Fix range pagenumbers in Python 3, prevent long cache filenames

pdfquery/pdfquery.py

@@ -7,6 +7,8 @@
 import numbers
 import re
 import chardet
+import sys
+import hashlib
 try:
     from collections import OrderedDict
 except ImportError:
@@ -84,7 +86,11 @@ def _comp_bbox(el, el2):
 
 
 # assorted helpers
-def _flatten(l, ltypes=(list, tuple)):
+LTYPES = (list, tuple)
+if sys.version_info.major > 2:
+    LTYPES = (list, tuple, range)
+
+def _flatten(l, ltypes=LTYPES):
     # via http://rightfootin.blogspot.com/2006/09/more-on-python-flatten.html
     ltype = type(l)
     l = list(l)
@@ -458,7 +464,9 @@ def get_tree(self, *page_numbers):
             Return lxml.etree.ElementTree for entire document, or page numbers
             given if any.
         """
-        cache_key = "_".join(map(str, _flatten(page_numbers)))
+        hasher = hashlib.md5()
+        hasher.update(str(page_numbers).encode('UTF-8'))
+        cache_key = "_{}".format(hasher.hexdigest())
         tree = self._parse_tree_cacher.get(cache_key)
         if tree is None:
             # set up root

tests/tests.py

@@ -8,6 +8,7 @@
 import sys
 import pdfquery
 from pdfquery.cache import FileCache
+import tempfile
 
 from .utils import BaseTestCase
 
@@ -157,3 +158,34 @@ def test_annot_dereferencing(self):
         pdf.load()
         pdf = pdfquery.PDFQuery("tests/samples/bug42.pdf")
         pdf.load()
+
+class TestPageRange(BaseTestCase):
+    """
+        Test various page numbers
+    """
+
+    @classmethod
+    def setUpClass(cls):
+        cache_dir = "{}/".format(tempfile.gettempdir())
+        print(cache_dir)
+        cls.pdf = pdfquery.PDFQuery("tests/samples/bug67.pdf", parse_tree_cacher=FileCache(cache_dir))
+
+    def test_page_int(self):
+        self.pdf.load(3)
+        self.assertEqual(len(self.pdf.pq('LTPage')), 1)
+        self.pdf.load(0, 10, 25, 49)
+        self.assertEqual(len(self.pdf.pq('LTPage')), 4)
+
+    def test_page_array(self):
+        self.pdf.load([0, 7, 11])
+        self.assertEqual(len(self.pdf.pq('LTPage')), 3)
+        self.pdf.load([10], [0, 12], [30, 40])
+        self.assertEqual(len(self.pdf.pq('LTPage')), 5)
+
+    def test_page_mixed(self):
+        self.pdf.load([0, 7, 11], [0, 44], 1)
+        self.assertEqual(len(self.pdf.pq('LTPage')), 6)
+
+    def test_page_range(self):
+        self.pdf.load(range(0, 150))
+        self.assertEqual(len(self.pdf.pq('LTPage')), 150)