/*------------------------------------------------------------------ * * Foreign data wrapper for TDS (Sybase and Microsoft SQL Server) * * Author: Geoff Montee * Name: tds_fdw * File: tds_fdw/src/deparse.c * * Description: * This is a PostgreSQL foreign data wrapper for use to connect to databases that use TDS, * such as Sybase databases and Microsoft SQL server. * * This foreign data wrapper requires requires a library that uses the DB-Library interface, * such as FreeTDS (http://www.freetds.org/). This has been tested with FreeTDS, but not * the proprietary implementations of DB-Library. *---------------------------------------------------------------------------- */ /*------------------------------------------------------------------------- * * deparse.c * Query deparser for tds_fdw. This file was copied from postgres_fdw * * This file includes functions that examine query WHERE clauses to see * whether they're safe to send to the remote server for execution, as * well as functions to construct the query text to be sent. The latter * functionality is annoyingly duplicative of ruleutils.c, but there are * enough special considerations that it seems best to keep this separate. * One saving grace is that we only need deparse logic for node types that * we consider safe to send. * * We assume that the remote session's search_path is exactly "pg_catalog", * and thus we need schema-qualify all and only names outside pg_catalog. * * We do not consider that it is ever safe to send COLLATE expressions to * the remote server: it might not have the same collation names we do. * (Later we might consider it safe to send COLLATE "C", but even that would * fail on old remote servers.) An expression is considered safe to send * only if all operator/function input collations used in it are traceable to * Var(s) of the foreign table. That implies that if the remote server gets * a different answer than we do, the foreign table's columns are not marked * with collations that match the remote table's columns, which we can * consider to be user error. * * Portions Copyright (c) 2012-2015, PostgreSQL Global Development Group * * IDENTIFICATION * contrib/postgres_fdw/deparse.c * *------------------------------------------------------------------------- */ #include "postgres.h" #include "access/heapam.h" #if (PG_VERSION_NUM >= 90300) #include "access/htup_details.h" #endif #include "access/sysattr.h" #include "catalog/pg_collation.h" #include "catalog/pg_namespace.h" #include "catalog/pg_operator.h" #include "catalog/pg_proc.h" #include "catalog/pg_type.h" #include "commands/defrem.h" #include "nodes/nodeFuncs.h" #include "optimizer/clauses.h" #include "optimizer/var.h" #include "parser/parsetree.h" #include "utils/builtins.h" #include "utils/lsyscache.h" #include "utils/rel.h" #include "utils/syscache.h" #include "tds_fdw.h" #include "deparse.h" /* * Global context for foreign_expr_walker's search of an expression tree. */ typedef struct foreign_glob_cxt { PlannerInfo *root; /* global planner state */ RelOptInfo *foreignrel; /* the foreign relation we are planning for */ } foreign_glob_cxt; /* * Local (per-tree-level) context for foreign_expr_walker's search. * This is concerned with identifying collations used in the expression. */ typedef enum { FDW_COLLATE_NONE, /* expression is of a noncollatable type, or * it has default collation that is not * traceable to a foreign Var */ FDW_COLLATE_SAFE, /* collation derives from a foreign Var */ FDW_COLLATE_UNSAFE /* collation is non-default and derives from * something other than a foreign Var */ } FDWCollateState; typedef struct foreign_loc_cxt { Oid collation; /* OID of current collation, if any */ FDWCollateState state; /* state of current collation choice */ } foreign_loc_cxt; /* * Context for deparseExpr */ typedef struct deparse_expr_cxt { PlannerInfo *root; /* global planner state */ RelOptInfo *foreignrel; /* the foreign relation we are planning for */ StringInfo buf; /* output buffer to append to */ List **params_list; /* exprs that will become remote Params */ } deparse_expr_cxt; /* * Functions to determine whether an expression can be evaluated safely on * remote server. */ static bool foreign_expr_walker(Node *node, foreign_glob_cxt *glob_cxt, foreign_loc_cxt *outer_cxt); static char *deparse_type_name(Oid type_oid, int32 typemod); /* * Functions to construct string representation of a node tree. */ static void deparseTargetList(StringInfo buf, PlannerInfo *root, Index rtindex, Relation rel, Bitmapset *attrs_used, List **retrieved_attrs, TdsFdwOptionSet* option_set); static void deparseReturningList(StringInfo buf, PlannerInfo *root, Index rtindex, Relation rel, bool trig_after_row, List *returningList, List **retrieved_attrs, TdsFdwOptionSet* option_set); static void deparseColumnRef(StringInfo buf, int varno, int varattno, PlannerInfo *root); static void deparseRelation(StringInfo buf, Relation rel); static void deparseExpr(Expr *expr, deparse_expr_cxt *context); static void deparseVar(Var *node, deparse_expr_cxt *context); static void deparseConst(Const *node, deparse_expr_cxt *context); static void deparseParam(Param *node, deparse_expr_cxt *context); static void deparseArrayRef(ArrayRef *node, deparse_expr_cxt *context); static void deparseFuncExpr(FuncExpr *node, deparse_expr_cxt *context); static void deparseOpExpr(OpExpr *node, deparse_expr_cxt *context); static void deparseOperatorName(StringInfo buf, Form_pg_operator opform); static void deparseDistinctExpr(DistinctExpr *node, deparse_expr_cxt *context); static void deparseScalarArrayOpExpr(ScalarArrayOpExpr *node, deparse_expr_cxt *context); static void deparseRelabelType(RelabelType *node, deparse_expr_cxt *context); static void deparseBoolExpr(BoolExpr *node, deparse_expr_cxt *context); static void deparseNullTest(NullTest *node, deparse_expr_cxt *context); static void deparseArrayExpr(ArrayExpr *node, deparse_expr_cxt *context); static void printRemoteParam(int paramindex, Oid paramtype, int32 paramtypmod, deparse_expr_cxt *context); static void printRemotePlaceholder(Oid paramtype, int32 paramtypmod, deparse_expr_cxt *context); /* * Examine each qual clause in input_conds, and classify them into two groups, * which are returned as two lists: * - remote_conds contains expressions that can be evaluated remotely * - local_conds contains expressions that can't be evaluated remotely */ void classifyConditions(PlannerInfo *root, RelOptInfo *baserel, List *input_conds, List **remote_conds, List **local_conds) { ListCell *lc; *remote_conds = NIL; *local_conds = NIL; foreach(lc, input_conds) { RestrictInfo *ri = (RestrictInfo *) lfirst(lc); if (is_foreign_expr(root, baserel, ri->clause)) *remote_conds = lappend(*remote_conds, ri); else *local_conds = lappend(*local_conds, ri); } } /* * Returns true if given expr is safe to evaluate on the foreign server. */ bool is_foreign_expr(PlannerInfo *root, RelOptInfo *baserel, Expr *expr) { foreign_glob_cxt glob_cxt; foreign_loc_cxt loc_cxt; /* * Check that the expression consists of nodes that are safe to execute * remotely. */ glob_cxt.root = root; glob_cxt.foreignrel = baserel; loc_cxt.collation = InvalidOid; loc_cxt.state = FDW_COLLATE_NONE; if (!foreign_expr_walker((Node *) expr, &glob_cxt, &loc_cxt)) return false; /* * If the expression has a valid collation that does not arise from a * foreign var, the expression can not be sent over. */ if (loc_cxt.state == FDW_COLLATE_UNSAFE) return false; /* * An expression which includes any mutable functions can't be sent over * because its result is not stable. For example, sending now() remote * side could cause confusion from clock offsets. Future versions might * be able to make this choice with more granularity. (We check this last * because it requires a lot of expensive catalog lookups.) */ if (contain_mutable_functions((Node *) expr)) return false; /* OK to evaluate on the remote server */ return true; } /* * Check if expression is safe to execute remotely, and return true if so. * * In addition, *outer_cxt is updated with collation information. * * We must check that the expression contains only node types we can deparse, * that all types/functions/operators are safe to send (they are "shippable"), * and that all collations used in the expression derive from Vars of the * foreign table. Because of the latter, the logic is pretty close to * assign_collations_walker() in parse_collate.c, though we can assume here * that the given expression is valid. Note function mutability is not * currently considered here. */ static bool foreign_expr_walker(Node *node, foreign_glob_cxt *glob_cxt, foreign_loc_cxt *outer_cxt) { bool check_type = true; TdsFdwRelationInfo *fpinfo; foreign_loc_cxt inner_cxt; Oid collation; FDWCollateState state; /* Need do nothing for empty subexpressions */ if (node == NULL) return true; /* May need server info from baserel's fdw_private struct */ fpinfo = (TdsFdwRelationInfo *) (glob_cxt->foreignrel->fdw_private); /* Set up inner_cxt for possible recursion to child nodes */ inner_cxt.collation = InvalidOid; inner_cxt.state = FDW_COLLATE_NONE; switch (nodeTag(node)) { /* deleted cases from postgres_fdw for T_ArrayRef, T_FuncExpr, T_ScalarArrayOpExpr, T_ArrayExpr, which should never be pushed down to Sybase / MS SQL Server from PostgreSQL */ case T_Var: { Var *var = (Var *) node; /* * If the Var is from the foreign table, we consider its * collation (if any) safe to use. If it is from another * table, we treat its collation the same way as we would a * Param's collation, ie it's not safe for it to have a * non-default collation. */ if (var->varno == glob_cxt->foreignrel->relid && var->varlevelsup == 0) { /* Var belongs to foreign table */ /* * System columns other than ctid should not be sent to * the remote, since we don't make any effort to ensure * that local and remote values match (tableoid, in * particular, almost certainly doesn't match). */ if (var->varattno < 0 && var->varattno != SelfItemPointerAttributeNumber) return false; /* Else check the collation */ collation = var->varcollid; state = OidIsValid(collation) ? FDW_COLLATE_SAFE : FDW_COLLATE_NONE; } else { /* Var belongs to some other table */ collation = var->varcollid; if (collation == InvalidOid || collation == DEFAULT_COLLATION_OID) { /* * It's noncollatable, or it's safe to combine with a * collatable foreign Var, so set state to NONE. */ state = FDW_COLLATE_NONE; } else { /* * Do not fail right away, since the Var might appear * in a collation-insensitive context. */ state = FDW_COLLATE_UNSAFE; } } } break; case T_Const: { Const *c = (Const *) node; /* * If the constant has nondefault collation, either it's of a * non-builtin type, or it reflects folding of a CollateExpr. * It's unsafe to send to the remote unless it's used in a * non-collation-sensitive context. */ collation = c->constcollid; if (collation == InvalidOid || collation == DEFAULT_COLLATION_OID) state = FDW_COLLATE_NONE; else state = FDW_COLLATE_UNSAFE; } break; case T_Param: { Param *p = (Param *) node; /* * Collation rule is same as for Consts and non-foreign Vars. */ collation = p->paramcollid; if (collation == InvalidOid || collation == DEFAULT_COLLATION_OID) state = FDW_COLLATE_NONE; else state = FDW_COLLATE_UNSAFE; } break; case T_OpExpr: case T_DistinctExpr: /* struct-equivalent to OpExpr */ { OpExpr *oe = (OpExpr *) node; /* * Similarly, only shippable operators can be sent to remote. * (If the operator is shippable, we assume its underlying * function is too.) */ if (!is_shippable(oe->opno, OperatorRelationId, fpinfo)) return false; /* * Recurse to input subexpressions. */ if (!foreign_expr_walker((Node *) oe->args, glob_cxt, &inner_cxt)) return false; /* * If operator's input collation is not derived from a foreign * Var, it can't be sent to remote. */ if (oe->inputcollid == InvalidOid) /* OK, inputs are all noncollatable */ ; else if (inner_cxt.state != FDW_COLLATE_SAFE || oe->inputcollid != inner_cxt.collation) return false; /* Result-collation handling is same as for functions */ collation = oe->opcollid; if (collation == InvalidOid) state = FDW_COLLATE_NONE; else if (inner_cxt.state == FDW_COLLATE_SAFE && collation == inner_cxt.collation) state = FDW_COLLATE_SAFE; else if (collation == DEFAULT_COLLATION_OID) state = FDW_COLLATE_NONE; else state = FDW_COLLATE_UNSAFE; } break; case T_RelabelType: { RelabelType *r = (RelabelType *) node; /* * Recurse to input subexpression. */ if (!foreign_expr_walker((Node *) r->arg, glob_cxt, &inner_cxt)) return false; /* * RelabelType must not introduce a collation not derived from * an input foreign Var (same logic as for a real function). */ collation = r->resultcollid; if (collation == InvalidOid) state = FDW_COLLATE_NONE; else if (inner_cxt.state == FDW_COLLATE_SAFE && collation == inner_cxt.collation) state = FDW_COLLATE_SAFE; else if (collation == DEFAULT_COLLATION_OID) state = FDW_COLLATE_NONE; else state = FDW_COLLATE_UNSAFE; } break; case T_BoolExpr: { BoolExpr *b = (BoolExpr *) node; /* * Recurse to input subexpressions. */ if (!foreign_expr_walker((Node *) b->args, glob_cxt, &inner_cxt)) return false; /* Output is always boolean and so noncollatable. */ collation = InvalidOid; state = FDW_COLLATE_NONE; } break; case T_NullTest: { NullTest *nt = (NullTest *) node; /* * Recurse to input subexpressions. */ if (!foreign_expr_walker((Node *) nt->arg, glob_cxt, &inner_cxt)) return false; /* Output is always boolean and so noncollatable. */ collation = InvalidOid; state = FDW_COLLATE_NONE; } break; case T_List: { List *l = (List *) node; ListCell *lc; /* * Recurse to component subexpressions. */ foreach(lc, l) { if (!foreign_expr_walker((Node *) lfirst(lc), glob_cxt, &inner_cxt)) return false; } /* * When processing a list, collation state just bubbles up * from the list elements. */ collation = inner_cxt.collation; state = inner_cxt.state; /* Don't apply exprType() to the list. */ check_type = false; } break; default: /* * If it's anything else, assume it's unsafe. This list can be * expanded later, but don't forget to add deparse support below. */ return false; } /* * If result type of given expression is not shippable, it can't be sent * to remote because it might have incompatible semantics on remote side. */ if (check_type && !is_shippable(exprType(node), TypeRelationId, fpinfo)) return false; /* * Now, merge my collation information into my parent's state. */ if (state > outer_cxt->state) { /* Override previous parent state */ outer_cxt->collation = collation; outer_cxt->state = state; } else if (state == outer_cxt->state) { /* Merge, or detect error if there's a collation conflict */ switch (state) { case FDW_COLLATE_NONE: /* Nothing + nothing is still nothing */ break; case FDW_COLLATE_SAFE: if (collation != outer_cxt->collation) { /* * Non-default collation always beats default. */ if (outer_cxt->collation == DEFAULT_COLLATION_OID) { /* Override previous parent state */ outer_cxt->collation = collation; } else if (collation != DEFAULT_COLLATION_OID) { /* * Conflict; show state as indeterminate. We don't * want to "return false" right away, since parent * node might not care about collation. */ outer_cxt->state = FDW_COLLATE_UNSAFE; } } break; case FDW_COLLATE_UNSAFE: /* We're still conflicted ... */ break; } } /* It looks OK */ return true; } #if (PG_VERSION_NUM >= 90600) bool is_builtin(Oid objectId) { return (objectId < FirstBootstrapObjectId); } #endif /* * Convert type OID + typmod info into a type name we can ship to the remote * server. Someplace else had better have verified that this type name is * expected to be known on the remote end. * * This is almost just format_type_with_typemod(), except that if left to its * own devices, that function will make schema-qualification decisions based * on the local search_path, which is wrong. We must schema-qualify all * type names that are not in pg_catalog. We assume here that built-in types * are all in pg_catalog and need not be qualified; otherwise, qualify. */ static char * deparse_type_name(Oid type_oid, int32 typemod) { #if (PG_VERSION_NUM >= 90600) if (is_builtin(type_oid)) return format_type_with_typemod(type_oid, typemod); else return format_type_with_typemod_qualified(type_oid, typemod); #else return format_type_with_typemod(type_oid, typemod); #endif } /* * Construct a simple SELECT statement that retrieves desired columns * of the specified foreign table, and append it to "buf". The output * contains just "SELECT ... FROM tablename". * * We also create an integer List of the columns being retrieved, which is * returned to *retrieved_attrs. */ void deparseSelectSql(StringInfo buf, PlannerInfo *root, RelOptInfo *baserel, Bitmapset *attrs_used, List **retrieved_attrs, TdsFdwOptionSet* option_set) { RangeTblEntry *rte = planner_rt_fetch(baserel->relid, root); Relation rel; /* * Core code already has some lock on each rel being planned, so we can * use NoLock here. */ rel = heap_open(rte->relid, NoLock); /* * Construct SELECT list */ appendStringInfoString(buf, "SELECT "); deparseTargetList(buf, root, baserel->relid, rel, attrs_used, retrieved_attrs, option_set); /* * Construct FROM clause */ appendStringInfoString(buf, " FROM "); deparseRelation(buf, rel); heap_close(rel, NoLock); } /* * Emit a target list that retrieves the columns specified in attrs_used. * This is used for both SELECT and RETURNING targetlists. * * The tlist text is appended to buf, and we also create an integer List * of the columns being retrieved, which is returned to *retrieved_attrs. */ static void deparseTargetList(StringInfo buf, PlannerInfo *root, Index rtindex, Relation rel, Bitmapset *attrs_used, List **retrieved_attrs, TdsFdwOptionSet* option_set) { TupleDesc tupdesc = RelationGetDescr(rel); bool have_wholerow; bool first; int i; *retrieved_attrs = NIL; /* If we are not matching remote and local column names, then we can't push down attributes */ if (!option_set->match_column_names) { appendStringInfoString(buf, "* "); return; } /* If there's a whole-row reference, we'll need all the columns. */ have_wholerow = bms_is_member(0 - FirstLowInvalidHeapAttributeNumber, attrs_used); first = true; for (i = 1; i <= tupdesc->natts; i++) { Form_pg_attribute attr = tupdesc->attrs[i - 1]; /* Ignore dropped attributes. */ if (attr->attisdropped) continue; if (have_wholerow || bms_is_member(i - FirstLowInvalidHeapAttributeNumber, attrs_used)) { if (!first) appendStringInfoString(buf, ", "); first = false; deparseColumnRef(buf, rtindex, i, root); *retrieved_attrs = lappend_int(*retrieved_attrs, i); } } /* * Add ctid if needed. We currently don't support retrieving any other * system columns. */ if (bms_is_member(SelfItemPointerAttributeNumber - FirstLowInvalidHeapAttributeNumber, attrs_used)) { if (!first) appendStringInfoString(buf, ", "); first = false; appendStringInfoString(buf, "ctid"); *retrieved_attrs = lappend_int(*retrieved_attrs, SelfItemPointerAttributeNumber); } /* Don't generate bad syntax if no undropped columns */ if (first) appendStringInfoString(buf, "NULL"); } /* * Deparse WHERE clauses in given list of RestrictInfos and append them to buf. * * baserel is the foreign table we're planning for. * * If no WHERE clause already exists in the buffer, is_first should be true. * * If params is not NULL, it receives a list of Params and other-relation Vars * used in the clauses; these values must be transmitted to the remote server * as parameter values. * * If params is NULL, we're generating the query for EXPLAIN purposes, * so Params and other-relation Vars should be replaced by dummy values. */ void appendWhereClause(StringInfo buf, PlannerInfo *root, RelOptInfo *baserel, List *exprs, bool is_first, List **params) { deparse_expr_cxt context; /*int nestlevel;*/ ListCell *lc; if (params) *params = NIL; /* initialize result list to empty */ /* Set up context struct for recursion */ context.root = root; context.foreignrel = baserel; context.buf = buf; context.params_list = params; /* Make sure any constants in the exprs are printed portably */ /*nestlevel = set_transmission_modes();*/ foreach(lc, exprs) { RestrictInfo *ri = (RestrictInfo *) lfirst(lc); /* Connect expressions with "AND" and parenthesize each condition. */ if (is_first) appendStringInfoString(buf, " WHERE "); else appendStringInfoString(buf, " AND "); appendStringInfoChar(buf, '('); deparseExpr(ri->clause, &context); appendStringInfoChar(buf, ')'); is_first = false; } /*reset_transmission_modes(nestlevel);*/ } /* * deparse remote INSERT statement * * The statement text is appended to buf, and we also create an integer List * of the columns being retrieved by RETURNING (if any), which is returned * to *retrieved_attrs. */ void deparseInsertSql(StringInfo buf, PlannerInfo *root, Index rtindex, Relation rel, List *targetAttrs, bool doNothing, List *returningList, List **retrieved_attrs, TdsFdwOptionSet* option_set) { AttrNumber pindex; bool first; ListCell *lc; appendStringInfoString(buf, "INSERT INTO "); deparseRelation(buf, rel); if (targetAttrs) { appendStringInfoChar(buf, '('); first = true; foreach(lc, targetAttrs) { int attnum = lfirst_int(lc); if (!first) appendStringInfoString(buf, ", "); first = false; deparseColumnRef(buf, rtindex, attnum, root); } appendStringInfoString(buf, ") VALUES ("); pindex = 1; first = true; foreach(lc, targetAttrs) { if (!first) appendStringInfoString(buf, ", "); first = false; appendStringInfo(buf, "$%d", pindex); pindex++; } appendStringInfoChar(buf, ')'); } else appendStringInfoString(buf, " DEFAULT VALUES"); if (doNothing) appendStringInfoString(buf, " ON CONFLICT DO NOTHING"); deparseReturningList(buf, root, rtindex, rel, rel->trigdesc && rel->trigdesc->trig_insert_after_row, returningList, retrieved_attrs, option_set); } /* * deparse remote UPDATE statement * * The statement text is appended to buf, and we also create an integer List * of the columns being retrieved by RETURNING (if any), which is returned * to *retrieved_attrs. */ void deparseUpdateSql(StringInfo buf, PlannerInfo *root, Index rtindex, Relation rel, List *targetAttrs, List *returningList, List **retrieved_attrs, TdsFdwOptionSet* option_set) { AttrNumber pindex; bool first; ListCell *lc; appendStringInfoString(buf, "UPDATE "); deparseRelation(buf, rel); appendStringInfoString(buf, " SET "); pindex = 2; /* ctid is always the first param */ first = true; foreach(lc, targetAttrs) { int attnum = lfirst_int(lc); if (!first) appendStringInfoString(buf, ", "); first = false; deparseColumnRef(buf, rtindex, attnum, root); appendStringInfo(buf, " = $%d", pindex); pindex++; } appendStringInfoString(buf, " WHERE ctid = $1"); deparseReturningList(buf, root, rtindex, rel, rel->trigdesc && rel->trigdesc->trig_update_after_row, returningList, retrieved_attrs, option_set); } /* * deparse remote DELETE statement * * The statement text is appended to buf, and we also create an integer List * of the columns being retrieved by RETURNING (if any), which is returned * to *retrieved_attrs. */ void deparseDeleteSql(StringInfo buf, PlannerInfo *root, Index rtindex, Relation rel, List *returningList, List **retrieved_attrs, TdsFdwOptionSet* option_set) { appendStringInfoString(buf, "DELETE FROM "); deparseRelation(buf, rel); appendStringInfoString(buf, " WHERE ctid = $1"); deparseReturningList(buf, root, rtindex, rel, rel->trigdesc && rel->trigdesc->trig_delete_after_row, returningList, retrieved_attrs, option_set); } /* * Add a RETURNING clause, if needed, to an INSERT/UPDATE/DELETE. */ static void deparseReturningList(StringInfo buf, PlannerInfo *root, Index rtindex, Relation rel, bool trig_after_row, List *returningList, List **retrieved_attrs, TdsFdwOptionSet* option_set) { Bitmapset *attrs_used = NULL; if (trig_after_row) { /* whole-row reference acquires all non-system columns */ attrs_used = bms_make_singleton(0 - FirstLowInvalidHeapAttributeNumber); } if (returningList != NIL) { /* * We need the attrs, non-system and system, mentioned in the local * query's RETURNING list. */ pull_varattnos((Node *) returningList, rtindex, &attrs_used); } if (attrs_used != NULL) { appendStringInfoString(buf, " RETURNING "); deparseTargetList(buf, root, rtindex, rel, attrs_used, retrieved_attrs, option_set); } else *retrieved_attrs = NIL; } /* * Construct SELECT statement to acquire size in blocks of given relation. * * Note: we use local definition of block size, not remote definition. * This is perhaps debatable. * * Note: pg_relation_size() exists in 8.1 and later. */ void deparseAnalyzeSizeSql(StringInfo buf, Relation rel) { StringInfoData relname; /* We'll need the remote relation name as a literal. */ initStringInfo(&relname); deparseRelation(&relname, rel); appendStringInfoString(buf, "SELECT pg_catalog.pg_relation_size("); deparseStringLiteral(buf, relname.data); appendStringInfo(buf, "::pg_catalog.regclass) / %d", BLCKSZ); } /* * Construct SELECT statement to acquire sample rows of given relation. * * SELECT command is appended to buf, and list of columns retrieved * is returned to *retrieved_attrs. */ void deparseAnalyzeSql(StringInfo buf, Relation rel, List **retrieved_attrs) { Oid relid = RelationGetRelid(rel); TupleDesc tupdesc = RelationGetDescr(rel); int i; char *colname; List *options; ListCell *lc; bool first = true; *retrieved_attrs = NIL; appendStringInfoString(buf, "SELECT "); for (i = 0; i < tupdesc->natts; i++) { /* Ignore dropped columns. */ if (tupdesc->attrs[i]->attisdropped) continue; if (!first) appendStringInfoString(buf, ", "); first = false; /* Use attribute name or column_name option. */ colname = NameStr(tupdesc->attrs[i]->attname); options = GetForeignColumnOptions(relid, i + 1); foreach(lc, options) { DefElem *def = (DefElem *) lfirst(lc); if (strcmp(def->defname, "column_name") == 0) { colname = defGetString(def); break; } } appendStringInfoString(buf, quote_identifier(colname)); *retrieved_attrs = lappend_int(*retrieved_attrs, i + 1); } /* Don't generate bad syntax for zero-column relation. */ if (first) appendStringInfoString(buf, "NULL"); /* * Construct FROM clause */ appendStringInfoString(buf, " FROM "); deparseRelation(buf, rel); } /* * Construct name to use for given column, and emit it into buf. * If it has a column_name FDW option, use that instead of attribute name. */ static void deparseColumnRef(StringInfo buf, int varno, int varattno, PlannerInfo *root) { RangeTblEntry *rte; char *colname = NULL; List *options; ListCell *lc; /* varno must not be any of OUTER_VAR, INNER_VAR and INDEX_VAR. */ Assert(!IS_SPECIAL_VARNO(varno)); /* Get RangeTblEntry from array in PlannerInfo. */ rte = planner_rt_fetch(varno, root); /* * If it's a column of a foreign table, and it has the column_name FDW * option, use that value. */ options = GetForeignColumnOptions(rte->relid, varattno); foreach(lc, options) { DefElem *def = (DefElem *) lfirst(lc); if (strcmp(def->defname, "column_name") == 0) { colname = defGetString(def); break; } } /* * If it's a column of a regular table or it doesn't have column_name FDW * option, use attribute name. */ if (colname == NULL) colname = get_relid_attribute_name(rte->relid, varattno); appendStringInfoString(buf, quote_identifier(colname)); } /* * Append remote name of specified foreign table to buf. * Use value of table_name FDW option (if any) instead of relation's name. * Similarly, schema_name FDW option overrides schema name. */ static void deparseRelation(StringInfo buf, Relation rel) { ForeignTable *table; const char *nspname = NULL; const char *relname = NULL; ListCell *lc; /* obtain additional catalog information. */ table = GetForeignTable(RelationGetRelid(rel)); /* * Use value of FDW options if any, instead of the name of object itself. */ foreach(lc, table->options) { DefElem *def = (DefElem *) lfirst(lc); if (strcmp(def->defname, "schema_name") == 0) nspname = defGetString(def); else if (strcmp(def->defname, "table_name") == 0) relname = defGetString(def); else if (strcmp(def->defname, "table") == 0) relname = defGetString(def); } /* * Note: we could skip printing the schema name if it's pg_catalog, but * that doesn't seem worth the trouble. */ /*if (nspname == NULL) nspname = get_namespace_name(RelationGetNamespace(rel)); if (relname == NULL) relname = RelationGetRelationName(rel);*/ if (nspname == NULL) appendStringInfo(buf, "%s", relname); else appendStringInfo(buf, "%s.%s", quote_identifier(nspname), quote_identifier(relname)); } /* * Append a SQL string literal representing "val" to buf. */ void deparseStringLiteral(StringInfo buf, const char *val) { const char *valptr; /* * Rather than making assumptions about the remote server's value of * standard_conforming_strings, always use E'foo' syntax if there are any * backslashes. This will fail on remote servers before 8.1, but those * are long out of support. */ if (strchr(val, '\\') != NULL) appendStringInfoChar(buf, ESCAPE_STRING_SYNTAX); appendStringInfoChar(buf, '\''); for (valptr = val; *valptr; valptr++) { char ch = *valptr; if (SQL_STR_DOUBLE(ch, true)) appendStringInfoChar(buf, ch); appendStringInfoChar(buf, ch); } appendStringInfoChar(buf, '\''); } /* * Deparse given expression into context->buf. * * This function must support all the same node types that foreign_expr_walker * accepts. * * Note: unlike ruleutils.c, we just use a simple hard-wired parenthesization * scheme: anything more complex than a Var, Const, function call or cast * should be self-parenthesized. */ static void deparseExpr(Expr *node, deparse_expr_cxt *context) { if (node == NULL) return; switch (nodeTag(node)) { case T_Var: deparseVar((Var *) node, context); break; case T_Const: deparseConst((Const *) node, context); break; case T_Param: deparseParam((Param *) node, context); break; case T_ArrayRef: deparseArrayRef((ArrayRef *) node, context); break; case T_FuncExpr: deparseFuncExpr((FuncExpr *) node, context); break; case T_OpExpr: deparseOpExpr((OpExpr *) node, context); break; case T_DistinctExpr: deparseDistinctExpr((DistinctExpr *) node, context); break; case T_ScalarArrayOpExpr: deparseScalarArrayOpExpr((ScalarArrayOpExpr *) node, context); break; case T_RelabelType: deparseRelabelType((RelabelType *) node, context); break; case T_BoolExpr: deparseBoolExpr((BoolExpr *) node, context); break; case T_NullTest: deparseNullTest((NullTest *) node, context); break; case T_ArrayExpr: deparseArrayExpr((ArrayExpr *) node, context); break; default: elog(ERROR, "unsupported expression type for deparse: %d", (int) nodeTag(node)); break; } } /* * Deparse given Var node into context->buf. * * If the Var belongs to the foreign relation, just print its remote name. * Otherwise, it's effectively a Param (and will in fact be a Param at * run time). Handle it the same way we handle plain Params --- see * deparseParam for comments. */ static void deparseVar(Var *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; if (node->varno == context->foreignrel->relid && node->varlevelsup == 0) { /* Var belongs to foreign table */ deparseColumnRef(buf, node->varno, node->varattno, context->root); } else { /* Treat like a Param */ if (context->params_list) { int pindex = 0; ListCell *lc; /* find its index in params_list */ foreach(lc, *context->params_list) { pindex++; if (equal(node, (Node *) lfirst(lc))) break; } if (lc == NULL) { /* not in list, so add it */ pindex++; *context->params_list = lappend(*context->params_list, node); } printRemoteParam(pindex, node->vartype, node->vartypmod, context); } else { printRemotePlaceholder(node->vartype, node->vartypmod, context); } } } /* * Deparse given constant value into context->buf. * * This function has to be kept in sync with ruleutils.c's get_const_expr. */ static void deparseConst(Const *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; Oid typoutput; bool typIsVarlena; char *extval; bool isfloat = false; bool needlabel; if (node->constisnull) { appendStringInfoString(buf, "NULL"); appendStringInfo(buf, "::%s", deparse_type_name(node->consttype, node->consttypmod)); return; } getTypeOutputInfo(node->consttype, &typoutput, &typIsVarlena); extval = OidOutputFunctionCall(typoutput, node->constvalue); switch (node->consttype) { case INT2OID: case INT4OID: case INT8OID: case OIDOID: case FLOAT4OID: case FLOAT8OID: case NUMERICOID: { /* * No need to quote unless it's a special value such as 'NaN'. * See comments in get_const_expr(). */ if (strspn(extval, "0123456789+-eE.") == strlen(extval)) { if (extval[0] == '+' || extval[0] == '-') appendStringInfo(buf, "(%s)", extval); else appendStringInfoString(buf, extval); if (strcspn(extval, "eE.") != strlen(extval)) isfloat = true; /* it looks like a float */ } else appendStringInfo(buf, "'%s'", extval); } break; case BITOID: case VARBITOID: appendStringInfo(buf, "B'%s'", extval); break; case BOOLOID: if (strcmp(extval, "t") == 0) appendStringInfoString(buf, "true"); else appendStringInfoString(buf, "false"); break; default: deparseStringLiteral(buf, extval); break; } /* * Append ::typename unless the constant will be implicitly typed as the * right type when it is read in. * * XXX this code has to be kept in sync with the behavior of the parser, * especially make_const. */ switch (node->consttype) { case BOOLOID: case INT4OID: case UNKNOWNOID: needlabel = false; break; case NUMERICOID: needlabel = !isfloat || (node->consttypmod >= 0); break; default: needlabel = true; break; } if (needlabel) appendStringInfo(buf, "::%s", deparse_type_name(node->consttype, node->consttypmod)); } /* * Deparse given Param node. * * If we're generating the query "for real", add the Param to * context->params_list if it's not already present, and then use its index * in that list as the remote parameter number. During EXPLAIN, there's * no need to identify a parameter number. */ static void deparseParam(Param *node, deparse_expr_cxt *context) { if (context->params_list) { int pindex = 0; ListCell *lc; /* find its index in params_list */ foreach(lc, *context->params_list) { pindex++; if (equal(node, (Node *) lfirst(lc))) break; } if (lc == NULL) { /* not in list, so add it */ pindex++; *context->params_list = lappend(*context->params_list, node); } printRemoteParam(pindex, node->paramtype, node->paramtypmod, context); } else { printRemotePlaceholder(node->paramtype, node->paramtypmod, context); } } /* * Deparse an array subscript expression. */ static void deparseArrayRef(ArrayRef *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; ListCell *lowlist_item; ListCell *uplist_item; /* Always parenthesize the expression. */ appendStringInfoChar(buf, '('); /* * Deparse referenced array expression first. If that expression includes * a cast, we have to parenthesize to prevent the array subscript from * being taken as typename decoration. We can avoid that in the typical * case of subscripting a Var, but otherwise do it. */ if (IsA(node->refexpr, Var)) deparseExpr(node->refexpr, context); else { appendStringInfoChar(buf, '('); deparseExpr(node->refexpr, context); appendStringInfoChar(buf, ')'); } /* Deparse subscript expressions. */ lowlist_item = list_head(node->reflowerindexpr); /* could be NULL */ foreach(uplist_item, node->refupperindexpr) { appendStringInfoChar(buf, '['); if (lowlist_item) { deparseExpr(lfirst(lowlist_item), context); appendStringInfoChar(buf, ':'); lowlist_item = lnext(lowlist_item); } deparseExpr(lfirst(uplist_item), context); appendStringInfoChar(buf, ']'); } appendStringInfoChar(buf, ')'); } /* * Deparse a function call. */ static void deparseFuncExpr(FuncExpr *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; HeapTuple proctup; Form_pg_proc procform; const char *proname; bool use_variadic = false; bool first; ListCell *arg; /* * If the function call came from an implicit coercion, then just show the * first argument. */ if (node->funcformat == COERCE_IMPLICIT_CAST) { deparseExpr((Expr *) linitial(node->args), context); return; } /* * If the function call came from a cast, then show the first argument * plus an explicit cast operation. */ if (node->funcformat == COERCE_EXPLICIT_CAST) { Oid rettype = node->funcresulttype; int32 coercedTypmod; /* Get the typmod if this is a length-coercion function */ (void) exprIsLengthCoercion((Node *) node, &coercedTypmod); deparseExpr((Expr *) linitial(node->args), context); appendStringInfo(buf, "::%s", deparse_type_name(rettype, coercedTypmod)); return; } /* * Normal function: display as proname(args). */ proctup = SearchSysCache1(PROCOID, ObjectIdGetDatum(node->funcid)); if (!HeapTupleIsValid(proctup)) elog(ERROR, "cache lookup failed for function %u", node->funcid); procform = (Form_pg_proc) GETSTRUCT(proctup); #if (PG_VERSION_NUM >= 90300) /* Check if need to print VARIADIC (cf. ruleutils.c) */ use_variadic = node->funcvariadic; #endif /* Print schema name only if it's not pg_catalog */ if (procform->pronamespace != PG_CATALOG_NAMESPACE) { const char *schemaname; schemaname = get_namespace_name(procform->pronamespace); appendStringInfo(buf, "%s.", quote_identifier(schemaname)); } /* Deparse the function name ... */ proname = NameStr(procform->proname); appendStringInfo(buf, "%s(", quote_identifier(proname)); /* ... and all the arguments */ first = true; foreach(arg, node->args) { if (!first) appendStringInfoString(buf, ", "); if (use_variadic && lnext(arg) == NULL) appendStringInfoString(buf, "VARIADIC "); deparseExpr((Expr *) lfirst(arg), context); first = false; } appendStringInfoChar(buf, ')'); ReleaseSysCache(proctup); } /* * Deparse given operator expression. To avoid problems around * priority of operations, we always parenthesize the arguments. */ static void deparseOpExpr(OpExpr *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; HeapTuple tuple; Form_pg_operator form; char oprkind; ListCell *arg; /* Retrieve information about the operator from system catalog. */ tuple = SearchSysCache1(OPEROID, ObjectIdGetDatum(node->opno)); if (!HeapTupleIsValid(tuple)) elog(ERROR, "cache lookup failed for operator %u", node->opno); form = (Form_pg_operator) GETSTRUCT(tuple); oprkind = form->oprkind; /* Sanity check. */ Assert((oprkind == 'r' && list_length(node->args) == 1) || (oprkind == 'l' && list_length(node->args) == 1) || (oprkind == 'b' && list_length(node->args) == 2)); /* Always parenthesize the expression. */ appendStringInfoChar(buf, '('); /* Deparse left operand. */ if (oprkind == 'r' || oprkind == 'b') { arg = list_head(node->args); deparseExpr(lfirst(arg), context); appendStringInfoChar(buf, ' '); } /* Deparse operator name. */ deparseOperatorName(buf, form); /* Deparse right operand. */ if (oprkind == 'l' || oprkind == 'b') { arg = list_tail(node->args); appendStringInfoChar(buf, ' '); deparseExpr(lfirst(arg), context); } appendStringInfoChar(buf, ')'); ReleaseSysCache(tuple); } /* * Print the name of an operator. */ static void deparseOperatorName(StringInfo buf, Form_pg_operator opform) { char *opname; /* opname is not a SQL identifier, so we should not quote it. */ opname = NameStr(opform->oprname); /* Print schema name only if it's not pg_catalog */ if (opform->oprnamespace != PG_CATALOG_NAMESPACE) { const char *opnspname; opnspname = get_namespace_name(opform->oprnamespace); /* Print fully qualified operator name. */ appendStringInfo(buf, "OPERATOR(%s.%s)", quote_identifier(opnspname), opname); } else { /* Just print operator name. */ appendStringInfoString(buf, opname); } } /* * Deparse IS DISTINCT FROM. */ static void deparseDistinctExpr(DistinctExpr *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; Assert(list_length(node->args) == 2); appendStringInfoChar(buf, '('); deparseExpr(linitial(node->args), context); appendStringInfoString(buf, " IS DISTINCT FROM "); deparseExpr(lsecond(node->args), context); appendStringInfoChar(buf, ')'); } /* * Deparse given ScalarArrayOpExpr expression. To avoid problems * around priority of operations, we always parenthesize the arguments. */ static void deparseScalarArrayOpExpr(ScalarArrayOpExpr *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; HeapTuple tuple; Form_pg_operator form; Expr *arg1; Expr *arg2; /* Retrieve information about the operator from system catalog. */ tuple = SearchSysCache1(OPEROID, ObjectIdGetDatum(node->opno)); if (!HeapTupleIsValid(tuple)) elog(ERROR, "cache lookup failed for operator %u", node->opno); form = (Form_pg_operator) GETSTRUCT(tuple); /* Sanity check. */ Assert(list_length(node->args) == 2); /* Always parenthesize the expression. */ appendStringInfoChar(buf, '('); /* Deparse left operand. */ arg1 = linitial(node->args); deparseExpr(arg1, context); appendStringInfoChar(buf, ' '); /* Deparse operator name plus decoration. */ deparseOperatorName(buf, form); appendStringInfo(buf, " %s (", node->useOr ? "ANY" : "ALL"); /* Deparse right operand. */ arg2 = lsecond(node->args); deparseExpr(arg2, context); appendStringInfoChar(buf, ')'); /* Always parenthesize the expression. */ appendStringInfoChar(buf, ')'); ReleaseSysCache(tuple); } /* * Deparse a RelabelType (binary-compatible cast) node. */ static void deparseRelabelType(RelabelType *node, deparse_expr_cxt *context) { deparseExpr(node->arg, context); if (node->relabelformat != COERCE_IMPLICIT_CAST) appendStringInfo(context->buf, "::%s", deparse_type_name(node->resulttype, node->resulttypmod)); } /* * Deparse a BoolExpr node. */ static void deparseBoolExpr(BoolExpr *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; const char *op = NULL; /* keep compiler quiet */ bool first; ListCell *lc; switch (node->boolop) { case AND_EXPR: op = "AND"; break; case OR_EXPR: op = "OR"; break; case NOT_EXPR: appendStringInfoString(buf, "(NOT "); deparseExpr(linitial(node->args), context); appendStringInfoChar(buf, ')'); return; } appendStringInfoChar(buf, '('); first = true; foreach(lc, node->args) { if (!first) appendStringInfo(buf, " %s ", op); deparseExpr((Expr *) lfirst(lc), context); first = false; } appendStringInfoChar(buf, ')'); } /* * Deparse IS [NOT] NULL expression. */ static void deparseNullTest(NullTest *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; appendStringInfoChar(buf, '('); deparseExpr(node->arg, context); if (node->nulltesttype == IS_NULL) appendStringInfoString(buf, " IS NULL)"); else appendStringInfoString(buf, " IS NOT NULL)"); } /* * Deparse ARRAY[...] construct. */ static void deparseArrayExpr(ArrayExpr *node, deparse_expr_cxt *context) { StringInfo buf = context->buf; bool first = true; ListCell *lc; appendStringInfoString(buf, "ARRAY["); foreach(lc, node->elements) { if (!first) appendStringInfoString(buf, ", "); deparseExpr(lfirst(lc), context); first = false; } appendStringInfoChar(buf, ']'); /* If the array is empty, we need an explicit cast to the array type. */ if (node->elements == NIL) appendStringInfo(buf, "::%s", deparse_type_name(node->array_typeid, -1)); } /* * Print the representation of a parameter to be sent to the remote side. * * Note: we always label the Param's type explicitly rather than relying on * transmitting a numeric type OID in PQexecParams(). This allows us to * avoid assuming that types have the same OIDs on the remote side as they * do locally --- they need only have the same names. */ static void printRemoteParam(int paramindex, Oid paramtype, int32 paramtypmod, deparse_expr_cxt *context) { StringInfo buf = context->buf; char *ptypename = deparse_type_name(paramtype, paramtypmod); appendStringInfo(buf, "$%d::%s", paramindex, ptypename); } /* * Print the representation of a placeholder for a parameter that will be * sent to the remote side at execution time. * * This is used when we're just trying to EXPLAIN the remote query. * We don't have the actual value of the runtime parameter yet, and we don't * want the remote planner to generate a plan that depends on such a value * anyway. Thus, we can't do something simple like "$1::paramtype". * Instead, we emit "((SELECT null::paramtype)::paramtype)". * In all extant versions of Postgres, the planner will see that as an unknown * constant value, which is what we want. This might need adjustment if we * ever make the planner flatten scalar subqueries. Note: the reason for the * apparently useless outer cast is to ensure that the representation as a * whole will be parsed as an a_expr and not a select_with_parens; the latter * would do the wrong thing in the context "x = ANY(...)". */ static void printRemotePlaceholder(Oid paramtype, int32 paramtypmod, deparse_expr_cxt *context) { StringInfo buf = context->buf; char *ptypename = deparse_type_name(paramtype, paramtypmod); appendStringInfo(buf, "((SELECT null::%s)::%s)", ptypename, ptypename); } /* * Deparse ORDER BY clause according to the given pathkeys for given base * relation. From given pathkeys expressions belonging entirely to the given * base relation are obtained and deparsed. */ void appendOrderByClause(StringInfo buf, PlannerInfo *root, RelOptInfo *baserel, List *pathkeys) { ListCell *lcell; deparse_expr_cxt context; /*int nestlevel;*/ char *delim = " "; /* Set up context struct for recursion */ context.root = root; context.foreignrel = baserel; context.buf = buf; context.params_list = NULL; /* Make sure any constants in the exprs are printed portably */ /*nestlevel = set_transmission_modes();*/ appendStringInfo(buf, " ORDER BY"); foreach(lcell, pathkeys) { PathKey *pathkey = lfirst(lcell); Expr *em_expr; em_expr = find_em_expr_for_rel(pathkey->pk_eclass, baserel); Assert(em_expr != NULL); appendStringInfoString(buf, delim); deparseExpr(em_expr, &context); if (pathkey->pk_strategy == BTLessStrategyNumber) appendStringInfoString(buf, " ASC"); else appendStringInfoString(buf, " DESC"); if (pathkey->pk_nulls_first) appendStringInfoString(buf, " NULLS FIRST"); delim = ", "; } /*reset_transmission_modes(nestlevel);*/ }