git@vger.kernel.org list mirror (unofficial, one of many)
 help / color / mirror / code / Atom feed
From: Andrey <ahippo@yandex.ru>
To: Philip McGraw <philip.mcgraw@bentley.com>
Cc: "git@vger.kernel.org" <git@vger.kernel.org>,
	"luke@diamand.org" <luke@diamand.org>
Subject: Re: [PATCH] git-p4: close temporary file before removing
Date: Thu, 01 Aug 2019 23:50:31 -0400	[thread overview]
Message-ID: <28612161564717831@iva4-ba508a90b0c0.qloud-c.yandex.net> (raw)
In-Reply-To: <BL0PR1901MB2097141B412696422CE957F4FFDE0@BL0PR1901MB2097.namprd19.prod.outlook.com>



01.08.2019, 11:30, "Philip McGraw" <philip.mcgraw@bentley.com>:
>>  From: Andrey <ahippo@yandex.ru>
>>  Sent: Wednesday, 31 July, 2019 21:35
>>  To: Philip McGraw <Philip.McGraw@bentley.com>
>>  Cc: git@vger.kernel.org; luke@diamand.org
>>  Subject: Re: [PATCH] git-p4: close temporary file before removing
>>
>>  31.07.2019, 17:52, "Philip McGraw" <philip.mcgraw@bentley.com>:
>>  > 2019.07.31 10:09 Andrey <ahippo@yandex.ru>
>>  >> 31.07.2019, 09:53, "Philip McGraw" <philip.mcgraw@bentley.com>:
>>  >>>>   30.07.2019, 13:37, "Philip McGraw" <philip.mcgraw@bentley.com>:
>>  >>>>   > python os.remove() throws exceptions on Windows platform when
>>  attempting
>>  >>>>   > to remove file while it is still open. Need to grab filename
>>  while file open,
>>  >>>>   > close file handle, then remove by name. Apparently other
>>  platforms are more
>>  >>>>   > permissive of removing files while busy.
>>  >>>>   > reference:
>>  >>>>   > ---
>>  >>>>   >  git-p4.py | 4 +++-
>>  >>>>   >  1 file changed, 3 insertions(+), 1 deletion(-)
>>  >>>>   >
>>  >>>>   > diff --git a/git-p4.py b/git-p4.py
>>  >>>>   > index c71a6832e2..6b9d2a8317 100755
>>  >>>>   > --- a/git-p4.py
>>  >>>>   > +++ b/git-p4.py
>>  >>>>   > @@ -1161,12 +1161,14 @@ def exceedsLargeFileThreshold(self,
>>  relPath, contents):
>>  >>>>   >                  return False
>>  >>>>   >              contentTempFile = self.generateTempFile(contents)
>>  >>>>   >              compressedContentFile =
>>  tempfile.NamedTemporaryFile(prefix='git-p4-large-file', delete=False)
>>  >>>>   > + compressedContentFileName = compressedContentFile.name
>>  >>>>   >              zf = zipfile.ZipFile(compressedContentFile.name,
>>  mode='w')
>>  >>>>   >              zf.write(contentTempFile,
>>  compress_type=zipfile.ZIP_DEFLATED)
>>  >>>>   >              zf.close()
>>  >>>>   >              compressedContentsSize =
>>  zf.infolist()[0].compress_size
>>  >>>>   >              os.remove(contentTempFile)
>>  >>>>   > - os.remove(compressedContentFile.name)
>>  >>>>   > + compressedContentFile.close()
>>  >>>>   > + os.remove(compressedContentFileName)
>>  >>>>
>>  >>>>   I'm not sure why NamedTemporaryFile() is called with delete=False
>>  above,
>>  >>>>   but it appears to me that it can have delete=True instead,
>>  >>>>   so that there is no need to call os.remove() explicitly
>>  >>>>   and thus worry about remove vs close ordering at all.
>>  >>>>
>>  >>>>   >              if compressedContentsSize > gitConfigInt('git-
>>  p4.largeFileCompressedThreshold'):
>>  >>>>   >                  return True
>>  >>>>   >          return False
>>  >>>>   > --
>>  >>>>   > 2.21.0.windows.1
>>  >>>>
>>  >>>>   Thank you,
>>  >>>>   Andrey.
>>  >>>
>>  >>>  Thanks Andrey; simpler is certainly better! I will test and re-submit
>>  v2 of patch with that approach.
>>  >>
>>  >> Thank you, that would be great!
>>  >>
>>  >> --
>>  >> Andrey.
>>  >
>>  > Unfortunately it wasn't as simple it seemed: upon testing with only
>>  changing delete=True,
>>  > found that the problem was not solved. Upon further debugging,
>>  recoded/refactored slightly adding
>>  > allocateTempFileName() locally scoped function to try to clarify how the
>>  NamedTemporaryFile()
>>  > was actually being used.
>>  >
>>  > We can't depend on the delete-on-close because the NamedTemporaryFile()
>>  is merely allocating
>>  > a temporary name for real use by the zipfile open-for-write which fails
>>  (on Windows) if file
>>  > was not explicitly closed first.
>>
>>  Oh, sorry for misguiding you!
>>  I didn't think of this aspect.
>
> No worries! I probably just misunderstood the implementation of your idea.

No, you understood what I was saying correctly.
It's just that I didn't think of opening the file twice.
(or rather that it would be a problem)

>>  > Hopefully the new patch
>>  (https://urldefense.proofpoint.com/v2/url?u=https-
>>  3A__github.com_gitgitgadget_git_pull_301&d=DwIDaQ&c=hmGTLOph1qd_VnCqj81HzE
>>  WkDaxmYdIWRBdoFggzhj8&r=b0ikFMJGw7xxhF3yjexiWJpLuNxlAh1SvUDuUJ-
>>  pHmE&m=1jGOrV_I1Mg5ajkJ7yFEcNlyLnD6zYNXqXB9Z5SIPyE&s=TdT4WHyQCk5WZty_CvajH
>>  XgrZJbmIOl1gbMcngmjmAs&e= ) will make this more clear.
>>
>>  The new changeset looks good to me.
>>  (I'll post a reply in that thread too)
>>
>>  > Open to other suggestions if still not clear.
>>
>>  Just as a thought, ZipFile() can take a file-like object instead of a file
>>  name,
>>  so can be passed the NamedTemporaryFile() object directly instead of its
>>  file name.
>>  This should hopefully avoid double-open issue on Windows.
>
> Another excellent idea that minimizes changes. I am testing this approach
> now and will submit v3 of the patch soon.

Thank you for willing to try the new approach!

>>  However, I'm good with your allocateTempFileName() changeset,
>>  so it's up to you to give it a try or not.
>>
>>  > Thanks again,
>>  > Philip
>>
>>  Thank you,
>>  Andrey.
>
> Thanks,
> Philip

-- 
Andrey.


      reply	other threads:[~2019-08-02  3:50 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-07-30 17:37 Philip McGraw
2019-07-31  1:48 ` Andrey
2019-07-31 13:53   ` Philip McGraw
2019-07-31 14:09     ` Andrey
2019-07-31 21:51       ` Philip McGraw
2019-08-01  1:34         ` Andrey
2019-08-01 15:30           ` Philip McGraw
2019-08-02  3:50             ` Andrey [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=28612161564717831@iva4-ba508a90b0c0.qloud-c.yandex.net \
    --to=ahippo@yandex.ru \
    --cc=git@vger.kernel.org \
    --cc=luke@diamand.org \
    --cc=philip.mcgraw@bentley.com \
    --subject='Re: [PATCH] git-p4: close temporary file before removing' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

git@vger.kernel.org list mirror (unofficial, one of many)

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://public-inbox.org/git
	git clone --mirror http://ou63pmih66umazou.onion/git
	git clone --mirror http://czquwvybam4bgbro.onion/git
	git clone --mirror http://hjrcffqmbrq6wope.onion/git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V1 git git/ https://public-inbox.org/git \
		git@vger.kernel.org
	public-inbox-index git

Example config snippet for mirrors.
Newsgroups are available over NNTP:
	nntp://news.public-inbox.org/inbox.comp.version-control.git
	nntp://7fh6tueqddpjyxjmgtdiueylzoqt6pt7hec3pukyptlmohoowvhde4yd.onion/inbox.comp.version-control.git
	nntp://ie5yzdi7fg72h7s4sdcztq5evakq23rdt33mfyfcddc5u3ndnw24ogqd.onion/inbox.comp.version-control.git
	nntp://4uok3hntl7oi7b4uf4rtfwefqeexfzil2w6kgk2jn5z2f764irre7byd.onion/inbox.comp.version-control.git
	nntp://news.gmane.io/gmane.comp.version-control.git
 note: .onion URLs require Tor: https://www.torproject.org/

code repositories for project(s) associated with this inbox:

	https://80x24.org/mirrors/git.git

AGPL code for this site: git clone https://public-inbox.org/public-inbox.git