Обсуждение: strangeness in pg_dump

Поиск
Список
Период
Сортировка

strangeness in pg_dump

От
Mathieu Arnold
Дата:
Hi

I have 2 servers, a linux and a freebsd, and I don't understand why when I
dump on one, the dumped strings looks like :

"foo\r\nbar"

and on the other one :

"foo^M\
bar"

does someone have an idea of what compile time option is responsible for
that ?

--
Mathieu Arnold

Re: strangeness in pg_dump

От
"Bjoern Metzdorf"
Дата:
> I have 2 servers, a linux and a freebsd, and I don't understand why when I
> dump on one, the dumped strings looks like :
>
> "foo\r\nbar"
>
> and on the other one :
>
> "foo^M\
> bar"
>
> does someone have an idea of what compile time option is responsible for
> that ?

Perhaps this is a viewer-issue? Copy the fbsd one over to linux and view it
again.

Greetings,
Bjoern



Re: strangeness in pg_dump

От
Jan Wieck
Дата:
Mathieu Arnold wrote:
> Hi
>
> I have 2 servers, a linux and a freebsd, and I don't understand why when I
> dump on one, the dumped strings looks like :
>
> "foo\r\nbar"
>
> and on the other one :
>
> "foo^M\
> bar"
>
> does someone have an idea of what compile time option is responsible for
> that ?

    Looks identical to me.

        ^M == \r == 0x0D == Carriage Return
        \ (at end of line) == \n == 0x0A == New Line

    Maybe you should use fully compatible commands to examine the
    data?


Jan

--

#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me.                                  #
#================================================== JanWieck@Yahoo.com #



Re: strangeness in pg_dump

От
Mathieu Arnold
Дата:

--On lundi 3 juin 2002 09:35 -0400 Jan Wieck <janwieck@yahoo.com> wrote:

> Mathieu Arnold wrote:
>> Hi
>>
>> I have 2 servers, a linux and a freebsd, and I don't understand why when
>> I dump on one, the dumped strings looks like :
>>
>> "foo\r\nbar"
>>
>> and on the other one :
>>
>> "foo^M\
>> bar"
>>
>> does someone have an idea of what compile time option is responsible for
>> that ?
>
>     Looks identical to me.

well, they do not I believe.

>         ^M == \r == 0x0D == Carriage Return
>         \ (at end of line) == \n == 0x0A == New Line

that is right, but, in the first case, it is not '\r' + '\n' but '\\' + 'r'
+ '\\' + '\n'

>     Maybe you should use fully compatible commands to examine the
>     data?



--On lundi 3 juin 2002 09:31 -0400 terry@greatgulfhomes.com wrote:

> Are you sure the files are different format???

well, I use :
pg_dump -c -h pgsqlhost -h file database
on the 2 servers, the database being on another server...
I don't understand.

--
Mathieu Arnold

Re: strangeness in pg_dump

От
Mathieu Arnold
Дата:

--On lundi 3 juin 2002 15:35 +0200 Bjoern Metzdorf
<bm@turtle-entertainment.de> wrote:

>> I have 2 servers, a linux and a freebsd, and I don't understand why when
>> I dump on one, the dumped strings looks like :
>>
>> "foo\r\nbar"
>>
>> and on the other one :
>>
>> "foo^M\
>> bar"
>>
>> does someone have an idea of what compile time option is responsible for
>> that ?
>
> Perhaps this is a viewer-issue? Copy the fbsd one over to linux and view
> it again.

no, no, it's really a psql issue, as copying the linux one on the freebsd
box (viva binary compatibility) still does the same.

--
Mathieu Arnold

Re: strangeness in pg_dump

От
Jan Wieck
Дата:
Mathieu Arnold wrote:
>
>
> --On lundi 3 juin 2002 09:35 -0400 Jan Wieck <janwieck@yahoo.com> wrote:
>
> >         ^M == \r == 0x0D == Carriage Return
> >         \ (at end of line) == \n == 0x0A == New Line
>
> that is right, but, in the first case, it is not '\r' + '\n' but '\\' + 'r'
> + '\\' + '\n'

    You're  right.  Shouldn't do anything before beeing done with
    at least my first coffee.

    Okay, need to look at the code in pg_dump the decides  if  to
    quote or not.


Jan

--

#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me.                                  #
#================================================== JanWieck@Yahoo.com #



Re: strangeness in pg_dump

От
Mathieu Arnold
Дата:

--On lundi 3 juin 2002 10:08 -0400 Jan Wieck <janwieck@yahoo.com> wrote:

> Mathieu Arnold wrote:
>>
>>
>> --On lundi 3 juin 2002 09:35 -0400 Jan Wieck <janwieck@yahoo.com> wrote:
>>
>> >         ^M == \r == 0x0D == Carriage Return
>> >         \ (at end of line) == \n == 0x0A == New Line
>>
>> that is right, but, in the first case, it is not '\r' + '\n' but '\\' +
>> 'r' + '\\' + '\n'
>
>     You're  right.  Shouldn't do anything before beeing done with
>     at least my first coffee.
>
>     Okay, need to look at the code in pg_dump the decides  if  to
>     quote or not.

that was what I was doing, but my grep "['\"]\\\\[rn]['\"]" was not helping
me very much.


--
Mathieu Arnold

Re: strangeness in pg_dump

От
Jan Wieck
Дата:
Mathieu Arnold wrote:
>
>
> --On lundi 3 juin 2002 10:08 -0400 Jan Wieck <janwieck@yahoo.com> wrote:
>
> >     Okay, need to look at the code in pg_dump the decides  if  to
> >     quote or not.
>
> that was what I was doing, but my grep "['\"]\\\\[rn]['\"]" was not helping
> me very much.

    I  don't  anything that could cause this in v7.2.1's pg_dump.
    Are the two servers  running  identical  configurations  (not
    only PG version, but multibyte support and encoding too)?

    For that matter, what version(s) are you running?


Jan

--

#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me.                                  #
#================================================== JanWieck@Yahoo.com #



Re: strangeness in pg_dump

От
Mathieu Arnold
Дата:

--On lundi 3 juin 2002 10:34 -0400 Jan Wieck <janwieck@yahoo.com> wrote:

> Mathieu Arnold wrote:
>>
>>
>> --On lundi 3 juin 2002 10:08 -0400 Jan Wieck <janwieck@yahoo.com> wrote:
>>
>> >     Okay, need to look at the code in pg_dump the decides  if  to
>> >     quote or not.
>>
>> that was what I was doing, but my grep "['\"]\\\\[rn]['\"]" was not
>> helping me very much.
>
>     I  don't  anything that could cause this in v7.2.1's pg_dump.
>     Are the two servers  running  identical  configurations  (not
>     only PG version, but multibyte support and encoding too)?
>
>     For that matter, what version(s) are you running?

both have multibyte support, both encoding LATIN9 (france, euro)
7.2 on the linux, 7.2.1 on the freebsd

--
Mathieu Arnold

Re: strangeness in pg_dump

От
Tom Lane
Дата:
Mathieu Arnold <mat@mat.cc> writes:
>> For that matter, what version(s) are you running?

> both have multibyte support, both encoding LATIN9 (france, euro)
> 7.2 on the linux, 7.2.1 on the freebsd

In that case I think the difference is explained by this 7.2.1 patch:

2002-02-12 16:25  tgl

    * doc/src/sgml/ref/copy.sgml, src/backend/commands/copy.c: Modify
    COPY TO to emit carriage returns and newlines as backslash escapes
    (backslash-r, backslash-n) for protection against
    newline-conversion munging.  In future we will also tweak COPY
    FROM, but this part of the change should be backwards-compatible.
    Per pghackers discussion.  Also, update COPY reference page to
    describe the backslash conversions more completely and accurately.

            regards, tom lane