Add tooling for migrating to relative md links

tool/filter_update_links.py

@@ -0,0 +1,104 @@
+###############################################################################
+## Update Links: tool to help change abc.html links to relative links to the md
+## source file for the same file.
+###############################################################################
+import re
+import os.path
+
+import ruamel.yaml
+yaml = ruamel.yaml.YAML()
+with open("tool/autosubs_cheatsheet.yml") as f:
+    AUTOSUBS = yaml.load(f)
+
+LOCAL_HTML_LINK = re.compile(r"^(?P<fname>[a-z_-]+\.html)(?P<anchor>#[a-z_-]+)?$")
+UNMATCHED_REFLINK = re.compile(r"\[([^\]]+)?\]\[(\w| )*\]")
+
+REPLACE_IN_PLACE = True # False: only print changes, don't apply. True: open and rewrite files.
+
+def find_page(html, pages):
+    for page in pages:
+        if page["html"] == html:
+            return page
+    return None
+
+def filter_soup(soup, currentpage={}, config={}, pages=[], logger=None, **kwargs):
+    if not currentpage.get("md", ""):
+        logger.debug("Skipping non-md page "+currentpage["html"])
+        return
+    links = soup.find_all("a", href=True)
+    #if links:
+    #    print("Source file:", currentpage["md"])
+
+    find_replace_list = []
+    pg_srcpath = os.path.dirname(currentpage["md"])
+
+    for match in links:
+        rawlink = match["href"]
+        m = LOCAL_HTML_LINK.match(rawlink)
+        if m:
+            link = m.group("fname")
+            anchor = m.group("anchor") or ""
+            linked_page = find_page(link, pages)
+            if not linked_page:
+                logger.warning("Link to missing page "+link)
+            elif not linked_page.get("md", ""):
+                logger.info("Link to page with no md source "+link)
+            else:
+                rel_path = os.path.relpath(linked_page["md"], start=pg_srcpath)
+                #TODO: do some actual substitution here if it seems promising
+                print("Old link:", link+anchor)
+                print("New link:", rel_path+anchor)
+                # actually only replace (link) so we hopefully don't mess up e.g. frontmatter
+                link = "(%s%s)"%(link,anchor)
+                rel_path = "(%s%s)"%(rel_path, anchor)
+                find_replace_list.append( (link, rel_path) )
+
+    # reference links too
+    for s in soup.strings:
+        matches = re.finditer(UNMATCHED_REFLINK, s)
+        for m in matches:
+            if m:
+                if m.group(2):
+                    # it's a [text][ref] style link, probs
+                    old_ref = m.group(2)
+                    linktext = m.group(1)
+                else:
+                    # probs [textref][] style link
+                    old_ref = m.group(1)
+                    linktext = old_ref
+
+                link = AUTOSUBS.get(old_ref, "")
+
+                if LOCAL_HTML_LINK.match(link):
+                    linked_page = find_page(link, pages)
+                    if not linked_page:
+                        logger.warning("Autolink to missing page "+link)
+                    elif not linked_page.get("md", ""):
+                        logger.info("Autolink to page with no md source "+link)
+                    else:
+                        rel_path = os.path.relpath(linked_page["md"], start=pg_srcpath)
+                        repl_str = "[%s](%s)" % (linktext, rel_path)
+                        print("Old link:", m.group(0))
+                        print("New link:", repl_str)
+                        find_replace_list.append( (m.group(0), repl_str) )
+                elif link:
+                    # probably a link to a full URL
+                    repl_str = "[%s](%s)" % (linktext, link)
+                    print("Old link:", m.group(0))
+                    print("New link:", repl_str)
+                    find_replace_list.append( (m.group(0), repl_str) )
+
+
+
+    if REPLACE_IN_PLACE and find_replace_list:
+        print("%s: Replacing %d links"%(currentpage["md"], len(find_replace_list)))
+        fpath = os.path.join(config["content_path"], currentpage["md"])
+        with open(fpath, "r") as f:
+            contents = f.read()
+        for old, new in find_replace_list:
+            contents = contents.replace(old, new)
+        with open(fpath, "w") as f:
+            f.write(contents)
+
+
+