From 6bebd55022c82829c0b15d24d2ca99edd22562df Mon Sep 17 00:00:00 2001
From: Charalampos Stratakis <cstratak@redhat.com>
Date: Wed, 4 May 2016 10:39:40 +0200
Subject: [PATCH] use Py_ssize_t for file offset and length computations in
 iteration

---
 Lib/test/test_file2k.py | 16 +++++++++++++++-
 Objects/fileobject.c    | 15 +++++++--------
 2 files changed, 22 insertions(+), 9 deletions(-)

diff --git a/Lib/test/test_file2k.py b/Lib/test/test_file2k.py
index 5a3c354..8f94cee 100644
--- a/Lib/test/test_file2k.py
+++ b/Lib/test/test_file2k.py
@@ -14,7 +14,7 @@ except ImportError:
     threading = None

 from test import test_support
-from test.test_support import TESTFN, run_unittest
+from test.test_support import TESTFN, run_unittest, requires
 from UserList import UserList

 class AutoFileTests(unittest.TestCase):
@@ -416,6 +416,20 @@ class OtherFileTests(unittest.TestCase):
         finally:
             os.unlink(TESTFN)

+    @test_support.precisionbigmemtest(2**31, 2.5, dry_run=False)
+    def test_very_long_line(self, size):
+        # Issue #22526
+        requires('largefile')
+        with open(TESTFN, "wb") as fp:
+            fp.seek(size - 1)
+            fp.write("\0")
+            with open(TESTFN, "rb") as fp:
+                for l in fp:
+                    pass
+            self.assertEqual(len(l), size)
+            self.assertEqual(l.count("\0"), size)
+            l = None
+
 class FileSubclassTests(unittest.TestCase):

     def testExit(self):
diff --git a/Objects/fileobject.c b/Objects/fileobject.c
index 76cdf74..fabe207 100644
--- a/Objects/fileobject.c
+++ b/Objects/fileobject.c
@@ -2230,7 +2230,7 @@ drop_readahead(PyFileObject *f)
    (unless at EOF) and no more than bufsize.  Returns negative value on
    error, will set MemoryError if bufsize bytes cannot be allocated. */
 static int
-readahead(PyFileObject *f, int bufsize)
+readahead(PyFileObject *f, Py_ssize_t bufsize)
 {
     Py_ssize_t chunksize;

@@ -2268,7 +2268,7 @@ readahead(PyFileObject *f, int bufsize)
    logarithmic buffer growth to about 50 even when reading a 1gb line. */

 static PyStringObject *
-readahead_get_line_skip(PyFileObject *f, int skip, int bufsize)
+readahead_get_line_skip(PyFileObject *f, Py_ssize_t skip, Py_ssize_t bufsize)
 {
     PyStringObject* s;
     char *bufptr;
@@ -2288,10 +2288,10 @@ readahead_get_line_skip(PyFileObject *f, int skip, int bufsize)
         bufptr++;                               /* Count the '\n' */
         len = bufptr - f->f_bufptr;
         s = (PyStringObject *)
-            PyString_FromStringAndSize(NULL, skip+len);
+            PyString_FromStringAndSize(NULL, skip + len);
         if (s == NULL)
             return NULL;
-        memcpy(PyString_AS_STRING(s)+skip, f->f_bufptr, len);
+        memcpy(PyString_AS_STRING(s) + skip, f->f_bufptr, len);
         f->f_bufptr = bufptr;
         if (bufptr == f->f_bufend)
             drop_readahead(f);
@@ -2299,14 +2299,13 @@ readahead_get_line_skip(PyFileObject *f, int skip, int bufsize)
         bufptr = f->f_bufptr;
         buf = f->f_buf;
         f->f_buf = NULL;                /* Force new readahead buffer */
-        assert(skip+len < INT_MAX);
-        s = readahead_get_line_skip(
-            f, (int)(skip+len), bufsize + (bufsize>>2) );
+        assert(len <= PY_SSIZE_T_MAX - skip);
+        s = readahead_get_line_skip(f, skip + len, bufsize + (bufsize>>2));
         if (s == NULL) {
             PyMem_Free(buf);
             return NULL;
         }
-        memcpy(PyString_AS_STRING(s)+skip, bufptr, len);
+        memcpy(PyString_AS_STRING(s) + skip, bufptr, len);
         PyMem_Free(buf);
     }
     return s;
--
2.5.5