# HG changeset patch
# User Henry S. Thompson <ht@inf.ed.ac.uk>
# Date 1659877009 -3600
# Node ID 04464ee31d66ac616a8a1dc4b73ee6d6bfebd7b3
# Parent  a9763cd189497e3fe4243516cc52907af17f9def
toward link extractions from pdf

diff -r a9763cd18949 -r 04464ee31d66 bin/plinks.py
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/bin/plinks.py	Sun Aug 07 13:56:49 2022 +0100
@@ -0,0 +1,25 @@
+#!/usr/bin/env python3
+import sys,pdfx,traceback
+from datetime import datetime
+
+def run(file):
+  global pdf
+  pdf=pdfx.PDFx(file)
+  return pdf.get_references_as_dict()
+
+f=sys.argv[1]
+
+try:
+  links=run(f)
+  if bool(links) and (links.get('scrape',False) or
+                      links.get('annot',False)):
+    for k in links.keys():
+      for l in links[k]:
+        print("%s\t%s"%(k,l))
+except Exception as e:
+  if str(e)=='Unexpected EOF':
+    print("%s:\t%s"%(datetime.now().isoformat(),e),file=sys.stderr)
+  else:
+    print("%s: "%(datetime.now().isoformat()),end='',file=sys.stderr)
+    traceback.print_exc(file=sys.stderr)
+