Main function.
+
+   Load all cross-linked content from series.txt and related.txt, and
+   rewrites all posts to update the cross-link part.
+
+   """
+   posts = [path.replace("./_posts/", "").replace(".md", "")
+            for path in glob("./_posts/*.md")]
+
+   series = load("./series.txt")
+   if not check(series, posts):
+      return 1
+
+   related = load("./related.txt")
+   if not check(related, posts):
+      return 1
+
+   contents, titles = find_contents_and_titles(posts)
+
+   for post in posts:
+      this_series = set()
+      this_related = set()
+
+      # Find posts to link
+      for group in series:
+         if post in group:
+            this_series = set(group)
+      for group in related:
+         if post in group:
+            this_related = set(group)
+
+      this_related.discard(post)
+
+      links = build_links(this_series, this_related, titles)
+
+      with open("./_posts/%s.md" % post, "w") as f:
+         f.write(contents[post] + links)
+
+   return 0
+
+
+if __name__ == "__main__":
+   sys.exit(run());