Re: Parsing mbox files with Windows Php [message #181714 is a reply to message #181713] |
Mon, 27 May 2013 13:25 |
Christoph Becker
Messages: 91 Registered: June 2012
Karma:
|
Member |
|
|
Jerry Stuckle wrote:
> On 5/27/2013 8:13 AM, Christoph Becker wrote:
>> Jerry Stuckle wrote:
>>> On 5/25/2013 6:39 PM, Chuck Anderson wrote:
>>>> Peter H. Coffin wrote:
>>>> > On Fri, 24 May 2013 15:57:14 -0600, Chuck Anderson wrote:
>>>> >> I have been slowly building a Php/MySQL based IMAP email client. The
>>>> >> Php IMAP functions are well documented and easy to use. I store
>>>> >> messages in a MySQL database (attachments in the file system) for
>>>> >> ease of organization, maintenance, backup and searching. I like it
>>>> >> enough that I want to add my existing archive of email in
>>>> >> Thunderbird. To do so, I need to parse mbox files and extract
>>>> >> message headers, parts, and attachments. Thankfully, the Php IMAP
>>>> >> functions can be used to open and parse an mbox file (or even a
>>>> >> single .eml file) as well as opening a stream to the server.
>>>> >>
>>>> >> Using this functionality I can write a script to run on my WAMP
>>>> >> development machine that reads the Thunderbird folder structure,
>>>> >> parses the mbox files and saves individual messages along with their
>>>> >> folder/subfolder path in the Thunderbird folder hierarchy. It should
>>>> >> be as easy as pointing it to the top of the folder structure and
>>>> >> letting it do all the work from there.
>>>> >>
>>>> >> Unfortunately, it appears that the Windows Php binary is unable to
>>>> >> connect to an mbox file, so to make my job easy, I would have to
>>>> >> upload the entire folder structure (it is about 200MBs) to my shared
>>>> >> host and process it there. I would much rather "toy around" with
>>>> >> this process on my Windows development machine and not on the shared
>>>> >> host.
>>>> >>
>>>> >> There is a Php bug filed for this, but it was determined that the
>>>> >> "underlying c-client function is unable to open a file."
>>>> >> https://bugs.php.net/bug.php?id=39880 - closed as "not a bug."
>>>> >>
>>>> >> $mbox = imap_open('pathto/mboxfile', '', '') // works on *nix, but
>>>> >> not on Windows.
>>>> >> - Notice: Unknown: Can't open mailbox mboxfiles/Inbox: no such
>>>> >> mailbox (errflg=2) in Unknown on line 0.
>>>> >> (This second error is the one coming from the underlying c-client
>>>> >> function.)
>>>> >
>>>> > Okay, step one: quit munging stuff around and give us the EXACT code,
>>>> > the EXACT contents of variables involved, and the EXACT error
>>>> > messages.
>>>> > You're not revealing national secrets by posting paths to
>>>> > filenames, and
>>>> > what the problem is may be in what you're changing to be more general.
>>>>
>>>> I used generic paths and filenames because I had tried several
>>>> variations.
>>>>
>>>> I have learned that imap_open only works on a file - on a *nix web
>>>> server - if the path to the mbox file is relative to $HOME. I have not
>>>> found any official documentation, just forum posts saying so, and
>>>> ... it
>>>> is the only way I have been able to make it work.
>>>>
>>>> So, imap_open works for me (on the remote Linux host) if, and only
>>>> if, I
>>>> use:
>>>> imap_open ('public_html/mboxfiles/Trash', '', '')
>>>> // Trash is an mbox file I uploaded directly from my Thunderbird
>>>> Profile
>>>>
>>>> On Windows
>>>> c:/localhost is the document root
>>>> c:/localhost/imap is where the Php scripts are located.
>>>> c:/localhost/imap/mboxfiles is where the mbox file "Trash" is located
>>>>
>>>> I have tried:
>>>> imap_open('localhost/imap/mboxfiles/Trash', '', ''); // equivalent of
>>>> being relative to $HOME on *nix
>>>> imap_open('/imap/mboxfiles/Trash', '', ''); // absolute path from
>>>> document root
>>>> imap_open('mboxfiles/Trash', '', ''); // relative path
>>>> imap_open('c:/localhost/imap/mboxfiles/Trash', '', ''); // real path on
>>>> disk
>>>>
>>>> On Windows I always get these two errors (the file path changes
>>>> accordingly):
>>>>
>>>> - Warning: imap_open(): Couldn't open stream
>>>> localhost/utilities/imap/eml/Trash in
>>>> localhost\utilities\imap\imap_save_mbox_file.php on line 127
>>>>
>>>> - Notice: Unknown: Can't open mailbox
>>>> localhost/utilities/imap/eml/Trash: no such mailbox (errflg=2) in
>>>> Unknown on line 0
>>>>
>>>> I have read that the second error is coming from the c-client (errflg=2
>>>> comes from there).
>>>>
>>>
>>> I don't see any of the code you listed listing the directories
>>> imap_open() is looking in.
>>
>> There seems to be a misunderstanding. Chuck is not trying to open an
>> IMAP stream to a mail *server*, but instead to an mbox *file*.
>> Apparently this works on Linux, but not on Windows. According to
>> <https://bugs.php.net/bug.php?id=39880> this seems to be a problem of
>> the c-client library on Windows.
>>
>
> No, there is no misunderstanding. But he has NOT been clear on his code
> and resulting messages.
Chuck has posted the code that fails on Windows, but works on Linux, and
posted the according error messages.
> Also, that bug is over six years old. I have no idea if the bug still
> exists or not. However, I do not consider a bug that old to still be
> valid until proven so.
The issue's state was changed to "not a bug" in 2006 already. I
consider this to be correct, but not for the reason given in the
comments. IMO the valid reason to see it as "not a bug" is, that it's
not documented that imap_open() can be used for mbox *files*.
However, the issue does still exist, even with the latest c-client
library, which is from 2007.
>>> Once again (and for the LAST TIME):
>>>
>>> What is the *EXACT CODE* you are using. And what is the *EXACT MESSAGE*
>>> you get in response?
>>>
>>> We can't help you if you give incomplete/confusing/incorrect
>>> information!
--
Christoph M. Becker
|
|
|