Fixes needed to be able to compile pdfminer.six with Cython

pdfminer/ccitt.py

@@ -15,6 +15,8 @@ import array
 
 import six  #Python 2+3 compatibility
 
+import unittest
+
 if six.PY3:
     def get_bytes(data):
         for byte in data:
@@ -598,7 +600,7 @@ def main(argv):
             pygame.image.save(self.img, 'out.bmp')
             return
     for path in argv[1:]:
-        fp = file(path, 'rb')
+        fp = open(path, 'rb')
         (_, _, k, w, h, _) = path.split('.')
         parser = Parser(int(w))
         parser.feedbytes(fp.read())

pdfminer/cmapdb.py

@@ -426,7 +426,7 @@ class CMapParser(PSStackParser):
 def main(argv):
     args = argv[1:]
     for fname in args:
-        fp = file(fname, 'rb')
+        fp = open(fname, 'rb')
         cmap = FileUnicodeMap()
         #cmap = FileCMap()
         CMapParser(cmap, fp).run()

setup.py

@@ -3,7 +3,7 @@ import sys
 
 import pdfminer as package
 
-requires = ['six', 'pycryptodome']
+requires = ['six', 'pycryptodome', 'pygame']
 if sys.version_info >= (3, 0):
     requires.append('chardet')
 

tools/pdf2html.cgi

@@ -48,7 +48,7 @@ def convert(infp, outfp, path, codec='utf-8',
             maxpages=0, maxfilesize=0, pagenos=None,
             html=True):
     # save the input file.
-    src = file(path, 'wb')
+    src = open(path, 'wb')
     nbytes = 0
     while 1:
         data = infp.read(4096)
@@ -68,7 +68,7 @@ def convert(infp, outfp, path, codec='utf-8',
                                layoutmode='exact')
     else:
         device = TextConverter(rsrcmgr, outfp, codec=codec, laparams=laparams)
-    fp = file(path, 'rb')
+    fp = open(path, 'rb')
     interpreter = PDFPageInterpreter(rsrcmgr, device)
     for page in PDFPage.get_pages(fp, pagenos, maxpages=maxpages):
         interpreter.process_page(page)