Code samples: parse READMEs better and use smarter default titles

template/page-code-samples.html.jinja

@@ -45,7 +45,7 @@
     } %}
 
     {% for card in code_samples %}
-      <a class="card" href="{{target.prefix}}{{card.href}}">
+      <a class="card" href="{{target.github_forkurl}}/tree/{{target.github_branch}}/{{card.href}}">
         <div class="card-header">
         {% for lang in card.langs %}
           <span class="circled-logo"><img src="{{lang_icons[lang]}}" /></span>

tool/filter_code_samples.py

@@ -9,6 +9,8 @@
 ################################################################################
 
 import os
+import re
+from bs4 import BeautifulSoup
 from markdown import markdown
 
 cs_dir = "content/_code-samples/" #TODO: make a configurable field
@@ -17,6 +19,18 @@ skip_dirs = [
     ".git",
     "__pycache__"
 ]
+words_to_caps = [
+    "xrp",
+]
+
+def to_title_case(s):
+    """
+    Convert a folder name string to title case, for code samples that don't have
+    a README file. The input string is expected to be in kebab-case or snake_case.
+    """
+    words = re.split(r"_|[^\w']", s)
+    words = [w.upper() if w in words_to_caps else w.title() for w in words if w] #remove empty splits
+    return " ".join(words).replace("'S", "'s").replace(" A ", " a ")
 
 def all_code_samples():
     cses = []
@@ -40,9 +54,20 @@ def all_code_samples():
         if "README.md" in filenames:
             with open(os.path.join(dirpath, "README.md"), "r") as f:
                 md = f.read()
-            cs["description"] = markdown(md)
+            md = markdown(md)
+            soup = BeautifulSoup(md, "html.parser")
+            h = soup.find(["h1", "h2", "h3"]) # get first header in the file
+            if h:
+                cs["title"] = h.get_text()
             else:
-            cs["description"] = os.path.basename(dirpath)
+                cs["title"] = to_title_case(os.path.basename(dirpath))
+            p = soup.find("p") # first paragraph defines the blurb
+            cs["description"] = p.get_text()
+
+        else:
+            cs["title"] = to_title_case(os.path.basename(dirpath))
+            cs["description"] = ""
+        cs["href"] = dirpath
         cses.append(cs)
 
     return cses