Merge branch 'feature/sum_apparent_size_only_for_files_and_symlinks' into 'main'

Align `G_FILE_MEASURE_APPARENT_SIZE` behaviour with `du` from GNU coreutils 9.2

Closes #2965

See merge request GNOME/glib!3358

3 files changed, 43 insertions, 82 deletions

diff --git a/gio/gioenums.h b/gio/gioenums.h
index 7fd74a43e..c820cd36d 100644
--- a/gio/gioenums.h
+++ b/gio/gioenums.h
@@ -224,6 +224,9 @@ typedef enum {
  *   sizes.  Normally, the block-size is used, if available, as this is a
  *   more accurate representation of disk space used.
  *   Compare with `du --apparent-size`.
+ *   Since GLib 2.78. and similarly to `du` since GNU Coreutils 9.2, this will
+ *   ignore the sizes of file types other than regular files and links, as the
+ *   sizes of other file types are not specified in a standard way.
  * @G_FILE_MEASURE_NO_XDEV: Do not cross mount point boundaries.
  *   Compare with `du -x`.
  *
diff --git a/gio/glocalfile.c b/gio/glocalfile.c
index 67d4b99fb..dbb56902d 100644
--- a/gio/glocalfile.c
+++ b/gio/glocalfile.c
@@ -86,6 +86,9 @@
 #define FILE_READ_ONLY_VOLUME           0x00080000
 #endif
 
+#ifndef S_ISREG
+#define S_ISREG(m) (((m) & _S_IFMT) == _S_IFREG)
+#endif
 #ifndef S_ISDIR
 #define S_ISDIR(m) (((m) & _S_IFMT) == _S_IFDIR)
 #endif
@@ -2777,6 +2780,39 @@ g_local_file_measure_size_of_contents (gint           fd,
                                        MeasureState  *state,
                                        GError       **error);
 
+/*
+ * _g_stat_is_size_usable:
+ * @buf: a #GLocalFileStat.
+ *
+ * Checks if the file type is such that the `st_size` field of `struct stat` is
+ * well-defined by POSIX.
+ * (see https://pubs.opengroup.org/onlinepubs/009696799/basedefs/sys/stat.h.html)
+ *
+ * This behaviour is aligned with `du` from GNU Coreutils 9.2+
+ * (see https://lists.gnu.org/archive/html/bug-coreutils/2023-03/msg00007.html)
+ * and makes apparent size sums well-defined; formerly, they depended on the
+ * implementation, and could differ across filesystems.
+ *
+ * Returns: %TRUE if the size field is well-defined, %FALSE otherwise.
+ **/
+inline static gboolean
+_g_stat_is_size_usable (const GLocalFileStat *buf)
+{
+#ifndef HAVE_STATX
+  /* Memory objects are defined by POSIX, but are not supported by statx nor Windows */
+#ifdef S_TYPEISSHM
+  if (S_TYPEISSHM (buf))
+    return TRUE;
+#endif
+#ifdef S_TYPEISTMO
+  if (S_TYPEISTMO (buf))
+    return TRUE;
+#endif
+#endif
+
+  return S_ISREG (_g_stat_mode (buf)) || S_ISLNK (_g_stat_mode (buf));
+}
+
 static gboolean
 g_local_file_measure_size_of_file (gint           parent_fd,
                                    GSList        *name,
@@ -2836,6 +2872,7 @@ g_local_file_measure_size_of_file (gint           parent_fd,
     state->disk_usage += _g_stat_blocks (&buf) * G_GUINT64_CONSTANT (512);
   else
 #endif
+  if (_g_stat_is_size_usable (&buf))
     state->disk_usage += _g_stat_size (&buf);
 
   if (S_ISDIR (_g_stat_mode (&buf)))
diff --git a/gio/tests/file.c b/gio/tests/file.c
index d16eda5c0..ad2f945f9 100644
--- a/gio/tests/file.c
+++ b/gio/tests/file.c
@@ -2515,75 +2515,10 @@ test_copy_preserve_mode (void)
 #endif
 }
 
-static gchar *
-splice_to_string (GInputStream   *stream,
-                  GError        **error)
-{
-  GMemoryOutputStream *buffer = NULL;
-  char *ret = NULL;
-
-  buffer = (GMemoryOutputStream*)g_memory_output_stream_new (NULL, 0, g_realloc, g_free);
-  if (g_output_stream_splice ((GOutputStream*)buffer, stream, 0, NULL, error) < 0)
-    goto out;
-
-  if (!g_output_stream_write ((GOutputStream*)buffer, "\0", 1, NULL, error))
-    goto out;
-
-  if (!g_output_stream_close ((GOutputStream*)buffer, NULL, error))
-    goto out;
-
-  ret = g_memory_output_stream_steal_data (buffer);
- out:
-  g_clear_object (&buffer);
-  return ret;
-}
-
-static gboolean
-get_size_from_du (const gchar *path, guint64 *size)
-{
-  GSubprocess *du;
-  gboolean ok;
-  gchar *result;
-  gchar *endptr;
-  GError *error = NULL;
-  gchar *du_path = NULL;
-
-#ifndef __APPLE__
-  du_path = g_find_program_in_path ("du");
-#endif
-
-  /* If we can’t find du, don’t try and run the test. */
-  if (du_path == NULL)
-    return FALSE;
-
-  g_free (du_path);
-
-  du = g_subprocess_new (G_SUBPROCESS_FLAGS_STDOUT_PIPE,
-                         &error,
-                         "du", "--bytes", "-s", path, NULL);
-  g_assert_no_error (error);
-
-  result = splice_to_string (g_subprocess_get_stdout_pipe (du), &error);
-  g_assert_no_error (error);
-
-  *size = g_ascii_strtoll (result, &endptr, 10);
-
-  g_subprocess_wait (du, NULL, &error);
-  g_assert_no_error (error);
-
-  ok = g_subprocess_get_successful (du);
-
-  g_object_unref (du);
-  g_free (result);
-
-  return ok;
-}
-
 static void
 test_measure (void)
 {
   GFile *file;
-  guint64 size;
   guint64 num_bytes;
   guint64 num_dirs;
   guint64 num_files;
@@ -2594,12 +2529,6 @@ test_measure (void)
   path = g_test_build_filename (G_TEST_DIST, "desktop-files", NULL);
   file = g_file_new_for_path (path);
 
-  if (!get_size_from_du (path, &size))
-    {
-      g_test_message ("du not found or fail to run, skipping byte measurement");
-      size = 0;
-    }
-
   ok = g_file_measure_disk_usage (file,
                                   G_FILE_MEASURE_APPARENT_SIZE,
                                   NULL,
@@ -2612,8 +2541,7 @@ test_measure (void)
   g_assert_true (ok);
   g_assert_no_error (error);
 
-  if (size > 0)
-    g_assert_cmpuint (num_bytes, ==, size);
+  g_assert_cmpuint (num_bytes, ==, 74478);
   g_assert_cmpuint (num_dirs, ==, 6);
   g_assert_cmpuint (num_files, ==, 32);
 
@@ -2665,8 +2593,7 @@ measure_done (GObject      *source,
   g_assert_true (ok);
   g_assert_no_error (error);
 
-  if (data->expected_bytes > 0)
-    g_assert_cmpuint (data->expected_bytes, ==, num_bytes);
+  g_assert_cmpuint (data->expected_bytes, ==, num_bytes);
   g_assert_cmpuint (data->expected_dirs, ==, num_dirs);
   g_assert_cmpuint (data->expected_files, ==, num_files);
 
@@ -2695,15 +2622,9 @@ test_measure_async (void)
 
   path = g_test_build_filename (G_TEST_DIST, "desktop-files", NULL);
   file = g_file_new_for_path (path);
-
-  if (!get_size_from_du (path, &data->expected_bytes))
-    {
-      g_test_message ("du not found or fail to run, skipping byte measurement");
-      data->expected_bytes = 0;
-    }
-
   g_free (path);
 
+  data->expected_bytes = 74478;
   data->expected_dirs = 6;
   data->expected_files = 32;