Hi, In the examples what ever they are doing is with in one session....But its not in my case. I have to open the LO in each and every function and have to write my data into it and close. I am trying to get the Length of it and setting the position to END....After first write when i check the length of it when it coems fro second write....the filedescriptor is showing zero again...Every time it is showing zero and only the last content is persisting....?How to come out from this problem? Even Global File descriptor too failed....to get this... No way can we open a LO and append the content to it....... please help me...anybody Thank you all, Regards Sandhya ----- Original Message ----- From: "Marco Bizzarri" <marco.bizzarri@xxxxxxxxx> To: "sandhya" <sandhyar@xxxxxxxxxxxxxx> Cc: "Postgres" <pgsql-admin@xxxxxxxxxxxxxx> Sent: Thursday, June 22, 2006 3:33 PM Subject: Re: [ADMIN] reg:lo_write > Did you look at this example? > > http://www.redhat.com/docs/manuals/database/RHDB-7.1.3-Manual/prog/largeobjects.html > > (I think you can find better example in the postgresql documentation, > this was the first one I found) > > On 6/22/06, Marco Bizzarri <marco.bizzarri@xxxxxxxxx> wrote: > > Hi Sandhya. > > > > On 6/22/06, sandhya <sandhyar@xxxxxxxxxxxxxx> wrote: > > > Hi Macro.. > > > If i assign my lobj_fd(say LO file descriptor) to one Global int. > > > But will my lo_write work if i pass that global int inplace of lobjfd.... > > > > Yes, I think so. > > > > > One more thing...It should be in "begin" and "end" sessions right! > > > > If you mean inside a single transaction, yes, the lobj_fd is sensible > > within one transaction. > > > > > How can i open a LO in many places and write the content into it.....Will it > > > get appended to the previous content? > > > > If you have to open and append to the previous content, you could use > > other lo_lseek, and pass it SEEK_END, so that you will position to the > > end of the LO. At that point, you could "append" to the end of the LO. > > > > > You said two different file descriptors i will be having?How to overcome > > > it?Any suggestion pls > > > > > > Anybody please give me suggestion how can i over come this problem?So that i > > > can proceed further > > > > > > > > > > > > > > > ----- Original Message ----- > > > From: "Marco Bizzarri" <marco.bizzarri@xxxxxxxxx> > > > To: "sandhya" <sandhyar@xxxxxxxxxxxxxx> > > > Cc: "Postgres" <pgsql-admin@xxxxxxxxxxxxxx> > > > Sent: Thursday, June 22, 2006 1:43 PM > > > Subject: Re: [ADMIN] reg:lo_write > > > > > > > > > > I'm not sure, but I think you will have problems. As far as I can > > > > understand, largeobject semantics is similar to files. Therefore, you > > > > open an existing LO, and you obtain a file descriptor, which will > > > > point (I think) at the start of the LO. If you write something, the > > > > file descriptor will be moved accordingly. If you open again (i.e. you > > > > issue another lo_open), you will have another file descriptor for the > > > > same LO, which will be at the start of the LO, and you will overwrite > > > > your previous LO. > > > > > > > > Please take care with these advices as I'm not a PostgreSQL expert, > > > > just reporting some experience. > > > > > > > > Regards > > > > Marco > > > > > > > > On 6/22/06, sandhya <sandhyar@xxxxxxxxxxxxxx> wrote: > > > > > > > > > > > > > > > Hi, > > > > > > > > > > If i use lo_write in different function to write data into an > > > object,Will > > > > > there be any problem? > > > > > like, > > > > > I open my object and add data into it and in another function if i again > > > > > open the same object and add data....It will append to the previous > > > content > > > > > right!Am i wrong? > > > > > > > > > > If i wants to append the content to the present object is there anything > > > i > > > > > need to do? > > > > > Please help me in doing this................ > > > > > I am unable to see the image which i am exporting... > > > > > Thank you, > > > > > Sandhya > > > > > > > > > > > > > > > > > -- > > > > Marco Bizzarri > > > > http://notenotturne.blogspot.com/ > > > > > > > > ---------------------------(end of broadcast)--------------------------- > > > > TIP 9: In versions below 8.0, the planner will ignore your desire to > > > > choose an index scan if your joining column's datatypes do not > > > > match > > > > > > > > > > > > > > > > > Regards > > Marco > > > > -- > > Marco Bizzarri > > http://notenotturne.blogspot.com/ > > > > > -- > Marco Bizzarri > http://notenotturne.blogspot.com/ >
