Add test for page list generation and fix backend kwarg

camelot/handlers.py

@@ -47,9 +47,9 @@ class PDFHandler(object):
             self.password = password
             if sys.version_info[0] < 3:
                 self.password = self.password.encode("ascii")
-        self.pages = self._get_pages(self.filepath, pages)
+        self.pages = self._get_pages(pages)
 
-    def _get_pages(self, filepath, pages):
+    def _get_pages(self, pages):
         """Converts pages string to list of ints.
 
         Parameters
@@ -67,25 +67,28 @@ class PDFHandler(object):
 
         """
         page_numbers = []
+
         if pages == "1":
             page_numbers.append({"start": 1, "end": 1})
         else:
-            instream = open(filepath, "rb")
+            with open(self.filepath, "rb") as f:
-            infile = PdfFileReader(instream, strict=False)
+                infile = PdfFileReader(f, strict=False)
-            if infile.isEncrypted:
+
-                infile.decrypt(self.password)
+                if infile.isEncrypted:
-            if pages == "all":
+                    infile.decrypt(self.password)
-                page_numbers.append({"start": 1, "end": infile.getNumPages()})
+
-            else:
+                if pages == "all":
-                for r in pages.split(","):
+                    page_numbers.append({"start": 1, "end": infile.getNumPages()})
-                    if "-" in r:
+                else:
-                        a, b = r.split("-")
+                    for r in pages.split(","):
-                        if b == "end":
+                        if "-" in r:
-                            b = infile.getNumPages()
+                            a, b = r.split("-")
-                        page_numbers.append({"start": int(a), "end": int(b)})
+                            if b == "end":
-                    else:
+                                b = infile.getNumPages()
-                        page_numbers.append({"start": int(r), "end": int(r)})
+                            page_numbers.append({"start": int(a), "end": int(b)})
-            instream.close()
+                        else:
+                            page_numbers.append({"start": int(r), "end": int(r)})
+
         P = []
         for p in page_numbers:
             P.extend(range(p["start"], p["end"] + 1))

tests/test_common.py

@@ -8,6 +8,7 @@ import pandas as pd
 from pandas.testing import assert_frame_equal
 
 import camelot
+from camelot.io import PDFHandler
 from camelot.core import Table, TableList
 from camelot.__version__ import generate_version
 from camelot.backends import ImageConversionBackend
@@ -60,9 +61,7 @@ def test_password():
 
 def test_repr_poppler():
     filename = os.path.join(testdir, "foo.pdf")
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(filename, backend="poppler")
-        filename, backend=ImageConversionBackend(backend="poppler", use_fallback=False)
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=219 x2=165 y2=234>"
@@ -71,10 +70,7 @@ def test_repr_poppler():
 @skip_on_windows
 def test_repr_ghostscript():
     filename = os.path.join(testdir, "foo.pdf")
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(filename, backend="ghostscript")
-        filename,
-        backend=ImageConversionBackend(backend="ghostscript", use_fallback=False),
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=218 x2=165 y2=234>"
@@ -82,9 +78,7 @@ def test_repr_ghostscript():
 
 def test_url_poppler():
     url = "https://camelot-py.readthedocs.io/en/master/_static/pdf/foo.pdf"
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(url, backend="poppler")
-        url, backend=ImageConversionBackend(backend="poppler", use_fallback=False)
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=219 x2=165 y2=234>"
@@ -93,9 +87,7 @@ def test_url_poppler():
 @skip_on_windows
 def test_url_ghostscript():
     url = "https://camelot-py.readthedocs.io/en/master/_static/pdf/foo.pdf"
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(url, backend="ghostscript")
-        url, backend=ImageConversionBackend(backend="ghostscript", use_fallback=False)
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=218 x2=165 y2=234>"
@@ -103,27 +95,17 @@ def test_url_ghostscript():
 
 def test_pages_poppler():
     url = "https://camelot-py.readthedocs.io/en/master/_static/pdf/foo.pdf"
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(url, backend="poppler")
-        url, backend=ImageConversionBackend(backend="poppler", use_fallback=False)
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=219 x2=165 y2=234>"
 
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(url, pages="1-end", backend="poppler")
-        url,
-        pages="1-end",
-        backend=ImageConversionBackend(backend="poppler", use_fallback=False),
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=219 x2=165 y2=234>"
 
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(url, pages="all", backend="poppler")
-        url,
-        pages="all",
-        backend=ImageConversionBackend(backend="poppler", use_fallback=False),
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=219 x2=165 y2=234>"
@@ -132,27 +114,17 @@ def test_pages_poppler():
 @skip_on_windows
 def test_pages_ghostscript():
     url = "https://camelot-py.readthedocs.io/en/master/_static/pdf/foo.pdf"
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(url, backend="ghostscript")
-        url, backend=ImageConversionBackend(backend="ghostscript", use_fallback=False)
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=218 x2=165 y2=234>"
 
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(url, pages="1-end", backend="ghostscript")
-        url,
-        pages="1-end",
-        backend=ImageConversionBackend(backend="ghostscript", use_fallback=False),
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=218 x2=165 y2=234>"
 
-    tables = camelot.read_pdf(
+    tables = camelot.read_pdf(url, pages="all", backend="ghostscript")
-        url,
-        pages="all",
-        backend=ImageConversionBackend(backend="ghostscript", use_fallback=False),
-    )
     assert repr(tables) == "<TableList n=1>"
     assert repr(tables[0]) == "<Table shape=(7, 7)>"
     assert repr(tables[0].cells[0][0]) == "<Cell x1=120 y1=218 x2=165 y2=234>"
@@ -181,3 +153,22 @@ def test_table_order():
         (1, 2),
         (1, 1),
     ]
+
+
+def test_handler_pages_generator():
+    filename = os.path.join(testdir, "foo.pdf")
+
+    handler = PDFHandler(filename)
+    assert handler._get_pages("1") == [1]
+
+    handler = PDFHandler(filename)
+    assert handler._get_pages("all") == [1]
+
+    handler = PDFHandler(filename)
+    assert handler._get_pages("1-end") == [1]
+
+    handler = PDFHandler(filename)
+    assert handler._get_pages("1,2,3,4") == [1, 2, 3, 4]
+
+    handler = PDFHandler(filename)
+    assert handler._get_pages("1,2,5-10") == [1, 2, 5, 6, 7, 8, 9, 10]