[WIP] PDFHandler accepts file like objects

camelot/handlers.py

@@ -3,12 +3,13 @@
 import os
 import sys
 
+from builtins import str
 from PyPDF2 import PdfFileReader, PdfFileWriter
 
 from .core import TableList
 from .parsers import Stream, Lattice
 from .utils import (TemporaryDirectory, get_page_layout, get_text_objects,
-                    get_rotation)
+                    get_rotation, is_url)
 
 
 class PDFHandler(object):
@@ -27,11 +28,11 @@ class PDFHandler(object):
         Password for decryption.
 
     """
-    def __init__(self, filename, pages='1', password=None):
-        self.filename = filename
-        if not filename.lower().endswith('.pdf'):
+    def __init__(self, io, pages='1', password=None):
+        self.io = io
+        if isinstance(self.io, str) and not self.io.endswith('.pdf'):
             raise NotImplementedError("File format not supported")
-        self.pages = self._get_pages(self.filename, pages)
+        self.pages = self._get_pages(self.io, pages)
         if password is None:
             self.password = ''
         else:
@@ -149,7 +150,7 @@ def parse(self, flavor='lattice', **kwargs):
         tables = []
         with TemporaryDirectory() as tempdir:
             for p in self.pages:
-                self._save_page(self.filename, p, tempdir)
+                self._save_page(self.io, p, tempdir)
             pages = [os.path.join(tempdir, 'page-{0}.pdf'.format(p))
                      for p in self.pages]
             parser = Lattice(**kwargs) if flavor == 'lattice' else Stream(**kwargs)

camelot/utils.py

@@ -1,9 +1,16 @@
 from __future__ import division
+from future.standard_library import hooks
 import shutil
 import tempfile
 import warnings
 from itertools import groupby
 from operator import itemgetter
+from contextlib import contextmanager, closing
+
+with hooks():
+    from urllib.parse import (urlparse, uses_relative, uses_netloc, uses_params,
+                              urlencode, urljoin)
+    from urllib.request import urlopen
 
 import numpy as np
 
@@ -17,7 +24,8 @@
 from pdfminer.layout import (LAParams, LTAnno, LTChar, LTTextLineHorizontal,
                              LTTextLineVertical)
 
-
+_VALID_URLS = set(uses_relative + uses_netloc + uses_params)
+_VALID_URLS.discard('')
 stream_kwargs = [
     'columns',
     'row_close_tol',
@@ -639,3 +647,30 @@ def get_text_objects(layout, ltype="char", t=None):
     except AttributeError:
         pass
     return t
+
+
+def is_url(url):
+    """Check to see if a URL has a valid protocol.
+
+    Parameters
+    ----------
+    url : str or unicode
+
+    Returns
+    -------
+    isurl : bool
+        If `url` has a valid protocol return True otherwise False.
+    """
+    try:
+        return urlparse(url).scheme in _VALID_URLS
+    except Exception:
+        return False
+
+
+@contextmanager
+def pdf_file_reader(io):
+    if is_url(io):
+        with closing(urlopen(io)) as f:
+            yield f
+    else:
+        pass

requirements.txt

@@ -1,8 +1,9 @@
-click>=6.7
-matplotlib>=2.2.3
-numpy>=1.13.3
-opencv-python>=3.4.2.17
-openpyxl>=2.5.8
-pandas>=0.23.4
-pdfminer.six>=20170720
-PyPDF2>=1.26.0
+click>=6.7
+matplotlib>=2.2.3
+numpy>=1.13.3
+opencv-python>=3.4.2.17
+openpyxl>=2.5.8
+pandas>=0.23.4
+pdfminer.six>=20170720
+PyPDF2>=1.26.0
+future>=0.16.0