From 18473bd57f9255ba2cd7fe9a75881abf601df7b1 Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Thu, 14 Nov 2019 00:30:08 -0800
Subject: start of hrmars.com ingest support

---
 python/sandcrawler/html.py | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'python/sandcrawler/html.py')

diff --git a/python/sandcrawler/html.py b/python/sandcrawler/html.py
index 7e1e10d..2117eb0 100644
--- a/python/sandcrawler/html.py
+++ b/python/sandcrawler/html.py
@@ -91,4 +91,6 @@ def extract_fulltext_url(html_url, html_body):
         if iframe and '.pdf' in iframe['src']:
             return dict(pdf_url=iframe['src'])
 
+    # TODO: hrmars.com. anchor with .pdf href, and anchor text is "PDF"
+
     return dict()
-- 
cgit v1.2.3