Re: [Tracker] REVIEW: external-crawler branch, part1

[Philip Van Hoof <philip codeminded be>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8/08/2014 10:56, Martyn Russell wrote:

Yah, just fix it anyway ;-)

Poor Aleksander. git blame is too good.


> > c2863b39b4e8f178c4d87ceee5d96ad0c3647585
> >
> > if (fs->priv->item_queue_blocker) { trace_eq ("   cancelled: item
> > queue blocked waiting for file '%s'", - -
> > g_file_get_path (fs->priv->item_queue_blocker)); +
> > g_file_get_uri (fs->priv->item_queue_blocker)); return;
> >
> > "  a string containing the GFile's URI. The returned string
> > should be freed with g_free() when no longer needed."
> >
> > Afaik this was and is in both cases a memory leak. You need to 
> > g_free() the result of g_file_get_uri. You probably just search
> > and replaced here, but it found a leak ;-)
>
> You're right, but so does _get_path() both leak. The trace_*()
> functions are usually disabled, making this a moot point, it
> depends on how particular you want to be. This existed before I got
> here, I think we can blame Aleksander :)


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.20 (MingW32)

iQEcBAEBAgAGBQJT5JOoAAoJEEP2NSGEz4aDQBoIAKli677exUp9jQ+xfD7uLyDB
Zp2kAtuwrA+8xpedOuUugbzT7cGNrvbpVLasdRwE5DWx1WfNM7rNu02cWXvMSd7v
XScp5L3DAIWpCMkBOCLkJvhlEMwfcWzCuCPT3sakT/k6Y4RJoSdxG/tceC6AoZlV
VY/tLL/kNZSjsSbIRhwXWz5Oiohud0rslc8F9Q3g/N1m9Lpq0rt2XIVe7OBgt1Ot
X5zj27q4JL9eQ2uRdtZ2Di/wd6jVgbE2jFWkYfyja/pYjsHqqFMHDpJNk/qa4XNy
VASXgcI5necTWVOk7RyjCq80h1EDGnbRBLeV6OCjgHjrA0WnIcDOEA3/OleXaOI=
=pAwe
-----END PGP SIGNATURE-----