Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat (limited to 'doc/manual/lobj.html')
-rw-r--r--doc/manual/lobj.html429
1 files changed, 429 insertions, 0 deletions
diff --git a/doc/manual/lobj.html b/doc/manual/lobj.html
new file mode 100644
index 00000000000..c8d0e518e89
--- /dev/null
+++ b/doc/manual/lobj.html
@@ -0,0 +1,429 @@
+<HTML>
+<HEAD>
+ <TITLE>The POSTGRES95 User Manual - LARGE OBJECTS</TITLE>
+</HEAD>
+
+<BODY>
+
+<font size=-1>
+<A HREF="pg95user.html">[ TOC ]</A>
+<A HREF="libpq.html">[ Previous ]</A>
+<A HREF="rules.html">[ Next ]</A>
+</font>
+<HR>
+<H1>13. LARGE OBJECTS</H1>
+<HR>
+ In POSTGRES, data values are stored in tuples and
+ individual tuples cannot span data pages. Since the size of
+ a data page is 8192 bytes, the upper limit on the size
+ of a data value is relatively low. To support the storage
+ of larger atomic values, POSTGRES provides a large
+ object interface. This interface provides file
+ oriented access to user data that has been declared to
+ be a large type.
+ This section describes the implementation and the
+ programmatic and query language interfaces to POSTGRES
+ large object data.
+
+<H2><A NAME="historical-note">13.1. Historical Note</A></H2>
+ Originally, <B>POSTGRES 4.2</B> supports three standard
+ implementations of large objects: as files external
+ to POSTGRES, as <B>UNIX</B> files managed by POSTGRES, and as data
+ stored within the POSTGRES database. It causes
+ considerable confusion among users. As a result, we only
+ support large objects as data stored within the POSTGRES
+ database in <B>POSTGRES95</B>. Even though is is slower to
+ access, it provides stricter data integrity and time
+ travel. For historical reasons, they are called
+ Inversion large objects. (We will use Inversion and large
+ objects interchangeably to mean the same thing in this
+ section.)
+
+<H2><A NAME="inversion-large-objects">13.2. Inversion Large Objects</A></H2>
+ The Inversion large object implementation breaks large
+ objects up into "chunks" and stores the chunks in
+ tuples in the database. A B-tree index guarantees fast
+ searches for the correct chunk number when doing random
+ access reads and writes.
+
+<H2><A NAME="large-object-interfaces">13.3. Large Object Interfaces</A></H2>
+ The facilities POSTGRES provides to access large
+ objects, both in the backend as part of user-defined
+ functions or the front end as part of an application
+ using the interface, are described below. (For users
+ familiar with <B>POSTGRES 4.2</B>, <B>POSTGRES95</B> has a new set of
+ functions providing a more coherent interface. The
+ interface is the same for dynamically-loaded C
+ functions as well as for .
+ The POSTGRES large object interface is modeled after
+ the <B>UNIX</B> file system interface, with analogues of
+ <B>open(2), read(2), write(2), lseek(2)</B>, etc. User
+ functions call these routines to retrieve only the data of
+ interest from a large object. For example, if a large
+ object type called mugshot existed that stored
+ photographs of faces, then a function called beard could
+ be declared on mugshot data. Beard could look at the
+ lower third of a photograph, and determine the color of
+ the beard that appeared there, if any. The entire
+ large object value need not be buffered, or even
+ examined, by the beard function.
+ Large objects may be accessed from dynamically-loaded <B>C</B>
+ functions or database client programs that link the
+ library. POSTGRES provides a set of routines that
+ support opening, reading, writing, closing, and seeking on
+ large objects.
+<p>
+<H3><A NAME="creating-large-objects">13.3.1. Creating a Large Object</A></H3>
+ The routine
+<pre> Oid lo_creat(PGconn &#42;conn, int mode)
+</pre>
+ creates a new large object. The mode is a bitmask
+ describing several different attributes of the new
+ object. The symbolic constants listed here are defined
+ in
+<pre> /usr/local/postgres95/src/backend/libpq/libpq-fs.h
+</pre>
+ The access type (read, write, or both) is controlled by
+ OR ing together the bits <B>INV_READ</B> and <B>INV_WRITE</B>. If
+ the large object should be archived -- that is, if
+ historical versions of it should be moved periodically to
+ a special archive relation -- then the <B>INV_ARCHIVE</B> bit
+ should be set. The low-order sixteen bits of mask are
+ the storage manager number on which the large object
+ should reside. For sites other than Berkeley, these
+ bits should always be zero.
+ The commands below create an (Inversion) large object:
+<pre> inv_oid = lo_creat(INV_READ|INV_WRITE|INV_ARCHIVE);
+</pre>
+
+<H3><A NAME="importing-a-large-object">13.3.2. Importing a Large Object</A></H3>
+To import a <B>UNIX</B> file as
+ a large object, call
+<pre> Oid
+ lo_import(PGconn &#42;conn, text &#42;filename)
+</pre>
+ The filename argument specifies the <B>UNIX</B> pathname of
+ the file to be imported as a large object.
+<p>
+<H3><A NAME="exporting-a-large-object">13.3.3. Exporting a Large Object</A></H3>
+To export a large object
+ into <B>UNIX</B> file, call
+<pre> int
+ lo_export(PGconn &#42;conn, Oid lobjId, text &#42;filename)
+</pre>
+ The lobjId argument specifies the Oid of the large
+ object to export and the filename argument specifies
+ the <B>UNIX</B> pathname of the file.
+<p>
+<H3><A NAME="opening-an-existing-large-object">13.3.4. Opening an Existing Large Object</A></H3>
+ To open an existing large object, call
+<pre> int
+ lo_open(PGconn &#42;conn, Oid lobjId, int mode, ...)
+</pre>
+ The lobjId argument specifies the Oid of the large
+ object to open. The mode bits control whether the
+ object is opened for reading INV_READ), writing or
+ both.
+ A large object cannot be opened before it is created.
+ lo_open returns a large object descriptor for later use
+ in lo_read, lo_write, lo_lseek, lo_tell, and lo_close.
+<p>
+<H3><A NAME="writing-data-to-a-large-object">13.3.5. Writing Data to a Large Object</A></H3>
+ The routine
+<pre> int
+ lo_write(PGconn &#42;conn, int fd, char &#42;buf, int len)
+</pre>
+ writes len bytes from buf to large object fd. The fd
+ argument must have been returned by a previous lo_open.
+ The number of bytes actually written is returned. In
+ the event of an error, the return value is negative.
+<p>
+<H3><A NAME="seeking-on-a-large-object">13.3.6. Seeking on a Large Object</A></H3>
+ To change the current read or write location on a large
+ object, call
+<pre> int
+ lo_lseek(PGconn &#42;conn, int fd, int offset, int whence)
+</pre>
+ This routine moves the current location pointer for the
+ large object described by fd to the new location specified
+ by offset. The valid values for .i whence are
+ SEEK_SET SEEK_CUR and SEEK_END.
+<p>
+<H3><A NAME="closing-a-large-object-descriptor">13.3.7. Closing a Large Object Descriptor</A></H3>
+ A large object may be closed by calling
+<pre> int
+ lo_close(PGconn &#42;conn, int fd)
+</pre>
+ where fd is a large object descriptor returned by
+ lo_open. On success, <B>lo_close</B> returns zero. On error,
+ the return value is negative.
+
+<H2><A NAME="built-in-registered-functions">13.4. Built in registered functions</A></H2>
+ There are two built-in registered functions, <B>lo_import</B>
+ and <B>lo_export</B> which are convenient for use in <B>SQL</B>
+ queries.
+ Here is an example of there use
+<pre> CREATE TABLE image (
+ name text,
+ raster oid
+ );
+
+ INSERT INTO image (name, raster)
+ VALUES ('beautiful image', lo_import('/etc/motd'));
+
+ SELECT lo_export(image.raster, "/tmp/motd") from image
+ WHERE name = 'beautiful image';
+</pre>
+<H2><A NAME="accessing-large-objects-from-libpq">13.5. Accessing Large Objects from LIBPQ</A></H2>
+ Below is a sample program which shows how the large object
+ interface
+ in LIBPQ can be used. Parts of the program are
+ commented out but are left in the source for the readers
+ benefit. This program can be found in
+<pre> ../src/test/examples
+</pre>
+ Frontend applications which use the large object interface
+ in LIBPQ should include the header file
+ libpq/libpq-fs.h and link with the libpq library.
+
+<H2><A NAME="sample-program">13.6. Sample Program</A></H2>
+<pre> /&#42;--------------------------------------------------------------
+ &#42;
+ &#42; testlo.c--
+ &#42; test using large objects with libpq
+ &#42;
+ &#42; Copyright (c) 1994, Regents of the University of California
+ &#42;
+ &#42;
+ &#42; IDENTIFICATION
+ &#42; /usr/local/devel/pglite/cvs/src/doc/manual.me,v 1.16 1995/09/01 23:55:00 jolly Exp
+ &#42;
+ &#42;--------------------------------------------------------------
+ &#42;/
+ #include &lt;stdio.h&gt;
+ #include "libpq-fe.h"
+ #include "libpq/libpq-fs.h"
+<p>
+ #define BUFSIZE 1024
+<p>
+ /&#42;
+ &#42; importFile &#42; import file "in_filename" into database as large object "lobjOid"
+ &#42;
+ &#42;/
+ Oid importFile(PGconn &#42;conn, char &#42;filename)
+ {
+ Oid lobjId;
+ int lobj_fd;
+ char buf[BUFSIZE];
+ int nbytes, tmp;
+ int fd;
+<p>
+ /&#42;
+ &#42; open the file to be read in
+ &#42;/
+ fd = open(filename, O_RDONLY, 0666);
+ if (fd &lt; 0) { /&#42; error &#42;/
+ fprintf(stderr, "can't open unix file
+ }
+<p>
+ /&#42;
+ &#42; create the large object
+ &#42;/
+ lobjId = lo_creat(conn, INV_READ|INV_WRITE);
+ if (lobjId == 0) {
+ fprintf(stderr, "can't create large object");
+ }
+<p>
+ lobj_fd = lo_open(conn, lobjId, INV_WRITE);
+ /&#42;
+ &#42; read in from the Unix file and write to the inversion file
+ &#42;/
+ while ((nbytes = read(fd, buf, BUFSIZE)) &gt; 0) {
+ tmp = lo_write(conn, lobj_fd, buf, nbytes);
+ if (tmp &lt; nbytes) {
+ fprintf(stderr, "error while reading
+ }
+ }
+<p>
+ (void) close(fd);
+ (void) lo_close(conn, lobj_fd);
+<p>
+ return lobjId;
+ }
+<p>
+ void pickout(PGconn &#42;conn, Oid lobjId, int start, int len)
+ {
+ int lobj_fd;
+ char&#42; buf;
+ int nbytes;
+ int nread;
+<p>
+ lobj_fd = lo_open(conn, lobjId, INV_READ);
+ if (lobj_fd &lt; 0) {
+ fprintf(stderr,"can't open large object &#37;d",
+ lobjId);
+ }
+<p>
+ lo_lseek(conn, lobj_fd, start, SEEK_SET);
+ buf = malloc(len+1);
+<p>
+ nread = 0;
+ while (len - nread &gt; 0) {
+ nbytes = lo_read(conn, lobj_fd, buf, len - nread);
+ buf[nbytes] = ' ';
+ fprintf(stderr,"&gt;&gt;&gt; &#37;s", buf);
+ nread += nbytes;
+ }
+ fprintf(stderr,"0);
+ lo_close(conn, lobj_fd);
+ }
+<p>
+ void overwrite(PGconn &#42;conn, Oid lobjId, int start, int len)
+ {
+ int lobj_fd;
+ char&#42; buf;
+ int nbytes;
+ int nwritten;
+ int i;
+<p>
+ lobj_fd = lo_open(conn, lobjId, INV_READ);
+ if (lobj_fd &lt; 0) {
+ fprintf(stderr,"can't open large object &#37;d",
+ lobjId);
+ }
+<p>
+ lo_lseek(conn, lobj_fd, start, SEEK_SET);
+ buf = malloc(len+1);
+<p>
+ for (i=0;i&lt;len;i++)
+ buf[i] = 'X';
+ buf[i] = ' ';
+<p>
+ nwritten = 0;
+ while (len - nwritten &gt; 0) {
+ nbytes = lo_write(conn, lobj_fd, buf + nwritten, len - nwritten);
+ nwritten += nbytes;
+ }
+ fprintf(stderr,"0);
+ lo_close(conn, lobj_fd);
+ }
+<p>
+
+ /&#42;
+ &#42; exportFile &#42; export large object "lobjOid" to file "out_filename"
+ &#42;
+ &#42;/
+ void exportFile(PGconn &#42;conn, Oid lobjId, char &#42;filename)
+ {
+ int lobj_fd;
+ char buf[BUFSIZE];
+ int nbytes, tmp;
+ int fd;
+<p>
+ /&#42;
+ &#42; create an inversion "object"
+ &#42;/
+ lobj_fd = lo_open(conn, lobjId, INV_READ);
+ if (lobj_fd &lt; 0) {
+ fprintf(stderr,"can't open large object &#37;d",
+ lobjId);
+ }
+<p>
+ /&#42;
+ &#42; open the file to be written to
+ &#42;/
+ fd = open(filename, O_CREAT|O_WRONLY, 0666);
+ if (fd &lt; 0) { /&#42; error &#42;/
+ fprintf(stderr, "can't open unix file
+ filename);
+ }
+<p>
+ /&#42;
+ &#42; read in from the Unix file and write to the inversion file
+ &#42;/
+ while ((nbytes = lo_read(conn, lobj_fd, buf, BUFSIZE)) &gt; 0) {
+ tmp = write(fd, buf, nbytes);
+ if (tmp &lt; nbytes) {
+ fprintf(stderr,"error while writing
+ filename);
+ }
+ }
+<p>
+ (void) lo_close(conn, lobj_fd);
+ (void) close(fd);
+<p>
+ return;
+ }
+<p>
+ void
+ exit_nicely(PGconn&#42; conn)
+ {
+ PQfinish(conn);
+ exit(1);
+ }
+<p>
+ int
+ main(int argc, char &#42;&#42;argv)
+ {
+ char &#42;in_filename, &#42;out_filename;
+ char &#42;database;
+ Oid lobjOid;
+ PGconn &#42;conn;
+ PGresult &#42;res;
+<p>
+ if (argc != 4) {
+ fprintf(stderr, "Usage: &#37;s database_name in_filename out_filename0,
+ argv[0]);
+ exit(1);
+ }
+<p>
+ database = argv[1];
+ in_filename = argv[2];
+ out_filename = argv[3];
+<p>
+ /&#42;
+ &#42; set up the connection
+ &#42;/
+ conn = PQsetdb(NULL, NULL, NULL, NULL, database);
+<p>
+ /&#42; check to see that the backend connection was successfully made &#42;/
+ if (PQstatus(conn) == CONNECTION_BAD) {
+ fprintf(stderr,"Connection to database '&#37;s' failed.0, database);
+ fprintf(stderr,"&#37;s",PQerrorMessage(conn));
+ exit_nicely(conn);
+ }
+<p>
+ res = PQexec(conn, "begin");
+ PQclear(res);
+
+ printf("importing file
+ /&#42; lobjOid = importFile(conn, in_filename); &#42;/
+ lobjOid = lo_import(conn, in_filename);
+ /&#42;
+ printf("as large object &#37;d.0, lobjOid);
+<p>
+ printf("picking out bytes 1000-2000 of the large object0);
+ pickout(conn, lobjOid, 1000, 1000);
+<p>
+ printf("overwriting bytes 1000-2000 of the large object with X's0);
+ overwrite(conn, lobjOid, 1000, 1000);
+ &#42;/
+<p>
+ printf("exporting large object to file
+ /&#42; exportFile(conn, lobjOid, out_filename); &#42;/
+ lo_export(conn, lobjOid,out_filename);
+<p>
+ res = PQexec(conn, "end");
+ PQclear(res);
+ PQfinish(conn);
+ exit(0);
+ }
+</pre>
+<HR>
+<font size=-1>
+<A HREF="pg95user.html">[ TOC ]</A>
+<A HREF="libpq.html">[ Previous ]</A>
+<A HREF="rules.html">[ Next ]</A>
+</font>
+</BODY>
+</HTML>