Issue #9962: GzipFile now has the peek() method.

2 files changed, 46 insertions, 1 deletions

diff --git a/Lib/gzip.py b/Lib/gzip.py
index 3edc839..58e866b 100644
--- a/Lib/gzip.py
+++ b/Lib/gzip.py
@@ -204,7 +204,10 @@ class GzipFile(io.BufferedIOBase):
         return self.name
 
     def __repr__(self):
-        s = repr(self.fileobj)
+        fileobj = self.fileobj
+        if isinstance(fileobj, _PaddedFile):
+            fileobj = fileobj.file
+        s = repr(fileobj)
         return '<gzip ' + s[1:-1] + ' ' + hex(id(self)) + '>'
 
     def _init_write(self, filename):
@@ -336,6 +339,26 @@ class GzipFile(io.BufferedIOBase):
         self.offset += size
         return chunk
 
+    def peek(self, n):
+        if self.mode != READ:
+            import errno
+            raise IOError(errno.EBADF, "read() on write-only GzipFile object")
+
+        # Do not return ridiculously small buffers
+        if n < 100:
+            n = 100
+        if self.extrasize == 0:
+            if self.fileobj is None:
+                return b''
+            try:
+                self._read(max(self.max_read_chunk, n))
+            except EOFError:
+                pass
+        offset = self.offset - self.extrastart
+        remaining = self.extrasize
+        assert remaining == len(self.extrabuf) - offset
+        return self.extrabuf[offset:offset + n]
+
     def _unread(self, buf):
         self.extrasize = len(buf) + self.extrasize
         self.offset -= len(buf)
diff --git a/Lib/test/test_gzip.py b/Lib/test/test_gzip.py
index e49fe00..8e493b5 100644
--- a/Lib/test/test_gzip.py
+++ b/Lib/test/test_gzip.py
@@ -286,6 +286,28 @@ class TestGzip(unittest.TestCase):
         with gzip.GzipFile(fileobj=buf, mode="rb") as f:
             self.assertEqual(f.read(), uncompressed)
 
+    def test_peek(self):
+        uncompressed = data1 * 200
+        with gzip.GzipFile(self.filename, "wb") as f:
+            f.write(uncompressed)
+
+        def sizes():
+            while True:
+                for n in range(5, 50, 10):
+                    yield n
+
+        with gzip.GzipFile(self.filename, "rb") as f:
+            f.max_read_chunk = 33
+            nread = 0
+            for n in sizes():
+                s = f.peek(n)
+                if s == b'':
+                    break
+                self.assertEqual(f.read(len(s)), s)
+                nread += len(s)
+            self.assertEqual(f.read(100), b'')
+            self.assertEqual(nread, len(uncompressed))
+
     # Testing compress/decompress shortcut functions
 
     def test_compress(self):