Hi Masahiro, On 20 August 2014 22:02, Masahiro Yamada <yamad...@jp.panasonic.com> wrote: > Hi Simon, > > > > On Wed, 20 Aug 2014 13:13:13 -0600 > Simon Glass <s...@chromium.org> wrote: >> > +def output_is_new(): >> > + """Check if the boards.cfg file is up to date. >> > + >> > + Returns: >> > + True if the boards.cfg file exists and is newer than any of >> > + *_defconfig, MAINTAINERS and Kconfig*. False otherwise. >> > + """ >> > + try: >> > + ctime = os.path.getctime(BOARD_FILE) >> > + except OSError, exception: >> > + if exception.errno == errno.ENOENT: >> > + # return False on 'No such file or directory' error >> > + return False >> > + else: >> > + raise >> >> if not os.path.exists(BOARD_FILE) >> return False >> >> would probably be enough. >> > > > Actually my first code was as follows: > > ------------>8------------------------ > if not os.path.exists(BOARD_FILE) > return False > > ctime = os.path.getctime(BOARD_FILE) > -------------8<---------------------- > > > > But what if someone deletes BOARD_FILE > between os.path.exists(BOARD_FILE) and os.path.getctime(BOARD_FILE)? > > I know it is ridiculous to consider such a rare case. > > > But I believe it is Python style to follow > "try something and then handle an exception" thing. > > I am trying to be a bit strict to this rule > when invoking OS system calls where there is possibility of failure.
Failure in this case seems safe IMO :-) Anyway I will leave this up to you, it is not important. Regards, Simon _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
