PYTHON-1695 GridOut/GridIn more closely implement io.IOBase (#387)

Allows GridOut to be wrapped with zipfile.ZipFile from the stdlib.

Author: terencehonles

doc/contributors.rst

@@ -85,3 +85,4 @@ The following is a list of people who have contributed to
 - Jagrut Trivedi(Jagrut)
 - Shrey Batra(shreybatra)
 - Felipe Rodrigues(fbidu)
+- Terence Honles (terencehonles)

gridfs/grid_file.py

@@ -15,6 +15,7 @@
 """Tools for representing files stored in GridFS."""
 import datetime
 import hashlib
+import io
 import math
 import os
 
@@ -312,6 +313,15 @@ def close(self):
             self.__flush()
             object.__setattr__(self, "_closed", True)
 
+    def read(self, size=-1):
+        raise io.UnsupportedOperation('read')
+
+    def readable(self):
+        return False
+
+    def seekable(self):
+        return False
+
     def write(self, data):
         """Write data to the file. There is no return value.
 
@@ -379,6 +389,9 @@ def writelines(self, sequence):
         for line in sequence:
             self.write(line)
 
+    def writeable(self):
+        return True
+
     def __enter__(self):
         """Support for the context manager protocol.
         """
@@ -472,6 +485,9 @@ def __getattr__(self, name):
             return self._file[name]
         raise AttributeError("GridOut object has no attribute '%s'" % name)
 
+    def readable(self):
+        return True
+
     def readchunk(self):
         """Reads a chunk at a time. If the current position is within a
         chunk the remainder of the chunk is returned.
@@ -617,6 +633,9 @@ def seek(self, pos, whence=_SEEK_SET):
             self.__chunk_iter.close()
             self.__chunk_iter = None
 
+    def seekable(self):
+        return True
+
     def __iter__(self):
         """Return an iterator over all of this file's data.
 
@@ -625,6 +644,11 @@ def __iter__(self):
         useful when serving files using a webserver that handles
         such an iterator efficiently.
 
+        .. note::
+           This is different from :py:class:`io.IOBase` which iterates over
+           *lines* in the file. Use :meth:`GridOut.readline` to read line by
+           line instead of chunk by chunk.
+
         .. versionchanged:: 3.8
            The iterator now raises :class:`CorruptGridFile` when encountering
            any truncated, missing, or extra chunk in a file. The previous
@@ -639,6 +663,9 @@ def close(self):
             self.__chunk_iter.close()
             self.__chunk_iter = None
 
+    def write(self, value):
+        raise io.UnsupportedOperation('write')
+
     def __enter__(self):
         """Makes it possible to use :class:`GridOut` files
         with the context manager protocol.

test/test_grid_file.py

@@ -19,6 +19,7 @@
 
 import datetime
 import sys
+import zipfile
 sys.path[0:0] = [""]
 
 from bson.objectid import ObjectId
@@ -644,6 +645,24 @@ def test_survive_cursor_not_found(self):
         # Paranoid, ensure that a getMore was actually sent.
         self.assertIn("getMore", listener.started_command_names())
 
+    def test_zip(self):
+        zf = StringIO()
+        z = zipfile.ZipFile(zf, "w")
+        z.writestr("test.txt", b"hello world")
+        z.close()
+        zf.seek(0)
+
+        f = GridIn(self.db.fs, filename="test.zip")
+        f.write(zf)
+        f.close()
+        self.assertEqual(1, self.db.fs.files.count_documents({}))
+        self.assertEqual(1, self.db.fs.chunks.count_documents({}))
+
+        g = GridOut(self.db.fs, f._id)
+        z = zipfile.ZipFile(g)
+        self.assertSequenceEqual(z.namelist(), ["test.txt"])
+        self.assertEqual(z.read("test.txt"), b"hello world")
+
 
 if __name__ == "__main__":
     unittest.main()