postgresql/src/backend/executor/functions.c

1942 lines
56 KiB
C
Raw Normal View History

/*-------------------------------------------------------------------------
*
* functions.c
* Execution of SQL-language functions
*
2017-01-03 19:48:53 +01:00
* Portions Copyright (c) 1996-2017, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
2010-09-20 22:08:53 +02:00
* src/backend/executor/functions.c
*
*-------------------------------------------------------------------------
*/
#include "postgres.h"
#include "access/htup_details.h"
#include "access/xact.h"
#include "catalog/pg_proc.h"
#include "catalog/pg_type.h"
1999-07-16 07:00:38 +02:00
#include "executor/functions.h"
#include "funcapi.h"
#include "miscadmin.h"
#include "nodes/makefuncs.h"
#include "nodes/nodeFuncs.h"
#include "parser/parse_coerce.h"
#include "parser/parse_func.h"
#include "storage/proc.h"
1996-10-26 06:15:05 +02:00
#include "tcop/utility.h"
#include "utils/builtins.h"
#include "utils/datum.h"
#include "utils/lsyscache.h"
#include "utils/memutils.h"
#include "utils/snapmgr.h"
#include "utils/syscache.h"
/*
* Specialized DestReceiver for collecting query output in a SQL function
*/
typedef struct
{
DestReceiver pub; /* publicly-known function pointers */
Tuplestorestate *tstore; /* where to put result tuples */
MemoryContext cxt; /* context containing tstore */
JunkFilter *filter; /* filter to convert tuple type */
} DR_sqlfunction;
/*
* We have an execution_state record for each query in a function. Each
* record contains a plantree for its query. If the query is currently in
* F_EXEC_RUN state then there's a QueryDesc too.
*
* The "next" fields chain together all the execution_state records generated
* from a single original parsetree. (There will only be more than one in
* case of rule expansion of the original parsetree.)
*/
typedef enum
{
F_EXEC_START, F_EXEC_RUN, F_EXEC_DONE
} ExecStatus;
typedef struct execution_state
{
struct execution_state *next;
ExecStatus status;
bool setsResult; /* true if this query produces func's result */
bool lazyEval; /* true if should fetch one row at a time */
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
PlannedStmt *stmt; /* plan for this query */
QueryDesc *qd; /* null unless status == RUN */
} execution_state;
/*
* An SQLFunctionCache record is built during the first call,
* and linked to from the fn_extra field of the FmgrInfo struct.
*
* Note that currently this has only the lifespan of the calling query.
* Someday we should rewrite this code to use plancache.c to save parse/plan
* results for longer than that.
*
* Physically, though, the data has the lifespan of the FmgrInfo that's used
* to call the function, and there are cases (particularly with indexes)
* where the FmgrInfo might survive across transactions. We cannot assume
* that the parse/plan trees are good for longer than the (sub)transaction in
* which parsing was done, so we must mark the record with the LXID/subxid of
* its creation time, and regenerate everything if that's obsolete. To avoid
* memory leakage when we do have to regenerate things, all the data is kept
* in a sub-context of the FmgrInfo's fn_mcxt.
*/
typedef struct
{
char *fname; /* function name (for error msgs) */
char *src; /* function body text (for error msgs) */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
SQLFunctionParseInfoPtr pinfo; /* data for parser callback hooks */
Oid rettype; /* actual return type */
int16 typlen; /* length of the return type */
bool typbyval; /* true if return type is pass by value */
bool returnsSet; /* true if returning multiple rows */
bool returnsTuple; /* true if returning whole tuple result */
bool shutdown_reg; /* true if registered shutdown callback */
bool readonly_func; /* true to run in "read only" mode */
bool lazyEval; /* true if using lazyEval for result query */
ParamListInfo paramLI; /* Param list representing current args */
Tuplestorestate *tstore; /* where we accumulate result tuples */
JunkFilter *junkFilter; /* will be NULL if function returns VOID */
/*
* func_state is a List of execution_state records, each of which is the
* first for its original parsetree, with any additional records chained
* to it via the "next" fields. This sublist structure is needed to keep
* track of where the original query boundaries are.
*/
List *func_state;
MemoryContext fcontext; /* memory context holding this struct and all
* subsidiary data */
LocalTransactionId lxid; /* lxid in which cache was made */
SubTransactionId subxid; /* subxid in which cache was made */
} SQLFunctionCache;
typedef SQLFunctionCache *SQLFunctionCachePtr;
/*
* Data structure needed by the parser callback hooks to resolve parameter
* references during parsing of a SQL function's body. This is separate from
* SQLFunctionCache since we sometimes do parsing separately from execution.
*/
typedef struct SQLFunctionParseInfo
{
char *fname; /* function's name */
int nargs; /* number of input arguments */
Oid *argtypes; /* resolved types of input arguments */
char **argnames; /* names of input arguments; NULL if none */
/* Note that argnames[i] can be NULL, if some args are unnamed */
Oid collation; /* function's input collation, if known */
2017-06-21 20:39:04 +02:00
} SQLFunctionParseInfo;
/* non-export function prototypes */
static Node *sql_fn_param_ref(ParseState *pstate, ParamRef *pref);
static Node *sql_fn_post_column_ref(ParseState *pstate,
ColumnRef *cref, Node *var);
static Node *sql_fn_make_param(SQLFunctionParseInfoPtr pinfo,
int paramno, int location);
static Node *sql_fn_resolve_param_name(SQLFunctionParseInfoPtr pinfo,
const char *paramname, int location);
static List *init_execution_state(List *queryTree_list,
SQLFunctionCachePtr fcache,
bool lazyEvalOK);
static void init_sql_fcache(FmgrInfo *finfo, Oid collation, bool lazyEvalOK);
static void postquel_start(execution_state *es, SQLFunctionCachePtr fcache);
static bool postquel_getnext(execution_state *es, SQLFunctionCachePtr fcache);
static void postquel_end(execution_state *es);
static void postquel_sub_params(SQLFunctionCachePtr fcache,
2003-08-04 02:43:34 +02:00
FunctionCallInfo fcinfo);
static Datum postquel_get_single_result(TupleTableSlot *slot,
FunctionCallInfo fcinfo,
SQLFunctionCachePtr fcache,
MemoryContext resultcontext);
static void sql_exec_error_callback(void *arg);
static void ShutdownSQLFunction(Datum arg);
static void sqlfunction_startup(DestReceiver *self, int operation, TupleDesc typeinfo);
static bool sqlfunction_receive(TupleTableSlot *slot, DestReceiver *self);
static void sqlfunction_shutdown(DestReceiver *self);
static void sqlfunction_destroy(DestReceiver *self);
/*
* Prepare the SQLFunctionParseInfo struct for parsing a SQL function body
*
* This includes resolving actual types of polymorphic arguments.
*
* call_expr can be passed as NULL, but then we will fail if there are any
* polymorphic arguments.
*/
SQLFunctionParseInfoPtr
prepare_sql_fn_parse_info(HeapTuple procedureTuple,
Node *call_expr,
Oid inputCollation)
{
SQLFunctionParseInfoPtr pinfo;
Form_pg_proc procedureStruct = (Form_pg_proc) GETSTRUCT(procedureTuple);
int nargs;
pinfo = (SQLFunctionParseInfoPtr) palloc0(sizeof(SQLFunctionParseInfo));
/* Function's name (only) can be used to qualify argument names */
pinfo->fname = pstrdup(NameStr(procedureStruct->proname));
/* Save the function's input collation */
pinfo->collation = inputCollation;
/*
* Copy input argument types from the pg_proc entry, then resolve any
* polymorphic types.
*/
pinfo->nargs = nargs = procedureStruct->pronargs;
if (nargs > 0)
{
Oid *argOidVect;
int argnum;
argOidVect = (Oid *) palloc(nargs * sizeof(Oid));
memcpy(argOidVect,
procedureStruct->proargtypes.values,
nargs * sizeof(Oid));
for (argnum = 0; argnum < nargs; argnum++)
{
Oid argtype = argOidVect[argnum];
if (IsPolymorphicType(argtype))
{
argtype = get_call_expr_argtype(call_expr, argnum);
if (argtype == InvalidOid)
ereport(ERROR,
(errcode(ERRCODE_DATATYPE_MISMATCH),
errmsg("could not determine actual type of argument declared %s",
format_type_be(argOidVect[argnum]))));
argOidVect[argnum] = argtype;
}
}
pinfo->argtypes = argOidVect;
}
/*
* Collect names of arguments, too, if any
*/
if (nargs > 0)
{
Datum proargnames;
Datum proargmodes;
int n_arg_names;
bool isNull;
proargnames = SysCacheGetAttr(PROCNAMEARGSNSP, procedureTuple,
Anum_pg_proc_proargnames,
&isNull);
if (isNull)
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
proargnames = PointerGetDatum(NULL); /* just to be sure */
proargmodes = SysCacheGetAttr(PROCNAMEARGSNSP, procedureTuple,
Anum_pg_proc_proargmodes,
&isNull);
if (isNull)
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
proargmodes = PointerGetDatum(NULL); /* just to be sure */
n_arg_names = get_func_input_arg_names(proargnames, proargmodes,
&pinfo->argnames);
/* Paranoia: ignore the result if too few array entries */
if (n_arg_names < nargs)
pinfo->argnames = NULL;
}
else
pinfo->argnames = NULL;
return pinfo;
}
/*
* Parser setup hook for parsing a SQL function body.
*/
void
sql_fn_parser_setup(struct ParseState *pstate, SQLFunctionParseInfoPtr pinfo)
{
pstate->p_pre_columnref_hook = NULL;
pstate->p_post_columnref_hook = sql_fn_post_column_ref;
pstate->p_paramref_hook = sql_fn_param_ref;
/* no need to use p_coerce_param_hook */
pstate->p_ref_hook_state = (void *) pinfo;
}
/*
* sql_fn_post_column_ref parser callback for ColumnRefs
*/
static Node *
sql_fn_post_column_ref(ParseState *pstate, ColumnRef *cref, Node *var)
{
SQLFunctionParseInfoPtr pinfo = (SQLFunctionParseInfoPtr) pstate->p_ref_hook_state;
int nnames;
Node *field1;
Node *subfield = NULL;
const char *name1;
const char *name2 = NULL;
Node *param;
/*
* Never override a table-column reference. This corresponds to
* considering the parameter names to appear in a scope outside the
* individual SQL commands, which is what we want.
*/
if (var != NULL)
return NULL;
/*----------
* The allowed syntaxes are:
*
* A A = parameter name
* A.B A = function name, B = parameter name
* OR: A = record-typed parameter name, B = field name
* (the first possibility takes precedence)
* A.B.C A = function name, B = record-typed parameter name,
* C = field name
* A.* Whole-row reference to composite parameter A.
* A.B.* Same, with A = function name, B = parameter name
*
* Here, it's sufficient to ignore the "*" in the last two cases --- the
* main parser will take care of expanding the whole-row reference.
*----------
*/
nnames = list_length(cref->fields);
if (nnames > 3)
return NULL;
if (IsA(llast(cref->fields), A_Star))
nnames--;
field1 = (Node *) linitial(cref->fields);
Assert(IsA(field1, String));
name1 = strVal(field1);
if (nnames > 1)
{
subfield = (Node *) lsecond(cref->fields);
Assert(IsA(subfield, String));
name2 = strVal(subfield);
}
if (nnames == 3)
{
/*
* Three-part name: if the first part doesn't match the function name,
* we can fail immediately. Otherwise, look up the second part, and
* take the third part to be a field reference.
*/
if (strcmp(name1, pinfo->fname) != 0)
return NULL;
param = sql_fn_resolve_param_name(pinfo, name2, cref->location);
subfield = (Node *) lthird(cref->fields);
Assert(IsA(subfield, String));
}
else if (nnames == 2 && strcmp(name1, pinfo->fname) == 0)
{
/*
* Two-part name with first part matching function name: first see if
* second part matches any parameter name.
*/
param = sql_fn_resolve_param_name(pinfo, name2, cref->location);
if (param)
{
/* Yes, so this is a parameter reference, no subfield */
subfield = NULL;
}
else
{
/* No, so try to match as parameter name and subfield */
param = sql_fn_resolve_param_name(pinfo, name1, cref->location);
}
}
else
{
/* Single name, or parameter name followed by subfield */
param = sql_fn_resolve_param_name(pinfo, name1, cref->location);
}
if (!param)
return NULL; /* No match */
if (subfield)
{
/*
* Must be a reference to a field of a composite parameter; otherwise
* ParseFuncOrColumn will return NULL, and we'll fail back at the
* caller.
*/
param = ParseFuncOrColumn(pstate,
list_make1(subfield),
list_make1(param),
Disallow set-returning functions inside CASE or COALESCE. When we reimplemented SRFs in commit 69f4b9c85, our initial choice was to allow the behavior to vary from historical practice in cases where a SRF call appeared within a conditional-execution construct (currently, only CASE or COALESCE). But that was controversial to begin with, and subsequent discussion has resulted in a consensus that it's better to throw an error instead of executing the query differently from before, so long as we can provide a reasonably clear error message and a way to rewrite the query. Hence, add a parser mechanism to allow detection of such cases during parse analysis. The mechanism just requires storing, in the ParseState, a pointer to the set-returning FuncExpr or OpExpr most recently emitted by parse analysis. Then the parsing functions for CASE and COALESCE can detect the presence of a SRF in their arguments by noting whether this pointer changes while analyzing their arguments. Furthermore, if it does, it provides a suitable error cursor location for the complaint. (This means that if there's more than one SRF in the arguments, the error will point at the last one to be analyzed not the first. While connoisseurs of parsing behavior might find that odd, it's unlikely the average user would ever notice.) While at it, we can also provide more specific error messages than before about some pre-existing restrictions, such as no-SRFs-within-aggregates. Also, reject at parse time cases where a NULLIF or IS DISTINCT FROM construct would need to return a set. We've never supported that, but the restriction is depended on in more subtle ways now, so it seems wise to detect it at the start. Also, provide some documentation about how to rewrite a SRF-within-CASE query using a custom wrapper SRF. It turns out that the information_schema.user_mapping_options view contained an instance of exactly the behavior we're now forbidding; but rewriting it makes it more clear and safer too. initdb forced because of user_mapping_options change. Patch by me, with error message suggestions from Alvaro Herrera and Andres Freund, pursuant to a complaint from Regina Obe. Discussion: https://postgr.es/m/000001d2d5de$d8d66170$8a832450$@pcorp.us
2017-06-14 05:46:39 +02:00
pstate->p_last_srf,
Support ordered-set (WITHIN GROUP) aggregates. This patch introduces generic support for ordered-set and hypothetical-set aggregate functions, as well as implementations of the instances defined in SQL:2008 (percentile_cont(), percentile_disc(), rank(), dense_rank(), percent_rank(), cume_dist()). We also added mode() though it is not in the spec, as well as versions of percentile_cont() and percentile_disc() that can compute multiple percentile values in one pass over the data. Unlike the original submission, this patch puts full control of the sorting process in the hands of the aggregate's support functions. To allow the support functions to find out how they're supposed to sort, a new API function AggGetAggref() is added to nodeAgg.c. This allows retrieval of the aggregate call's Aggref node, which may have other uses beyond the immediate need. There is also support for ordered-set aggregates to install cleanup callback functions, so that they can be sure that infrastructure such as tuplesort objects gets cleaned up. In passing, make some fixes in the recently-added support for variadic aggregates, and make some editorial adjustments in the recent FILTER additions for aggregates. Also, simplify use of IsBinaryCoercible() by allowing it to succeed whenever the target type is ANY or ANYELEMENT. It was inconsistent that it dealt with other polymorphic target types but not these. Atri Sharma and Andrew Gierth; reviewed by Pavel Stehule and Vik Fearing, and rather heavily editorialized upon by Tom Lane
2013-12-23 22:11:35 +01:00
NULL,
cref->location);
}
return param;
}
/*
* sql_fn_param_ref parser callback for ParamRefs ($n symbols)
*/
static Node *
sql_fn_param_ref(ParseState *pstate, ParamRef *pref)
{
SQLFunctionParseInfoPtr pinfo = (SQLFunctionParseInfoPtr) pstate->p_ref_hook_state;
int paramno = pref->number;
/* Check parameter number is valid */
if (paramno <= 0 || paramno > pinfo->nargs)
return NULL; /* unknown parameter number */
return sql_fn_make_param(pinfo, paramno, pref->location);
}
/*
* sql_fn_make_param construct a Param node for the given paramno
*/
static Node *
sql_fn_make_param(SQLFunctionParseInfoPtr pinfo,
int paramno, int location)
{
Param *param;
param = makeNode(Param);
param->paramkind = PARAM_EXTERN;
param->paramid = paramno;
param->paramtype = pinfo->argtypes[paramno - 1];
param->paramtypmod = -1;
param->paramcollid = get_typcollation(param->paramtype);
param->location = location;
/*
* If we have a function input collation, allow it to override the
* type-derived collation for parameter symbols. (XXX perhaps this should
* not happen if the type collation is not default?)
*/
if (OidIsValid(pinfo->collation) && OidIsValid(param->paramcollid))
param->paramcollid = pinfo->collation;
return (Node *) param;
}
/*
* Search for a function parameter of the given name; if there is one,
* construct and return a Param node for it. If not, return NULL.
* Helper function for sql_fn_post_column_ref.
*/
static Node *
sql_fn_resolve_param_name(SQLFunctionParseInfoPtr pinfo,
const char *paramname, int location)
{
int i;
if (pinfo->argnames == NULL)
return NULL;
for (i = 0; i < pinfo->nargs; i++)
{
if (pinfo->argnames[i] && strcmp(pinfo->argnames[i], paramname) == 0)
return sql_fn_make_param(pinfo, i + 1, location);
}
return NULL;
}
/*
* Set up the per-query execution_state records for a SQL function.
*
* The input is a List of Lists of parsed and rewritten, but not planned,
* querytrees. The sublist structure denotes the original query boundaries.
*/
static List *
init_execution_state(List *queryTree_list,
SQLFunctionCachePtr fcache,
bool lazyEvalOK)
{
List *eslist = NIL;
execution_state *lasttages = NULL;
ListCell *lc1;
foreach(lc1, queryTree_list)
{
List *qtlist = lfirst_node(List, lc1);
execution_state *firstes = NULL;
execution_state *preves = NULL;
ListCell *lc2;
foreach(lc2, qtlist)
{
Query *queryTree = lfirst_node(Query, lc2);
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
PlannedStmt *stmt;
execution_state *newes;
/* Plan the query if needed */
if (queryTree->commandType == CMD_UTILITY)
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
{
/* Utility commands require no planning. */
stmt = makeNode(PlannedStmt);
stmt->commandType = CMD_UTILITY;
stmt->canSetTag = queryTree->canSetTag;
stmt->utilityStmt = queryTree->utilityStmt;
stmt->stmt_location = queryTree->stmt_location;
stmt->stmt_len = queryTree->stmt_len;
}
else
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
stmt = pg_plan_query(queryTree,
CURSOR_OPT_PARALLEL_OK,
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
NULL);
/*
* Precheck all commands for validity in a function. This should
* generally match the restrictions spi.c applies.
*/
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
if (stmt->commandType == CMD_UTILITY)
{
if (IsA(stmt->utilityStmt, CopyStmt) &&
((CopyStmt *) stmt->utilityStmt)->filename == NULL)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("cannot COPY to/from client in a SQL function")));
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
if (IsA(stmt->utilityStmt, TransactionStmt))
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
/* translator: %s is a SQL statement name */
errmsg("%s is not allowed in a SQL function",
CreateCommandTag(stmt->utilityStmt))));
}
if (fcache->readonly_func && !CommandIsReadOnly(stmt))
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
/* translator: %s is a SQL statement name */
errmsg("%s is not allowed in a non-volatile function",
CreateCommandTag((Node *) stmt))));
if (IsInParallelMode() && !CommandIsReadOnly(stmt))
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
PreventCommandIfParallelMode(CreateCommandTag((Node *) stmt));
/* OK, build the execution_state for this query */
newes = (execution_state *) palloc(sizeof(execution_state));
if (preves)
preves->next = newes;
else
firstes = newes;
newes->next = NULL;
newes->status = F_EXEC_START;
2011-04-10 17:42:00 +02:00
newes->setsResult = false; /* might change below */
newes->lazyEval = false; /* might change below */
newes->stmt = stmt;
newes->qd = NULL;
if (queryTree->canSetTag)
lasttages = newes;
preves = newes;
}
eslist = lappend(eslist, firstes);
}
/*
* Mark the last canSetTag query as delivering the function result; then,
* if it is a plain SELECT, mark it for lazy evaluation. If it's not a
* SELECT we must always run it to completion.
*
* Note: at some point we might add additional criteria for whether to use
* lazy eval. However, we should prefer to use it whenever the function
* doesn't return set, since fetching more than one row is useless in that
* case.
*
* Note: don't set setsResult if the function returns VOID, as evidenced
* by not having made a junkfilter. This ensures we'll throw away any
* output from a utility statement that check_sql_fn_retval deemed to not
* have output.
*/
if (lasttages && fcache->junkFilter)
{
lasttages->setsResult = true;
if (lazyEvalOK &&
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
lasttages->stmt->commandType == CMD_SELECT &&
!lasttages->stmt->hasModifyingCTE)
fcache->lazyEval = lasttages->lazyEval = true;
}
return eslist;
}
/*
* Initialize the SQLFunctionCache for a SQL function
*/
static void
init_sql_fcache(FmgrInfo *finfo, Oid collation, bool lazyEvalOK)
{
Oid foid = finfo->fn_oid;
MemoryContext fcontext;
MemoryContext oldcontext;
Oid rettype;
HeapTuple procedureTuple;
Form_pg_proc procedureStruct;
SQLFunctionCachePtr fcache;
List *raw_parsetree_list;
List *queryTree_list;
List *flat_query_list;
ListCell *lc;
Datum tmp;
bool isNull;
/*
* Create memory context that holds all the SQLFunctionCache data. It
* must be a child of whatever context holds the FmgrInfo.
*/
fcontext = AllocSetContextCreate(finfo->fn_mcxt,
"SQL function data",
Add macros to make AllocSetContextCreate() calls simpler and safer. I found that half a dozen (nearly 5%) of our AllocSetContextCreate calls had typos in the context-sizing parameters. While none of these led to especially significant problems, they did create minor inefficiencies, and it's now clear that expecting people to copy-and-paste those calls accurately is not a great idea. Let's reduce the risk of future errors by introducing single macros that encapsulate the common use-cases. Three such macros are enough to cover all but two special-purpose contexts; those two calls can be left as-is, I think. While this patch doesn't in itself improve matters for third-party extensions, it doesn't break anything for them either, and they can gradually adopt the simplified notation over time. In passing, change TopMemoryContext to use the default allocation parameters. Formerly it could only be extended 8K at a time. That was probably reasonable when this code was written; but nowadays we create many more contexts than we did then, so that it's not unusual to have a couple hundred K in TopMemoryContext, even without considering various dubious code that sticks other things there. There seems no good reason not to let it use growing blocks like most other contexts. Back-patch to 9.6, mostly because that's still close enough to HEAD that it's easy to do so, and keeping the branches in sync can be expected to avoid some future back-patching pain. The bugs fixed by these changes don't seem to be significant enough to justify fixing them further back. Discussion: <21072.1472321324@sss.pgh.pa.us>
2016-08-27 23:50:38 +02:00
ALLOCSET_DEFAULT_SIZES);
oldcontext = MemoryContextSwitchTo(fcontext);
/*
* Create the struct proper, link it to fcontext and fn_extra. Once this
* is done, we'll be able to recover the memory after failure, even if the
* FmgrInfo is long-lived.
*/
fcache = (SQLFunctionCachePtr) palloc0(sizeof(SQLFunctionCache));
fcache->fcontext = fcontext;
finfo->fn_extra = (void *) fcache;
/*
* get the procedure tuple corresponding to the given function Oid
*/
procedureTuple = SearchSysCache1(PROCOID, ObjectIdGetDatum(foid));
if (!HeapTupleIsValid(procedureTuple))
elog(ERROR, "cache lookup failed for function %u", foid);
procedureStruct = (Form_pg_proc) GETSTRUCT(procedureTuple);
/*
* copy function name immediately for use by error reporting callback
*/
fcache->fname = pstrdup(NameStr(procedureStruct->proname));
/*
2005-10-15 04:49:52 +02:00
* get the result type from the procedure tuple, and check for polymorphic
* result type; if so, find out the actual result type.
*/
rettype = procedureStruct->prorettype;
if (IsPolymorphicType(rettype))
{
rettype = get_fn_expr_rettype(finfo);
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
if (rettype == InvalidOid) /* this probably should not happen */
ereport(ERROR,
(errcode(ERRCODE_DATATYPE_MISMATCH),
errmsg("could not determine actual result type for function declared to return type %s",
2005-10-15 04:49:52 +02:00
format_type_be(procedureStruct->prorettype))));
}
fcache->rettype = rettype;
/* Fetch the typlen and byval info for the result type */
get_typlenbyval(rettype, &fcache->typlen, &fcache->typbyval);
/* Remember whether we're returning setof something */
fcache->returnsSet = procedureStruct->proretset;
/* Remember if function is STABLE/IMMUTABLE */
fcache->readonly_func =
(procedureStruct->provolatile != PROVOLATILE_VOLATILE);
/*
* We need the actual argument types to pass to the parser. Also make
* sure that parameter symbols are considered to have the function's
* resolved input collation.
*/
fcache->pinfo = prepare_sql_fn_parse_info(procedureTuple,
finfo->fn_expr,
collation);
/*
* And of course we need the function body text.
*/
tmp = SysCacheGetAttr(PROCOID,
procedureTuple,
Anum_pg_proc_prosrc,
&isNull);
if (isNull)
elog(ERROR, "null prosrc for function %u", foid);
fcache->src = TextDatumGetCString(tmp);
/*
* Parse and rewrite the queries in the function text. Use sublists to
* keep track of the original query boundaries. But we also build a
* "flat" list of the rewritten queries to pass to check_sql_fn_retval.
* This is because the last canSetTag query determines the result type
* independently of query boundaries --- and it might not be in the last
* sublist, for example if the last query rewrites to DO INSTEAD NOTHING.
* (It might not be unreasonable to throw an error in such a case, but
* this is the historical behavior and it doesn't seem worth changing.)
*
* Note: since parsing and planning is done in fcontext, we will generate
* a lot of cruft that lives as long as the fcache does. This is annoying
* but we'll not worry about it until the module is rewritten to use
* plancache.c.
*/
raw_parsetree_list = pg_parse_query(fcache->src);
queryTree_list = NIL;
flat_query_list = NIL;
foreach(lc, raw_parsetree_list)
{
RawStmt *parsetree = lfirst_node(RawStmt, lc);
List *queryTree_sublist;
queryTree_sublist = pg_analyze_and_rewrite_params(parsetree,
fcache->src,
(ParserSetupHook) sql_fn_parser_setup,
fcache->pinfo,
NULL);
queryTree_list = lappend(queryTree_list, queryTree_sublist);
flat_query_list = list_concat(flat_query_list,
list_copy(queryTree_sublist));
}
/*
2007-11-15 22:14:46 +01:00
* Check that the function returns the type it claims to. Although in
* simple cases this was already done when the function was defined, we
* have to recheck because database objects used in the function's queries
* might have changed type. We'd have to do it anyway if the function had
* any polymorphic arguments.
*
2007-11-15 22:14:46 +01:00
* Note: we set fcache->returnsTuple according to whether we are returning
* the whole tuple result or just a single column. In the latter case we
2007-11-15 22:14:46 +01:00
* clear returnsTuple because we need not act different from the scalar
* result case, even if it's a rowtype column. (However, we have to force
* lazy eval mode in that case; otherwise we'd need extra code to expand
* the rowtype column into multiple columns, since we have no way to
* notify the caller that it should do that.)
*
* check_sql_fn_retval will also construct a JunkFilter we can use to
* coerce the returned rowtype to the desired form (unless the result type
* is VOID, in which case there's nothing to coerce to).
*/
fcache->returnsTuple = check_sql_fn_retval(foid,
rettype,
flat_query_list,
NULL,
&fcache->junkFilter);
if (fcache->returnsTuple)
{
/* Make sure output rowtype is properly blessed */
BlessTupleDesc(fcache->junkFilter->jf_resultSlot->tts_tupleDescriptor);
}
else if (fcache->returnsSet && type_is_rowtype(fcache->rettype))
{
/*
* Returning rowtype as if it were scalar --- materialize won't work.
* Right now it's sufficient to override any caller preference for
* materialize mode, but to add more smarts in init_execution_state
* about this, we'd probably need a three-way flag instead of bool.
*/
lazyEvalOK = true;
}
/* Finally, plan the queries */
fcache->func_state = init_execution_state(queryTree_list,
fcache,
lazyEvalOK);
/* Mark fcache with time of creation to show it's valid */
fcache->lxid = MyProc->lxid;
fcache->subxid = GetCurrentSubTransactionId();
ReleaseSysCache(procedureTuple);
MemoryContextSwitchTo(oldcontext);
}
/* Start up execution of one execution_state node */
static void
postquel_start(execution_state *es, SQLFunctionCachePtr fcache)
{
DestReceiver *dest;
Assert(es->qd == NULL);
/* Caller should have ensured a suitable snapshot is active */
Assert(ActiveSnapshotSet());
/*
* If this query produces the function result, send its output to the
* tuplestore; else discard any output.
*/
if (es->setsResult)
{
DR_sqlfunction *myState;
dest = CreateDestReceiver(DestSQLFunction);
/* pass down the needed info to the dest receiver routines */
myState = (DR_sqlfunction *) dest;
Assert(myState->pub.mydest == DestSQLFunction);
myState->tstore = fcache->tstore;
myState->cxt = CurrentMemoryContext;
myState->filter = fcache->junkFilter;
}
else
dest = None_Receiver;
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
es->qd = CreateQueryDesc(es->stmt,
fcache->src,
GetActiveSnapshot(),
InvalidSnapshot,
dest,
fcache->paramLI,
es->qd ? es->qd->queryEnv : NULL,
0);
/* Utility commands don't need Executor. */
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
if (es->qd->operation != CMD_UTILITY)
{
/*
* In lazyEval mode, do not let the executor set up an AfterTrigger
* context. This is necessary not just an optimization, because we
* mustn't exit from the function execution with a stacked
* AfterTrigger level still active. We are careful not to select
* lazyEval mode for any statement that could possibly queue triggers.
*/
int eflags;
if (es->lazyEval)
eflags = EXEC_FLAG_SKIP_TRIGGERS;
else
eflags = 0; /* default run-to-completion flags */
ExecutorStart(es->qd, eflags);
}
es->status = F_EXEC_RUN;
}
/* Run one execution_state; either to completion or to first result row */
/* Returns true if we ran to completion */
static bool
postquel_getnext(execution_state *es, SQLFunctionCachePtr fcache)
{
bool result;
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
if (es->qd->operation == CMD_UTILITY)
{
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
ProcessUtility(es->qd->plannedstmt,
fcache->src,
PROCESS_UTILITY_QUERY,
es->qd->params,
es->qd->queryEnv,
es->qd->dest,
NULL);
result = true; /* never stops early */
}
else
{
/* Run regular commands to completion unless lazyEval */
uint64 count = (es->lazyEval) ? 1 : 0;
2017-03-23 18:05:48 +01:00
ExecutorRun(es->qd, ForwardScanDirection, count, !fcache->returnsSet || !es->lazyEval);
/*
* If we requested run to completion OR there was no tuple returned,
* command must be complete.
*/
result = (count == 0 || es->qd->estate->es_processed == 0);
}
return result;
}
/* Shut down execution of one execution_state node */
static void
postquel_end(execution_state *es)
{
/* mark status done to ensure we don't do ExecutorEnd twice */
es->status = F_EXEC_DONE;
/* Utility commands don't need Executor. */
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
if (es->qd->operation != CMD_UTILITY)
{
ExecutorFinish(es->qd);
ExecutorEnd(es->qd);
}
(*es->qd->dest->rDestroy) (es->qd->dest);
FreeQueryDesc(es->qd);
es->qd = NULL;
}
/* Build ParamListInfo array representing current arguments */
static void
postquel_sub_params(SQLFunctionCachePtr fcache,
FunctionCallInfo fcinfo)
{
int nargs = fcinfo->nargs;
if (nargs > 0)
{
ParamListInfo paramLI;
int i;
if (fcache->paramLI == NULL)
{
paramLI = (ParamListInfo)
palloc(offsetof(ParamListInfoData, params) +
nargs * sizeof(ParamExternData));
/* we have static list of params, so no hooks needed */
paramLI->paramFetch = NULL;
paramLI->paramFetchArg = NULL;
paramLI->parserSetup = NULL;
paramLI->parserSetupArg = NULL;
paramLI->numParams = nargs;
paramLI->paramMask = NULL;
fcache->paramLI = paramLI;
}
else
{
paramLI = fcache->paramLI;
Assert(paramLI->numParams == nargs);
}
for (i = 0; i < nargs; i++)
{
ParamExternData *prm = &paramLI->params[i];
prm->value = fcinfo->arg[i];
prm->isnull = fcinfo->argnull[i];
prm->pflags = 0;
prm->ptype = fcache->pinfo->argtypes[i];
}
}
else
fcache->paramLI = NULL;
}
/*
* Extract the SQL function's value from a single result row. This is used
* both for scalar (non-set) functions and for each row of a lazy-eval set
* result.
*/
static Datum
postquel_get_single_result(TupleTableSlot *slot,
FunctionCallInfo fcinfo,
SQLFunctionCachePtr fcache,
MemoryContext resultcontext)
{
Datum value;
MemoryContext oldcontext;
/*
2005-10-15 04:49:52 +02:00
* Set up to return the function value. For pass-by-reference datatypes,
* be sure to allocate the result in resultcontext, not the current memory
* context (which has query lifespan). We can't leave the data in the
* TupleTableSlot because we intend to clear the slot before returning.
*/
oldcontext = MemoryContextSwitchTo(resultcontext);
if (fcache->returnsTuple)
{
/* We must return the whole tuple as a Datum. */
fcinfo->isnull = false;
value = ExecFetchSlotTupleDatum(slot);
}
else
{
/*
2005-10-15 04:49:52 +02:00
* Returning a scalar, which we have to extract from the first column
* of the SELECT result, and then copy into result context if needed.
*/
value = slot_getattr(slot, 1, &(fcinfo->isnull));
if (!fcinfo->isnull)
value = datumCopy(value, fcache->typbyval, fcache->typlen);
}
MemoryContextSwitchTo(oldcontext);
return value;
}
/*
* fmgr_sql: function call manager for SQL functions
*/
Datum
fmgr_sql(PG_FUNCTION_ARGS)
{
SQLFunctionCachePtr fcache;
ErrorContextCallback sqlerrcontext;
MemoryContext oldcontext;
bool randomAccess;
bool lazyEvalOK;
bool is_first;
bool pushed_snapshot;
execution_state *es;
TupleTableSlot *slot;
Datum result;
2011-04-10 17:42:00 +02:00
List *eslist;
ListCell *eslc;
/*
* Setup error traceback support for ereport()
*/
sqlerrcontext.callback = sql_exec_error_callback;
sqlerrcontext.arg = fcinfo->flinfo;
sqlerrcontext.previous = error_context_stack;
error_context_stack = &sqlerrcontext;
/* Check call context */
if (fcinfo->flinfo->fn_retset)
{
ReturnSetInfo *rsi = (ReturnSetInfo *) fcinfo->resultinfo;
/*
* For simplicity, we require callers to support both set eval modes.
* There are cases where we must use one or must use the other, and
2010-02-26 03:01:40 +01:00
* it's not really worthwhile to postpone the check till we know. But
* note we do not require caller to provide an expectedDesc.
*/
if (!rsi || !IsA(rsi, ReturnSetInfo) ||
(rsi->allowedModes & SFRM_ValuePerCall) == 0 ||
(rsi->allowedModes & SFRM_Materialize) == 0)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("set-valued function called in context that cannot accept a set")));
randomAccess = rsi->allowedModes & SFRM_Materialize_Random;
lazyEvalOK = !(rsi->allowedModes & SFRM_Materialize_Preferred);
}
else
{
randomAccess = false;
lazyEvalOK = true;
}
/*
* Initialize fcache (build plans) if first time through; or re-initialize
* if the cache is stale.
*/
fcache = (SQLFunctionCachePtr) fcinfo->flinfo->fn_extra;
if (fcache != NULL)
{
if (fcache->lxid != MyProc->lxid ||
!SubTransactionIsActive(fcache->subxid))
{
/* It's stale; unlink and delete */
fcinfo->flinfo->fn_extra = NULL;
MemoryContextDelete(fcache->fcontext);
fcache = NULL;
}
}
if (fcache == NULL)
{
init_sql_fcache(fcinfo->flinfo, PG_GET_COLLATION(), lazyEvalOK);
fcache = (SQLFunctionCachePtr) fcinfo->flinfo->fn_extra;
}
/*
* Switch to context in which the fcache lives. This ensures that our
* tuplestore etc will have sufficient lifetime. The sub-executor is
* responsible for deleting per-tuple information. (XXX in the case of a
* long-lived FmgrInfo, this policy represents more memory leakage, but
* it's not entirely clear where to keep stuff instead.)
*/
oldcontext = MemoryContextSwitchTo(fcache->fcontext);
/*
* Find first unfinished query in function, and note whether it's the
* first query.
*/
eslist = fcache->func_state;
es = NULL;
is_first = true;
foreach(eslc, eslist)
{
es = (execution_state *) lfirst(eslc);
while (es && es->status == F_EXEC_DONE)
{
is_first = false;
es = es->next;
}
if (es)
break;
}
/*
2005-10-15 04:49:52 +02:00
* Convert params to appropriate format if starting a fresh execution. (If
* continuing execution, we can re-use prior params.)
*/
if (is_first && es && es->status == F_EXEC_START)
postquel_sub_params(fcache, fcinfo);
/*
* Build tuplestore to hold results, if we don't have one already. Note
* it's in the query-lifespan context.
*/
if (!fcache->tstore)
fcache->tstore = tuplestore_begin_heap(randomAccess, false, work_mem);
/*
* Execute each command in the function one after another until we either
* run out of commands or get a result row from a lazily-evaluated SELECT.
*
* Notes about snapshot management:
*
* In a read-only function, we just use the surrounding query's snapshot.
*
* In a non-read-only function, we rely on the fact that we'll never
* suspend execution between queries of the function: the only reason to
2011-04-10 17:42:00 +02:00
* suspend execution before completion is if we are returning a row from a
* lazily-evaluated SELECT. So, when first entering this loop, we'll
* either start a new query (and push a fresh snapshot) or re-establish
* the active snapshot from the existing query descriptor. If we need to
* start a new query in a subsequent execution of the loop, either we need
* a fresh snapshot (and pushed_snapshot is false) or the existing
* snapshot is on the active stack and we can just bump its command ID.
*/
pushed_snapshot = false;
while (es)
{
bool completed;
if (es->status == F_EXEC_START)
{
/*
* If not read-only, be sure to advance the command counter for
* each command, so that all work to date in this transaction is
* visible. Take a new snapshot if we don't have one yet,
* otherwise just bump the command ID in the existing snapshot.
*/
if (!fcache->readonly_func)
{
CommandCounterIncrement();
if (!pushed_snapshot)
{
PushActiveSnapshot(GetTransactionSnapshot());
pushed_snapshot = true;
}
else
UpdateActiveSnapshotCommandId();
}
postquel_start(es, fcache);
}
else if (!fcache->readonly_func && !pushed_snapshot)
{
/* Re-establish active snapshot when re-entering function */
PushActiveSnapshot(es->qd->snapshot);
pushed_snapshot = true;
}
completed = postquel_getnext(es, fcache);
/*
* If we ran the command to completion, we can shut it down now. Any
* row(s) we need to return are safely stashed in the tuplestore, and
* we want to be sure that, for example, AFTER triggers get fired
* before we return anything. Also, if the function doesn't return
* set, we can shut it down anyway because it must be a SELECT and we
* don't care about fetching any more result rows.
*/
if (completed || !fcache->returnsSet)
postquel_end(es);
/*
* Break from loop if we didn't shut down (implying we got a
* lazily-evaluated row). Otherwise we'll press on till the whole
* function is done, relying on the tuplestore to keep hold of the
* data to eventually be returned. This is necessary since an
* INSERT/UPDATE/DELETE RETURNING that sets the result might be
* followed by additional rule-inserted commands, and we want to
* finish doing all those commands before we return anything.
*/
if (es->status != F_EXEC_DONE)
break;
/*
* Advance to next execution_state, which might be in the next list.
*/
es = es->next;
while (!es)
{
eslc = lnext(eslc);
if (!eslc)
break; /* end of function */
es = (execution_state *) lfirst(eslc);
/*
2011-04-10 17:42:00 +02:00
* Flush the current snapshot so that we will take a new one for
* the new query list. This ensures that new snaps are taken at
2011-04-10 17:42:00 +02:00
* original-query boundaries, matching the behavior of interactive
* execution.
*/
if (pushed_snapshot)
{
PopActiveSnapshot();
pushed_snapshot = false;
}
}
}
1997-08-29 11:05:57 +02:00
/*
* The tuplestore now contains whatever row(s) we are supposed to return.
1997-08-29 11:05:57 +02:00
*/
if (fcache->returnsSet)
{
ReturnSetInfo *rsi = (ReturnSetInfo *) fcinfo->resultinfo;
if (es)
{
/*
* If we stopped short of being done, we must have a lazy-eval
* row.
*/
Assert(es->lazyEval);
/* Re-use the junkfilter's output slot to fetch back the tuple */
Assert(fcache->junkFilter);
slot = fcache->junkFilter->jf_resultSlot;
if (!tuplestore_gettupleslot(fcache->tstore, true, false, slot))
elog(ERROR, "failed to fetch lazy-eval tuple");
/* Extract the result as a datum, and copy out from the slot */
result = postquel_get_single_result(slot, fcinfo,
fcache, oldcontext);
/* Clear the tuplestore, but keep it for next time */
/* NB: this might delete the slot's content, but we don't care */
tuplestore_clear(fcache->tstore);
/*
* Let caller know we're not finished.
*/
rsi->isDone = ExprMultipleResult;
/*
* Ensure we will get shut down cleanly if the exprcontext is not
* run to completion.
*/
if (!fcache->shutdown_reg)
{
RegisterExprContextCallback(rsi->econtext,
ShutdownSQLFunction,
PointerGetDatum(fcache));
fcache->shutdown_reg = true;
}
}
else if (fcache->lazyEval)
{
/*
* We are done with a lazy evaluation. Clean up.
*/
tuplestore_clear(fcache->tstore);
/*
* Let caller know we're finished.
*/
rsi->isDone = ExprEndResult;
fcinfo->isnull = true;
result = (Datum) 0;
/* Deregister shutdown callback, if we made one */
if (fcache->shutdown_reg)
{
UnregisterExprContextCallback(rsi->econtext,
ShutdownSQLFunction,
PointerGetDatum(fcache));
fcache->shutdown_reg = false;
}
}
else
{
/*
* We are done with a non-lazy evaluation. Return whatever is in
* the tuplestore. (It is now caller's responsibility to free the
* tuplestore when done.)
*/
rsi->returnMode = SFRM_Materialize;
rsi->setResult = fcache->tstore;
fcache->tstore = NULL;
Faster expression evaluation and targetlist projection. This replaces the old, recursive tree-walk based evaluation, with non-recursive, opcode dispatch based, expression evaluation. Projection is now implemented as part of expression evaluation. This both leads to significant performance improvements, and makes future just-in-time compilation of expressions easier. The speed gains primarily come from: - non-recursive implementation reduces stack usage / overhead - simple sub-expressions are implemented with a single jump, without function calls - sharing some state between different sub-expressions - reduced amount of indirect/hard to predict memory accesses by laying out operation metadata sequentially; including the avoidance of nearly all of the previously used linked lists - more code has been moved to expression initialization, avoiding constant re-checks at evaluation time Future just-in-time compilation (JIT) has become easier, as demonstrated by released patches intended to be merged in a later release, for primarily two reasons: Firstly, due to a stricter split between expression initialization and evaluation, less code has to be handled by the JIT. Secondly, due to the non-recursive nature of the generated "instructions", less performance-critical code-paths can easily be shared between interpreted and compiled evaluation. The new framework allows for significant future optimizations. E.g.: - basic infrastructure for to later reduce the per executor-startup overhead of expression evaluation, by caching state in prepared statements. That'd be helpful in OLTPish scenarios where initialization overhead is measurable. - optimizing the generated "code". A number of proposals for potential work has already been made. - optimizing the interpreter. Similarly a number of proposals have been made here too. The move of logic into the expression initialization step leads to some backward-incompatible changes: - Function permission checks are now done during expression initialization, whereas previously they were done during execution. In edge cases this can lead to errors being raised that previously wouldn't have been, e.g. a NULL array being coerced to a different array type previously didn't perform checks. - The set of domain constraints to be checked, is now evaluated once during expression initialization, previously it was re-built every time a domain check was evaluated. For normal queries this doesn't change much, but e.g. for plpgsql functions, which caches ExprStates, the old set could stick around longer. The behavior around might still change. Author: Andres Freund, with significant changes by Tom Lane, changes by Heikki Linnakangas Reviewed-By: Tom Lane, Heikki Linnakangas Discussion: https://postgr.es/m/20161206034955.bh33paeralxbtluv@alap3.anarazel.de
2017-03-14 23:45:36 +01:00
/* must copy desc because execSRF.c will free it */
if (fcache->junkFilter)
rsi->setDesc = CreateTupleDescCopy(fcache->junkFilter->jf_cleanTupType);
fcinfo->isnull = true;
result = (Datum) 0;
/* Deregister shutdown callback, if we made one */
if (fcache->shutdown_reg)
{
UnregisterExprContextCallback(rsi->econtext,
ShutdownSQLFunction,
PointerGetDatum(fcache));
fcache->shutdown_reg = false;
}
}
}
else
{
/*
* Non-set function. If we got a row, return it; else return NULL.
*/
if (fcache->junkFilter)
{
/* Re-use the junkfilter's output slot to fetch back the tuple */
slot = fcache->junkFilter->jf_resultSlot;
if (tuplestore_gettupleslot(fcache->tstore, true, false, slot))
result = postquel_get_single_result(slot, fcinfo,
fcache, oldcontext);
else
{
fcinfo->isnull = true;
result = (Datum) 0;
}
}
else
{
/* Should only get here for VOID functions */
Assert(fcache->rettype == VOIDOID);
fcinfo->isnull = true;
result = (Datum) 0;
}
/* Clear the tuplestore, but keep it for next time */
tuplestore_clear(fcache->tstore);
}
/* Pop snapshot if we have pushed one */
if (pushed_snapshot)
PopActiveSnapshot();
/*
* If we've gone through every command in the function, we are done. Reset
* the execution states to start over again on next call.
*/
if (es == NULL)
{
foreach(eslc, fcache->func_state)
{
es = (execution_state *) lfirst(eslc);
while (es)
{
es->status = F_EXEC_START;
es = es->next;
}
}
}
error_context_stack = sqlerrcontext.previous;
MemoryContextSwitchTo(oldcontext);
return result;
}
/*
* error context callback to let us supply a call-stack traceback
*/
static void
sql_exec_error_callback(void *arg)
{
FmgrInfo *flinfo = (FmgrInfo *) arg;
SQLFunctionCachePtr fcache = (SQLFunctionCachePtr) flinfo->fn_extra;
int syntaxerrposition;
/*
* We can do nothing useful if init_sql_fcache() didn't get as far as
* saving the function name
*/
if (fcache == NULL || fcache->fname == NULL)
return;
/*
2005-10-15 04:49:52 +02:00
* If there is a syntax error position, convert to internal syntax error
*/
syntaxerrposition = geterrposition();
if (syntaxerrposition > 0 && fcache->src != NULL)
{
errposition(0);
internalerrposition(syntaxerrposition);
internalerrquery(fcache->src);
}
/*
2005-10-15 04:49:52 +02:00
* Try to determine where in the function we failed. If there is a query
* with non-null QueryDesc, finger it. (We check this rather than looking
2005-10-15 04:49:52 +02:00
* for F_EXEC_RUN state, so that errors during ExecutorStart or
* ExecutorEnd are blamed on the appropriate query; see postquel_start and
* postquel_end.)
*/
if (fcache->func_state)
{
execution_state *es;
2003-08-04 02:43:34 +02:00
int query_num;
ListCell *lc;
es = NULL;
query_num = 1;
foreach(lc, fcache->func_state)
{
es = (execution_state *) lfirst(lc);
while (es)
{
if (es->qd)
{
errcontext("SQL function \"%s\" statement %d",
fcache->fname, query_num);
break;
}
es = es->next;
}
if (es)
break;
query_num++;
}
if (es == NULL)
{
/*
* couldn't identify a running query; might be function entry,
* function exit, or between queries.
*/
errcontext("SQL function \"%s\"", fcache->fname);
}
}
else
{
/*
2010-07-06 21:19:02 +02:00
* Assume we failed during init_sql_fcache(). (It's possible that the
* function actually has an empty body, but in that case we may as
* well report all errors as being "during startup".)
*/
errcontext("SQL function \"%s\" during startup", fcache->fname);
}
}
/*
* callback function in case a function-returning-set needs to be shut down
* before it has been run to completion
*/
static void
ShutdownSQLFunction(Datum arg)
{
SQLFunctionCachePtr fcache = (SQLFunctionCachePtr) DatumGetPointer(arg);
execution_state *es;
2011-04-10 17:42:00 +02:00
ListCell *lc;
foreach(lc, fcache->func_state)
{
es = (execution_state *) lfirst(lc);
while (es)
{
/* Shut down anything still running */
if (es->status == F_EXEC_RUN)
{
/* Re-establish active snapshot for any called functions */
if (!fcache->readonly_func)
PushActiveSnapshot(es->qd->snapshot);
postquel_end(es);
if (!fcache->readonly_func)
PopActiveSnapshot();
}
/* Reset states to START in case we're called again */
es->status = F_EXEC_START;
es = es->next;
}
}
/* Release tuplestore if we have one */
if (fcache->tstore)
tuplestore_end(fcache->tstore);
fcache->tstore = NULL;
/* execUtils will deregister the callback... */
fcache->shutdown_reg = false;
}
/*
* check_sql_fn_retval() -- check return value of a list of sql parse trees.
*
* The return value of a sql function is the value returned by the last
* canSetTag query in the function. We do some ad-hoc type checking here
* to be sure that the user is returning the type he claims. There are
* also a couple of strange-looking features to assist callers in dealing
* with allowed special cases, such as binary-compatible result types.
*
* For a polymorphic function the passed rettype must be the actual resolved
* output type of the function; we should never see a polymorphic pseudotype
* such as ANYELEMENT as rettype. (This means we can't check the type during
* function definition of a polymorphic function.)
*
* This function returns true if the sql function returns the entire tuple
* result of its final statement, or false if it returns just the first column
* result of that statement. It throws an error if the final statement doesn't
* return the right type at all.
*
* Note that because we allow "SELECT rowtype_expression", the result can be
* false even when the declared function return type is a rowtype.
*
* If modifyTargetList isn't NULL, the function will modify the final
* statement's targetlist in two cases:
* (1) if the tlist returns values that are binary-coercible to the expected
* type rather than being exactly the expected type. RelabelType nodes will
* be inserted to make the result types match exactly.
* (2) if there are dropped columns in the declared result rowtype. NULL
* output columns will be inserted in the tlist to match them.
* (Obviously the caller must pass a parsetree that is okay to modify when
* using this flag.) Note that this flag does not affect whether the tlist is
* considered to be a legal match to the result type, only how we react to
* allowed not-exact-match cases. *modifyTargetList will be set true iff
* we had to make any "dangerous" changes that could modify the semantics of
* the statement. If it is set true, the caller should not use the modified
* statement, but for simplicity we apply the changes anyway.
*
* If junkFilter isn't NULL, then *junkFilter is set to a JunkFilter defined
* to convert the function's tuple result to the correct output tuple type.
* Exception: if the function is defined to return VOID then *junkFilter is
* set to NULL.
*/
bool
check_sql_fn_retval(Oid func_id, Oid rettype, List *queryTreeList,
bool *modifyTargetList,
JunkFilter **junkFilter)
{
Query *parse;
List **tlist_ptr;
List *tlist;
int tlistlen;
char fn_typtype;
Oid restype;
ListCell *lc;
AssertArg(!IsPolymorphicType(rettype));
if (modifyTargetList)
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
*modifyTargetList = false; /* initialize for no change */
if (junkFilter)
*junkFilter = NULL; /* initialize in case of VOID result */
/*
* Find the last canSetTag query in the list. This isn't necessarily the
* last parsetree, because rule rewriting can insert queries after what
* the user wrote.
*/
parse = NULL;
foreach(lc, queryTreeList)
{
Query *q = lfirst_node(Query, lc);
if (q->canSetTag)
parse = q;
}
/*
* If it's a plain SELECT, it returns whatever the targetlist says.
* Otherwise, if it's INSERT/UPDATE/DELETE with RETURNING, it returns
* that. Otherwise, the function return type must be VOID.
*
2007-11-15 22:14:46 +01:00
* Note: eventually replace this test with QueryReturnsTuples? We'd need
* a more general method of determining the output type, though. Also, it
* seems too dangerous to consider FETCH or EXECUTE as returning a
* determinable rowtype, since they depend on relatively short-lived
* entities.
*/
if (parse &&
Change representation of statement lists, and add statement location info. This patch makes several changes that improve the consistency of representation of lists of statements. It's always been the case that the output of parse analysis is a list of Query nodes, whatever the types of the individual statements in the list. This patch brings similar consistency to the outputs of raw parsing and planning steps: * The output of raw parsing is now always a list of RawStmt nodes; the statement-type-dependent nodes are one level down from that. * The output of pg_plan_queries() is now always a list of PlannedStmt nodes, even for utility statements. In the case of a utility statement, "planning" just consists of wrapping a CMD_UTILITY PlannedStmt around the utility node. This list representation is now used in Portal and CachedPlan plan lists, replacing the former convention of intermixing PlannedStmts with bare utility-statement nodes. Now, every list of statements has a consistent head-node type depending on how far along it is in processing. This allows changing many places that formerly used generic "Node *" pointers to use a more specific pointer type, thus reducing the number of IsA() tests and casts needed, as well as improving code clarity. Also, the post-parse-analysis representation of DECLARE CURSOR is changed so that it looks more like EXPLAIN, PREPARE, etc. That is, the contained SELECT remains a child of the DeclareCursorStmt rather than getting flipped around to be the other way. It's now true for both Query and PlannedStmt that utilityStmt is non-null if and only if commandType is CMD_UTILITY. That allows simplifying a lot of places that were testing both fields. (I think some of those were just defensive programming, but in many places, it was actually necessary to avoid confusing DECLARE CURSOR with SELECT.) Because PlannedStmt carries a canSetTag field, we're also able to get rid of some ad-hoc rules about how to reconstruct canSetTag for a bare utility statement; specifically, the assumption that a utility is canSetTag if and only if it's the only one in its list. While I see no near-term need for relaxing that restriction, it's nice to get rid of the ad-hocery. The API of ProcessUtility() is changed so that what it's passed is the wrapper PlannedStmt not just the bare utility statement. This will affect all users of ProcessUtility_hook, but the changes are pretty trivial; see the affected contrib modules for examples of the minimum change needed. (Most compilers should give pointer-type-mismatch warnings for uncorrected code.) There's also a change in the API of ExplainOneQuery_hook, to pass through cursorOptions instead of expecting hook functions to know what to pick. This is needed because of the DECLARE CURSOR changes, but really should have been done in 9.6; it's unlikely that any extant hook functions know about using CURSOR_OPT_PARALLEL_OK. Finally, teach gram.y to save statement boundary locations in RawStmt nodes, and pass those through to Query and PlannedStmt nodes. This allows more intelligent handling of cases where a source query string contains multiple statements. This patch doesn't actually do anything with the information, but a follow-on patch will. (Passing this information through cleanly is the true motivation for these changes; while I think this is all good cleanup, it's unlikely we'd have bothered without this end goal.) catversion bump because addition of location fields to struct Query affects stored rules. This patch is by me, but it owes a good deal to Fabien Coelho who did a lot of preliminary work on the problem, and also reviewed the patch. Discussion: https://postgr.es/m/alpine.DEB.2.20.1612200926310.29821@lancre
2017-01-14 22:02:35 +01:00
parse->commandType == CMD_SELECT)
{
tlist_ptr = &parse->targetList;
tlist = parse->targetList;
}
else if (parse &&
(parse->commandType == CMD_INSERT ||
parse->commandType == CMD_UPDATE ||
parse->commandType == CMD_DELETE) &&
parse->returningList)
{
tlist_ptr = &parse->returningList;
tlist = parse->returningList;
}
else
{
/* Empty function body, or last statement is a utility command */
if (rettype != VOIDOID)
ereport(ERROR,
(errcode(ERRCODE_INVALID_FUNCTION_DEFINITION),
errmsg("return type mismatch in function declared to return %s",
format_type_be(rettype)),
errdetail("Function's final statement must be SELECT or INSERT/UPDATE/DELETE RETURNING.")));
return false;
}
/*
* OK, check that the targetlist returns something matching the declared
* type. (We used to insist that the declared type not be VOID in this
2007-11-15 22:14:46 +01:00
* case, but that makes it hard to write a void function that exits after
* calling another void function. Instead, we insist that the tlist
2007-11-15 22:14:46 +01:00
* return void ... so void is treated as if it were a scalar type below.)
*/
/*
* Count the non-junk entries in the result targetlist.
*/
tlistlen = ExecCleanTargetListLength(tlist);
fn_typtype = get_typtype(rettype);
if (fn_typtype == TYPTYPE_BASE ||
fn_typtype == TYPTYPE_DOMAIN ||
fn_typtype == TYPTYPE_ENUM ||
fn_typtype == TYPTYPE_RANGE ||
rettype == VOIDOID)
{
/*
* For scalar-type returns, the target list must have exactly one
* non-junk entry, and its type must agree with what the user
* declared; except we allow binary-compatible types too.
*/
TargetEntry *tle;
if (tlistlen != 1)
ereport(ERROR,
(errcode(ERRCODE_INVALID_FUNCTION_DEFINITION),
errmsg("return type mismatch in function declared to return %s",
format_type_be(rettype)),
errdetail("Final statement must return exactly one column.")));
/* We assume here that non-junk TLEs must come first in tlists */
tle = (TargetEntry *) linitial(tlist);
Assert(!tle->resjunk);
restype = exprType((Node *) tle->expr);
if (!IsBinaryCoercible(restype, rettype))
ereport(ERROR,
(errcode(ERRCODE_INVALID_FUNCTION_DEFINITION),
errmsg("return type mismatch in function declared to return %s",
format_type_be(rettype)),
errdetail("Actual return type is %s.",
format_type_be(restype))));
if (modifyTargetList && restype != rettype)
{
tle->expr = (Expr *) makeRelabelType(tle->expr,
rettype,
-1,
get_typcollation(rettype),
COERCE_IMPLICIT_CAST);
/* Relabel is dangerous if TLE is a sort/group or setop column */
if (tle->ressortgroupref != 0 || parse->setOperations)
*modifyTargetList = true;
}
/* Set up junk filter if needed */
if (junkFilter)
*junkFilter = ExecInitJunkFilter(tlist, false, NULL);
}
else if (fn_typtype == TYPTYPE_COMPOSITE || rettype == RECORDOID)
{
/* Returns a rowtype */
TupleDesc tupdesc;
2005-10-15 04:49:52 +02:00
int tupnatts; /* physical number of columns in tuple */
int tuplogcols; /* # of nondeleted columns in tuple */
int colindex; /* physical column index */
List *newtlist; /* new non-junk tlist entries */
List *junkattrs; /* new junk tlist entries */
/*
2005-10-15 04:49:52 +02:00
* If the target list is of length 1, and the type of the varnode in
* the target list matches the declared return type, this is okay.
* This can happen, for example, where the body of the function is
* 'SELECT func2()', where func2 has the same composite return type as
* the function that's calling it.
*
* XXX Note that if rettype is RECORD, the IsBinaryCoercible check
* will succeed for any composite restype. For the moment we rely on
* runtime type checking to catch any discrepancy, but it'd be nice to
* do better at parse time.
*/
if (tlistlen == 1)
{
TargetEntry *tle = (TargetEntry *) linitial(tlist);
Assert(!tle->resjunk);
restype = exprType((Node *) tle->expr);
if (IsBinaryCoercible(restype, rettype))
{
if (modifyTargetList && restype != rettype)
{
tle->expr = (Expr *) makeRelabelType(tle->expr,
rettype,
-1,
get_typcollation(rettype),
COERCE_IMPLICIT_CAST);
/* Relabel is dangerous if sort/group or setop column */
if (tle->ressortgroupref != 0 || parse->setOperations)
*modifyTargetList = true;
}
/* Set up junk filter if needed */
if (junkFilter)
*junkFilter = ExecInitJunkFilter(tlist, false, NULL);
return false; /* NOT returning whole tuple */
}
}
/* Is the rowtype fixed, or determined only at runtime? */
if (get_func_result_type(func_id, NULL, &tupdesc) != TYPEFUNC_COMPOSITE)
{
/*
2005-10-15 04:49:52 +02:00
* Assume we are returning the whole tuple. Crosschecking against
* what the caller expects will happen at runtime.
*/
if (junkFilter)
*junkFilter = ExecInitJunkFilter(tlist, false, NULL);
return true;
}
Assert(tupdesc);
/*
2005-10-15 04:49:52 +02:00
* Verify that the targetlist matches the return tuple type. We scan
* the non-deleted attributes to ensure that they match the datatypes
* of the non-resjunk columns. For deleted attributes, insert NULL
* result columns if the caller asked for that.
*/
tupnatts = tupdesc->natts;
tuplogcols = 0; /* we'll count nondeleted cols as we go */
colindex = 0;
newtlist = NIL; /* these are only used if modifyTargetList */
junkattrs = NIL;
foreach(lc, tlist)
{
TargetEntry *tle = (TargetEntry *) lfirst(lc);
Form_pg_attribute attr;
Oid tletype;
Oid atttype;
if (tle->resjunk)
{
if (modifyTargetList)
junkattrs = lappend(junkattrs, tle);
continue;
}
do
{
colindex++;
if (colindex > tupnatts)
ereport(ERROR,
(errcode(ERRCODE_INVALID_FUNCTION_DEFINITION),
errmsg("return type mismatch in function declared to return %s",
format_type_be(rettype)),
errdetail("Final statement returns too many columns.")));
attr = tupdesc->attrs[colindex - 1];
if (attr->attisdropped && modifyTargetList)
{
2010-02-26 03:01:40 +01:00
Expr *null_expr;
/* The type of the null we insert isn't important */
null_expr = (Expr *) makeConst(INT4OID,
-1,
InvalidOid,
sizeof(int32),
(Datum) 0,
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
true, /* isnull */
true /* byval */ );
newtlist = lappend(newtlist,
makeTargetEntry(null_expr,
colindex,
NULL,
false));
/* NULL insertion is dangerous in a setop */
if (parse->setOperations)
*modifyTargetList = true;
}
} while (attr->attisdropped);
tuplogcols++;
tletype = exprType((Node *) tle->expr);
atttype = attr->atttypid;
if (!IsBinaryCoercible(tletype, atttype))
ereport(ERROR,
(errcode(ERRCODE_INVALID_FUNCTION_DEFINITION),
errmsg("return type mismatch in function declared to return %s",
format_type_be(rettype)),
errdetail("Final statement returns %s instead of %s at column %d.",
format_type_be(tletype),
format_type_be(atttype),
tuplogcols)));
if (modifyTargetList)
{
if (tletype != atttype)
{
tle->expr = (Expr *) makeRelabelType(tle->expr,
atttype,
-1,
get_typcollation(atttype),
COERCE_IMPLICIT_CAST);
/* Relabel is dangerous if sort/group or setop column */
if (tle->ressortgroupref != 0 || parse->setOperations)
*modifyTargetList = true;
}
tle->resno = colindex;
newtlist = lappend(newtlist, tle);
}
}
/* remaining columns in tupdesc had better all be dropped */
for (colindex++; colindex <= tupnatts; colindex++)
{
if (!tupdesc->attrs[colindex - 1]->attisdropped)
ereport(ERROR,
(errcode(ERRCODE_INVALID_FUNCTION_DEFINITION),
errmsg("return type mismatch in function declared to return %s",
format_type_be(rettype)),
errdetail("Final statement returns too few columns.")));
if (modifyTargetList)
{
2010-02-26 03:01:40 +01:00
Expr *null_expr;
/* The type of the null we insert isn't important */
null_expr = (Expr *) makeConst(INT4OID,
-1,
InvalidOid,
sizeof(int32),
(Datum) 0,
2010-02-26 03:01:40 +01:00
true, /* isnull */
true /* byval */ );
newtlist = lappend(newtlist,
makeTargetEntry(null_expr,
colindex,
NULL,
false));
/* NULL insertion is dangerous in a setop */
if (parse->setOperations)
*modifyTargetList = true;
}
}
if (modifyTargetList)
{
/* ensure resjunk columns are numbered correctly */
foreach(lc, junkattrs)
{
TargetEntry *tle = (TargetEntry *) lfirst(lc);
tle->resno = colindex++;
}
/* replace the tlist with the modified one */
*tlist_ptr = list_concat(newtlist, junkattrs);
}
/* Set up junk filter if needed */
if (junkFilter)
*junkFilter = ExecInitJunkFilterConversion(tlist,
CreateTupleDescCopy(tupdesc),
2005-10-15 04:49:52 +02:00
NULL);
/* Report that we are returning entire tuple result */
return true;
}
else
ereport(ERROR,
(errcode(ERRCODE_INVALID_FUNCTION_DEFINITION),
2005-10-15 04:49:52 +02:00
errmsg("return type %s is not supported for SQL functions",
format_type_be(rettype))));
return false;
}
/*
* CreateSQLFunctionDestReceiver -- create a suitable DestReceiver object
*/
DestReceiver *
CreateSQLFunctionDestReceiver(void)
{
DR_sqlfunction *self = (DR_sqlfunction *) palloc0(sizeof(DR_sqlfunction));
self->pub.receiveSlot = sqlfunction_receive;
self->pub.rStartup = sqlfunction_startup;
self->pub.rShutdown = sqlfunction_shutdown;
self->pub.rDestroy = sqlfunction_destroy;
self->pub.mydest = DestSQLFunction;
/* private fields will be set by postquel_start */
return (DestReceiver *) self;
}
/*
* sqlfunction_startup --- executor startup
*/
static void
sqlfunction_startup(DestReceiver *self, int operation, TupleDesc typeinfo)
{
/* no-op */
}
/*
* sqlfunction_receive --- receive one tuple
*/
static bool
sqlfunction_receive(TupleTableSlot *slot, DestReceiver *self)
{
DR_sqlfunction *myState = (DR_sqlfunction *) self;
/* Filter tuple as needed */
slot = ExecFilterJunk(myState->filter, slot);
/* Store the filtered tuple into the tuplestore */
tuplestore_puttupleslot(myState->tstore, slot);
return true;
}
/*
* sqlfunction_shutdown --- executor end
*/
static void
sqlfunction_shutdown(DestReceiver *self)
{
/* no-op */
}
/*
* sqlfunction_destroy --- release DestReceiver object
*/
static void
sqlfunction_destroy(DestReceiver *self)
{
pfree(self);
}