pg_backup_archiver.c

PostgreSQL7.4.6 for Linux

	/* Check if tablename only is wanted */
	if (ropt->selTypes)
	{
		if ((strcmp(te->desc, "TABLE") == 0) || (strcmp(te->desc, "TABLE DATA") == 0))
		{
			if (!ropt->selTable)
				return 0;
			if (ropt->tableNames && strcmp(ropt->tableNames, te->tag) != 0)
				return 0;
		}
		else if (strcmp(te->desc, "INDEX") == 0)
		{
			if (!ropt->selIndex)
				return 0;
			if (ropt->indexNames && strcmp(ropt->indexNames, te->tag) != 0)
				return 0;
		}
		else if (strcmp(te->desc, "FUNCTION") == 0)
		{
			if (!ropt->selFunction)
				return 0;
			if (ropt->functionNames && strcmp(ropt->functionNames, te->tag) != 0)
				return 0;
		}
		else if (strcmp(te->desc, "TRIGGER") == 0)
		{
			if (!ropt->selTrigger)
				return 0;
			if (ropt->triggerNames && strcmp(ropt->triggerNames, te->tag) != 0)
				return 0;
		}
		else
			return 0;
	}

	/*
	 * Check if we had a dataDumper. Indicates if the entry is schema or
	 * data
	 */
	if (!te->hadDumper)
	{
		/*
		 * Special Case: If 'SEQUENCE SET' then it is considered a data
		 * entry
		 */
		if (strcmp(te->desc, "SEQUENCE SET") == 0)
			res = res & REQ_DATA;
		else
			res = res & ~REQ_DATA;
	}

	/*
	 * Special case: <Init> type with <Max OID> tag; this is part of a
	 * DATA restore even though it has SQL.
	 */
	if ((strcmp(te->desc, "<Init>") == 0) && (strcmp(te->tag, "Max OID") == 0))
		res = REQ_DATA;

	/* Mask it if we only want schema */
	if (ropt->schemaOnly)
		res = res & REQ_SCHEMA;

	/* Mask it we only want data */
	if (ropt->dataOnly)
		res = res & REQ_DATA;

	/* Mask it if we don't have a schema contribution */
	if (!te->defn || strlen(te->defn) == 0)
		res = res & ~REQ_SCHEMA;

	/* Finally, if we used a list, limit based on that as well */
	if (ropt->limitToList && !ropt->idWanted[te->id - 1])
		return 0;

	return res;
}

/*
 * Issue SET commands for parameters that we want to have set the same way
 * at all times during execution of a restore script.
 */
static void
_doSetFixedOutputState(ArchiveHandle *AH)
{
	TocEntry   *te;

	/* If we have an encoding setting, emit that */
	te = AH->toc->next;
	while (te != AH->toc)
	{
		if (strcmp(te->desc, "ENCODING") == 0)
		{
			ahprintf(AH, "%s", te->defn);
			break;
		}
		te = te->next;
	}

	/* Make sure function checking is disabled */
	ahprintf(AH, "SET check_function_bodies = false;\n");

	ahprintf(AH, "\n");
}

/*
 * Issue a SET SESSION AUTHORIZATION command.  Caller is responsible
 * for updating state if appropriate.  If user is NULL or an empty string,
 * the specification DEFAULT will be used.
 */
static void
_doSetSessionAuth(ArchiveHandle *AH, const char *user)
{
	PQExpBuffer cmd = createPQExpBuffer();

	appendPQExpBuffer(cmd, "SET SESSION AUTHORIZATION ");

	/*
	 * SQL requires a string literal here.	Might as well be correct.
	 */
	if (user && *user)
		appendStringLiteral(cmd, user, false);
	else
		appendPQExpBuffer(cmd, "DEFAULT");
	appendPQExpBuffer(cmd, ";");

	if (RestoringToDB(AH))
	{
		PGresult   *res;

		res = PQexec(AH->connection, cmd->data);

		if (!res || PQresultStatus(res) != PGRES_COMMAND_OK)
			die_horribly(AH, modulename, "could not set session user to \"%s\": %s",
						 user, PQerrorMessage(AH->connection));

		PQclear(res);
	}
	else
		ahprintf(AH, "%s\n\n", cmd->data);

	destroyPQExpBuffer(cmd);
}


/*
 * Issue the commands to connect to the specified database
 * as the specified user.
 *
 * If we're currently restoring right into a database, this will
 * actually establish a connection. Otherwise it puts a \connect into
 * the script output.
 */
static void
_reconnectToDB(ArchiveHandle *AH, const char *dbname, const char *user)
{
	if (RestoringToDB(AH))
		ReconnectToServer(AH, dbname, user);
	else
	{
		PQExpBuffer qry = createPQExpBuffer();

		appendPQExpBuffer(qry, "\\connect %s",
						  dbname ? fmtId(dbname) : "-");
		appendPQExpBuffer(qry, " %s\n\n",
						  fmtId(user));

		ahprintf(AH, qry->data);

		destroyPQExpBuffer(qry);
	}

	/*
	 * NOTE: currUser keeps track of what the imaginary session user in
	 * our script is
	 */
	if (AH->currUser)
		free(AH->currUser);

	AH->currUser = strdup(user);

	/* don't assume we still know the output schema */
	if (AH->currSchema)
		free(AH->currSchema);
	AH->currSchema = strdup("");

	/* re-establish fixed state */
	_doSetFixedOutputState(AH);
}

/*
 * Become the specified user, and update state to avoid redundant commands
 *
 * NULL or empty argument is taken to mean restoring the session default
 */
static void
_becomeUser(ArchiveHandle *AH, const char *user)
{
	if (!user)
		user = "";				/* avoid null pointers */

	if (AH->currUser && strcmp(AH->currUser, user) == 0)
		return;					/* no need to do anything */

	_doSetSessionAuth(AH, user);

	/*
	 * NOTE: currUser keeps track of what the imaginary session user in
	 * our script is
	 */
	if (AH->currUser)
		free(AH->currUser);

	AH->currUser = strdup(user);
}

/*
 * Become the owner of the the given TOC entry object.  If
 * changes in ownership are not allowed, this doesn't do anything.
 */
static void
_becomeOwner(ArchiveHandle *AH, TocEntry *te)
{
	if (AH->ropt && AH->ropt->noOwner)
		return;

	_becomeUser(AH, te->owner);
}


/*
 * Issue the commands to select the specified schema as the current schema
 * in the target database.
 */
static void
_selectOutputSchema(ArchiveHandle *AH, const char *schemaName)
{
	PQExpBuffer qry;

	if (!schemaName || *schemaName == '\0' ||
		strcmp(AH->currSchema, schemaName) == 0)
		return;					/* no need to do anything */

	qry = createPQExpBuffer();

	appendPQExpBuffer(qry, "SET search_path = %s",
					  fmtId(schemaName));
	if (strcmp(schemaName, "pg_catalog") != 0)
		appendPQExpBuffer(qry, ", pg_catalog");

	if (RestoringToDB(AH))
	{
		PGresult   *res;

		res = PQexec(AH->connection, qry->data);

		if (!res || PQresultStatus(res) != PGRES_COMMAND_OK)
			die_horribly(AH, modulename, "could not set search_path to \"%s\": %s",
						 schemaName, PQerrorMessage(AH->connection));

		PQclear(res);
	}
	else
		ahprintf(AH, "%s;\n\n", qry->data);

	if (AH->currSchema)
		free(AH->currSchema);
	AH->currSchema = strdup(schemaName);

	destroyPQExpBuffer(qry);
}



static int
_printTocEntry(ArchiveHandle *AH, TocEntry *te, RestoreOptions *ropt, bool isData)
{
	char	   *pfx;

	/* Select owner and schema as necessary */
	_becomeOwner(AH, te);
	_selectOutputSchema(AH, te->namespace);

	if (isData)
		pfx = "Data for ";
	else
		pfx = "";

	ahprintf(AH, "--\n-- %sTOC entry %d (OID %s)\n-- Name: %s; Type: %s; Schema: %s; Owner: %s\n",
			 pfx, te->id, te->oid, te->tag, te->desc,
			 te->namespace ? te->namespace : "-",
			 te->owner);
	if (AH->PrintExtraTocPtr !=NULL)
		(*AH->PrintExtraTocPtr) (AH, te);
	ahprintf(AH, "--\n\n");

	/*
	 * Really crude hack for suppressing AUTHORIZATION clause of CREATE SCHEMA
	 * when --no-owner mode is selected.  This is ugly, but I see no other
	 * good way ...
	 */
	if (AH->ropt && AH->ropt->noOwner && strcmp(te->desc, "SCHEMA") == 0)
	{
		ahprintf(AH, "CREATE SCHEMA %s;\n\n\n", te->tag);
	}
	else
	{
		/* normal case */
		if (strlen(te->defn) > 0)
			ahprintf(AH, "%s\n\n", te->defn);
	}

	/*
	 * If it's an ACL entry, it might contain SET SESSION AUTHORIZATION
	 * commands, so we can no longer assume we know the current auth setting.
	 */
	if (strncmp(te->desc, "ACL", 3) == 0)
	{
		if (AH->currUser)
			free(AH->currUser);
		AH->currUser = NULL;
	}

	return 1;
}

void
WriteHead(ArchiveHandle *AH)
{
	struct tm	crtm;

	(*AH->WriteBufPtr) (AH, "PGDMP", 5);		/* Magic code */
	(*AH->WriteBytePtr) (AH, AH->vmaj);
	(*AH->WriteBytePtr) (AH, AH->vmin);
	(*AH->WriteBytePtr) (AH, AH->vrev);
	(*AH->WriteBytePtr) (AH, AH->intSize);
	(*AH->WriteBytePtr) (AH, AH->offSize);
	(*AH->WriteBytePtr) (AH, AH->format);

#ifndef HAVE_LIBZ
	if (AH->compression != 0)
		write_msg(modulename, "WARNING: requested compression not available in this "
				  "installation -- archive will be uncompressed\n");

	AH->compression = 0;
#endif

	WriteInt(AH, AH->compression);

	crtm = *localtime(&AH->createDate);
	WriteInt(AH, crtm.tm_sec);
	WriteInt(AH, crtm.tm_min);
	WriteInt(AH, crtm.tm_hour);
	WriteInt(AH, crtm.tm_mday);
	WriteInt(AH, crtm.tm_mon);
	WriteInt(AH, crtm.tm_year);
	WriteInt(AH, crtm.tm_isdst);
	WriteStr(AH, PQdb(AH->connection));
}

void
ReadHead(ArchiveHandle *AH)
{
	char		tmpMag[7];
	int			fmt;
	struct tm	crtm;

	/* If we haven't already read the header... */
	if (!AH->readHeader)
	{

		(*AH->ReadBufPtr) (AH, tmpMag, 5);

		if (strncmp(tmpMag, "PGDMP", 5) != 0)
			die_horribly(AH, modulename, "did not find magic string in file header\n");

		AH->vmaj = (*AH->ReadBytePtr) (AH);
		AH->vmin = (*AH->ReadBytePtr) (AH);

		if (AH->vmaj > 1 || ((AH->vmaj == 1) && (AH->vmin > 0)))		/* Version > 1.0 */
			AH->vrev = (*AH->ReadBytePtr) (AH);
		else
			AH->vrev = 0;

		AH->version = ((AH->vmaj * 256 + AH->vmin) * 256 + AH->vrev) * 256 + 0;


		if (AH->version < K_VERS_1_0 || AH->version > K_VERS_MAX)
			die_horribly(AH, modulename, "unsupported version (%d.%d) in file header\n",
						 AH->vmaj, AH->vmin);

		AH->intSize = (*AH->ReadBytePtr) (AH);
		if (AH->intSize > 32)
			die_horribly(AH, modulename, "sanity check on integer size (%lu) failed\n",
						 (unsigned long) AH->intSize);

		if (AH->intSize > sizeof(int))
			write_msg(modulename, "WARNING: archive was made on a machine with larger integers, some operations may fail\n");

		if (AH->version >= K_VERS_1_7)
			AH->offSize = (*AH->ReadBytePtr) (AH);
		else
			AH->offSize = AH->intSize;

		fmt = (*AH->ReadBytePtr) (AH);

		if (AH->format != fmt)
			die_horribly(AH, modulename, "expected format (%d) differs from format found in file (%d)\n",
						 AH->format, fmt);
	}

	if (AH->version >= K_VERS_1_2)
	{
		if (AH->version < K_VERS_1_4)
			AH->compression = (*AH->ReadBytePtr) (AH);
		else
			AH->compression = ReadInt(AH);
	}
	else
		AH->compression = Z_DEFAULT_COMPRESSION;

#ifndef HAVE_LIBZ
	if (AH->compression != 0)
		write_msg(modulename, "WARNING: archive is compressed, but this installation does not support compression -- no data will be available\n");
#endif

	if (AH->version >= K_VERS_1_4)
	{
		crtm.tm_sec = ReadInt(AH);
		crtm.tm_min = ReadInt(AH);
		crtm.tm_hour = ReadInt(AH);
		crtm.tm_mday = ReadInt(AH);
		crtm.tm_mon = ReadInt(AH);
		crtm.tm_year = ReadInt(AH);
		crtm.tm_isdst = ReadInt(AH);

		AH->archdbname = ReadStr(AH);

		AH->createDate = mktime(&crtm);

		if (AH->createDate == (time_t) -1)
			write_msg(modulename, "WARNING: invalid creation date in header\n");
	}

}


/*
 * checkSeek
 *	  check to see if fseek can be performed.
 */

bool
checkSeek(FILE *fp)
{

	if (fseeko(fp, 0, SEEK_CUR) != 0)
		return false;
	else if (sizeof(off_t) > sizeof(long))

		/*
		 * At this point, off_t is too large for long, so we return based
		 * on whether an off_t version of fseek is available.
		 */
#ifdef HAVE_FSEEKO
		return true;
#else
		return false;
#endif
	else
		return true;
}


static void
_SortToc(ArchiveHandle *AH, TocSortCompareFn fn)
{
	TocEntry  **tea;
	TocEntry   *te;
	int			i;

	/* Allocate an array for quicksort (TOC size + head & foot) */
	tea = (TocEntry **) malloc(sizeof(TocEntry *) * (AH->tocCount + 2));

	/* Build array of toc entries, including header at start and end */
	te = AH->toc;
	for (i = 0; i <= AH->tocCount + 1; i++)
	{
		/*
		 * printf("%d: %x (%x, %x) - %u\n", i, te, te->prev, te->next,
		 * te->oidVal);
		 */
		tea[i] = te;
		te = te->next;
	}

	/* Sort it, but ignore the header entries */
	qsort(&(tea[1]), AH->tocCount, sizeof(TocEntry *), fn);

	/* Rebuild list: this works because we have headers at each end */
	for (i = 1; i <= AH->tocCount; i++)
	{
		tea[i]->next = tea[i + 1];
		tea[i]->prev = tea[i - 1];
	}


	te = AH->toc;
	for (i = 0; i <= AH->tocCount + 1; i++)
	{
		/*
		 * printf("%d: %x (%x, %x) - %u\n", i, te, te->prev, te->next,
		 * te->oidVal);
		 */
		te = te->next;
	}


	AH->toc->next = tea[1];
	AH->toc->prev = tea[AH->tocCount];
}

static int
_tocSortCompareByOIDNum(const void *p1, const void *p2)
{
	TocEntry   *te1 = *(TocEntry **) p1;
	TocEntry   *te2 = *(TocEntry **) p2;
	Oid			id1 = te1->maxOidVal;
	Oid			id2 = te2->maxOidVal;
	int			cmpval;

	/* printf("Comparing %u to %u\n", id1, id2); */

	cmpval = oidcmp(id1, id2);

	/* If we have a deterministic answer, return it. */
	if (cmpval != 0)
		return cmpval;

	/* More comparisons required */
	if (oideq(id1, te1->maxDepOidVal))	/* maxOid1 came from deps */
	{
		if (oideq(id2, te2->maxDepOidVal))		/* maxOid2 also came from
												 * deps */
		{
			cmpval = oidcmp(te1->oidVal, te2->oidVal);	/* Just compare base
														 * OIDs */
		}
		else
/* MaxOid2 was entry OID */
		{
			return 1;			/* entry1 > entry2 */
		};
	}
	else
/* must have oideq(id1, te1->oidVal) => maxOid1 = Oid1 */
	{
		if (oideq(id2, te2->maxDepOidVal))		/* maxOid2 came from deps */
		{
			return -1;			/* entry1 < entry2 */
		}
		else
/* MaxOid2 was entry OID - deps don't matter */
		{
			cmpval = 0;
		};
	};

	/*
	 * If we get here, then we've done another comparison Once again, a 0
	 * result means we require even more
	 */
	if (cmpval != 0)
		return cmpval;

	/*
	 * Entire OID details match, so use ID number (ie. original pg_dump
	 * order)
	 */
	return _tocSortCompareByIDNum(te1, te2);
}

static int
_tocSortCompareByIDNum(const void *p1, const void *p2)
{
	TocEntry   *te1 = *(TocEntry **) p1;
	TocEntry   *te2 = *(TocEntry **) p2;
	int			id1 = te1->id;
	int			id2 = te2->id;

	/* printf("Comparing %d to %d\n", id1, id2); */

	if (id1 < id2)
		return -1;
	else if (id1 > id2)
		return 1;
	else
		return 0;
}

/*
 * Assuming Oid and depOid are set, work out the various
 * Oid values used in sorting.
 */
static void
_fixupOidInfo(TocEntry *te)
{
	te->oidVal = atooid(te->oid);
	te->maxDepOidVal = _findMaxOID(te->depOid);

	/* For the purpose of sorting, find the max OID. */
	if (oidcmp(te->oidVal, te->maxDepOidVal) >= 0)
		te->maxOidVal = te->oidVal;
	else
		te->maxOidVal = te->maxDepOidVal;
}

/*
 * Find the max OID value for a given list of string Oid values
 */
static Oid
_findMaxOID(const char *((*deps)[]))
{
	const char *dep;
	int			i;
	Oid			maxOid = (Oid) 0;
	Oid			currOid;

	if (!deps)
		return maxOid;

	i = 0;
	while ((dep = (*deps)[i++]) != NULL)
	{
		currOid = atooid(dep);
		if (oidcmp(maxOid, currOid) < 0)
			maxOid = currOid;
	}

	return maxOid;
}

/*
 * Maybe I can use this somewhere...
 *
 *create table pgdump_blob_path(p text);
 *insert into pgdump_blob_path values('/home/pjw/work/postgresql-cvs/pgsql/src/bin/pg_dump_140');
 *
 *insert into dump_blob_xref select 12345,lo_import(p || '/q.q') from pgdump_blob_path;
 */