codekingpro/portable-devtools
115k
1<?xml version="1.0" encoding="UTF-8" standalone="no"?>2<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>64.2. Index Access Method Functions</title><link rel="stylesheet" type="text/css" href="stylesheet.css" /><link rev="made" href="pgsql-docs@lists.postgresql.org" /><meta name="generator" content="DocBook XSL Stylesheets Vsnapshot" /><link rel="prev" href="index-api.html" title="64.1. Basic API Structure for Indexes" /><link rel="next" href="index-scanning.html" title="64.3. Index Scanning" /></head><body id="docContent" class="container-fluid col-10"><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">64.2. Index Access Method Functions</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="index-api.html" title="64.1. Basic API Structure for Indexes">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="indexam.html" title="Chapter 64. Index Access Method Interface Definition">Up</a></td><th width="60%" align="center">Chapter 64. Index Access Method Interface Definition</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 16.3 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="index-scanning.html" title="64.3. Index Scanning">Next</a></td></tr></table><hr /></div><div class="sect1" id="INDEX-FUNCTIONS"><div class="titlepage"><div><div><h2 class="title" style="clear: both">64.2. Index Access Method Functions <a href="#INDEX-FUNCTIONS" class="id_link">#</a></h2></div></div></div><p>3 The index construction and maintenance functions that an index access4 method must provide in <code class="structname">IndexAmRoutine</code> are:5 </p><p>6</p><pre class="programlisting">7IndexBuildResult *8ambuild (Relation heapRelation,9 Relation indexRelation,10 IndexInfo *indexInfo);11</pre><p>12 Build a new index. The index relation has been physically created,13 but is empty. It must be filled in with whatever fixed data the14 access method requires, plus entries for all tuples already existing15 in the table. Ordinarily the <code class="function">ambuild</code> function will call16 <code class="function">table_index_build_scan()</code> to scan the table for existing tuples17 and compute the keys that need to be inserted into the index.18 The function must return a palloc'd struct containing statistics about19 the new index.20 </p><p>21</p><pre class="programlisting">22void23ambuildempty (Relation indexRelation);24</pre><p>25 Build an empty index, and write it to the initialization fork (<code class="symbol">INIT_FORKNUM</code>)26 of the given relation. This method is called only for unlogged indexes; the27 empty index written to the initialization fork will be copied over the main28 relation fork on each server restart.29 </p><p>30</p><pre class="programlisting">31bool32aminsert (Relation indexRelation,33 Datum *values,34 bool *isnull,35 ItemPointer heap_tid,36 Relation heapRelation,37 IndexUniqueCheck checkUnique,38 bool indexUnchanged,39 IndexInfo *indexInfo);40</pre><p>41 Insert a new tuple into an existing index. The <code class="literal">values</code> and42 <code class="literal">isnull</code> arrays give the key values to be indexed, and43 <code class="literal">heap_tid</code> is the TID to be indexed.44 If the access method supports unique indexes (its45 <code class="structfield">amcanunique</code> flag is true) then46 <code class="literal">checkUnique</code> indicates the type of uniqueness check to47 perform. This varies depending on whether the unique constraint is48 deferrable; see <a class="xref" href="index-unique-checks.html" title="64.5. Index Uniqueness Checks">Section 64.5</a> for details.49 Normally the access method only needs the <code class="literal">heapRelation</code>50 parameter when performing uniqueness checking (since then it will have to51 look into the heap to verify tuple liveness).52 </p><p>53 The <code class="literal">indexUnchanged</code> Boolean value gives a hint54 about the nature of the tuple to be indexed. When it is true,55 the tuple is a duplicate of some existing tuple in the index. The56 new tuple is a logically unchanged successor MVCC tuple version. This57 happens when an <code class="command">UPDATE</code> takes place that does not58 modify any columns covered by the index, but nevertheless requires a59 new version in the index. The index AM may use this hint to decide60 to apply bottom-up index deletion in parts of the index where many61 versions of the same logical row accumulate. Note that updating a non-key62 column or a column that only appears in a partial index predicate does not63 affect the value of <code class="literal">indexUnchanged</code>. The core code64 determines each tuple's <code class="literal">indexUnchanged</code> value using a low65 overhead approach that allows both false positives and false negatives.66 Index AMs must not treat <code class="literal">indexUnchanged</code> as an67 authoritative source of information about tuple visibility or versioning.68 </p><p>69 The function's Boolean result value is significant only when70 <code class="literal">checkUnique</code> is <code class="literal">UNIQUE_CHECK_PARTIAL</code>.71 In this case a true result means the new entry is known unique, whereas72 false means it might be non-unique (and a deferred uniqueness check must73 be scheduled). For other cases a constant false result is recommended.74 </p><p>75 Some indexes might not index all tuples. If the tuple is not to be76 indexed, <code class="function">aminsert</code> should just return without doing anything.77 </p><p>78 If the index AM wishes to cache data across successive index insertions79 within an SQL statement, it can allocate space80 in <code class="literal">indexInfo->ii_Context</code> and store a pointer to the81 data in <code class="literal">indexInfo->ii_AmCache</code> (which will be NULL82 initially).83 </p><p>84</p><pre class="programlisting">85IndexBulkDeleteResult *86ambulkdelete (IndexVacuumInfo *info,87 IndexBulkDeleteResult *stats,88 IndexBulkDeleteCallback callback,89 void *callback_state);90</pre><p>91 Delete tuple(s) from the index. This is a <span class="quote">“<span class="quote">bulk delete</span>”</span> operation92 that is intended to be implemented by scanning the whole index and checking93 each entry to see if it should be deleted.94 The passed-in <code class="literal">callback</code> function must be called, in the style95 <code class="literal">callback(<em class="replaceable"><code>TID</code></em>, callback_state) returns bool</code>,96 to determine whether any particular index entry, as identified by its97 referenced TID, is to be deleted. Must return either NULL or a palloc'd98 struct containing statistics about the effects of the deletion operation.99 It is OK to return NULL if no information needs to be passed on to100 <code class="function">amvacuumcleanup</code>.101 </p><p>102 Because of limited <code class="varname">maintenance_work_mem</code>,103 <code class="function">ambulkdelete</code> might need to be called more than once when many104 tuples are to be deleted. The <code class="literal">stats</code> argument is the result105 of the previous call for this index (it is NULL for the first call within a106 <code class="command">VACUUM</code> operation). This allows the AM to accumulate statistics107 across the whole operation. Typically, <code class="function">ambulkdelete</code> will108 modify and return the same struct if the passed <code class="literal">stats</code> is not109 null.110 </p><p>111</p><pre class="programlisting">112IndexBulkDeleteResult *113amvacuumcleanup (IndexVacuumInfo *info,114 IndexBulkDeleteResult *stats);115</pre><p>116 Clean up after a <code class="command">VACUUM</code> operation (zero or more117 <code class="function">ambulkdelete</code> calls). This does not have to do anything118 beyond returning index statistics, but it might perform bulk cleanup119 such as reclaiming empty index pages. <code class="literal">stats</code> is whatever the120 last <code class="function">ambulkdelete</code> call returned, or NULL if121 <code class="function">ambulkdelete</code> was not called because no tuples needed to be122 deleted. If the result is not NULL it must be a palloc'd struct.123 The statistics it contains will be used to update <code class="structname">pg_class</code>,124 and will be reported by <code class="command">VACUUM</code> if <code class="literal">VERBOSE</code> is given.125 It is OK to return NULL if the index was not changed at all during the126 <code class="command">VACUUM</code> operation, but otherwise correct stats should127 be returned.128 </p><p>129 <code class="function">amvacuumcleanup</code> will also be called at completion of an130 <code class="command">ANALYZE</code> operation. In this case <code class="literal">stats</code> is always131 NULL and any return value will be ignored. This case can be distinguished132 by checking <code class="literal">info->analyze_only</code>. It is recommended133 that the access method do nothing except post-insert cleanup in such a134 call, and that only in an autovacuum worker process.135 </p><p>136</p><pre class="programlisting">137bool138amcanreturn (Relation indexRelation, int attno);139</pre><p>140 Check whether the index can support <a class="link" href="indexes-index-only-scans.html" title="11.9. Index-Only Scans and Covering Indexes"><em class="firstterm">index-only scans</em></a> on141 the given column, by returning the column's original indexed value.142 The attribute number is 1-based, i.e., the first column's attno is 1.143 Returns true if supported, else false.144 This function should always return true for included columns145 (if those are supported), since there's little point in an included146 column that can't be retrieved.147 If the access method does not support index-only scans at all,148 the <code class="structfield">amcanreturn</code> field in its <code class="structname">IndexAmRoutine</code>149 struct can be set to NULL.150 </p><p>151</p><pre class="programlisting">152void153amcostestimate (PlannerInfo *root,154 IndexPath *path,155 double loop_count,156 Cost *indexStartupCost,157 Cost *indexTotalCost,158 Selectivity *indexSelectivity,159 double *indexCorrelation,160 double *indexPages);161</pre><p>162 Estimate the costs of an index scan. This function is described fully163 in <a class="xref" href="index-cost-estimation.html" title="64.6. Index Cost Estimation Functions">Section 64.6</a>, below.164 </p><p>165</p><pre class="programlisting">166bytea *167amoptions (ArrayType *reloptions,168 bool validate);169</pre><p>170 Parse and validate the reloptions array for an index. This is called only171 when a non-null reloptions array exists for the index.172 <em class="parameter"><code>reloptions</code></em> is a <code class="type">text</code> array containing entries of the173 form <em class="replaceable"><code>name</code></em><code class="literal">=</code><em class="replaceable"><code>value</code></em>.174 The function should construct a <code class="type">bytea</code> value, which will be copied175 into the <code class="structfield">rd_options</code> field of the index's relcache entry.176 The data contents of the <code class="type">bytea</code> value are open for the access177 method to define; most of the standard access methods use struct178 <code class="structname">StdRdOptions</code>.179 When <em class="parameter"><code>validate</code></em> is true, the function should report a suitable180 error message if any of the options are unrecognized or have invalid181 values; when <em class="parameter"><code>validate</code></em> is false, invalid entries should be182 silently ignored. (<em class="parameter"><code>validate</code></em> is false when loading options183 already stored in <code class="structname">pg_catalog</code>; an invalid entry could only184 be found if the access method has changed its rules for options, and in185 that case ignoring obsolete entries is appropriate.)186 It is OK to return NULL if default behavior is wanted.187 </p><p>188</p><pre class="programlisting">189bool190amproperty (Oid index_oid, int attno,191 IndexAMProperty prop, const char *propname,192 bool *res, bool *isnull);193</pre><p>194 The <code class="function">amproperty</code> method allows index access methods to override195 the default behavior of <code class="function">pg_index_column_has_property</code>196 and related functions.197 If the access method does not have any special behavior for index property198 inquiries, the <code class="structfield">amproperty</code> field in199 its <code class="structname">IndexAmRoutine</code> struct can be set to NULL.200 Otherwise, the <code class="function">amproperty</code> method will be called with201 <em class="parameter"><code>index_oid</code></em> and <em class="parameter"><code>attno</code></em> both zero for202 <code class="function">pg_indexam_has_property</code> calls,203 or with <em class="parameter"><code>index_oid</code></em> valid and <em class="parameter"><code>attno</code></em> zero for204 <code class="function">pg_index_has_property</code> calls,205 or with <em class="parameter"><code>index_oid</code></em> valid and <em class="parameter"><code>attno</code></em> greater than206 zero for <code class="function">pg_index_column_has_property</code> calls.207 <em class="parameter"><code>prop</code></em> is an enum value identifying the property being tested,208 while <em class="parameter"><code>propname</code></em> is the original property name string.209 If the core code does not recognize the property name210 then <em class="parameter"><code>prop</code></em> is <code class="literal">AMPROP_UNKNOWN</code>.211 Access methods can define custom property names by212 checking <em class="parameter"><code>propname</code></em> for a match (use <code class="function">pg_strcasecmp</code>213 to match, for consistency with the core code); for names known to the core214 code, it's better to inspect <em class="parameter"><code>prop</code></em>.215 If the <code class="structfield">amproperty</code> method returns <code class="literal">true</code> then216 it has determined the property test result: it must set <code class="literal">*res</code>217 to the Boolean value to return, or set <code class="literal">*isnull</code>218 to <code class="literal">true</code> to return a NULL. (Both of the referenced variables219 are initialized to <code class="literal">false</code> before the call.)220 If the <code class="structfield">amproperty</code> method returns <code class="literal">false</code> then221 the core code will proceed with its normal logic for determining the222 property test result.223 </p><p>224 Access methods that support ordering operators should225 implement <code class="literal">AMPROP_DISTANCE_ORDERABLE</code> property testing, as the226 core code does not know how to do that and will return NULL. It may227 also be advantageous to implement <code class="literal">AMPROP_RETURNABLE</code> testing,228 if that can be done more cheaply than by opening the index and calling229 <code class="function">amcanreturn</code>, which is the core code's default behavior.230 The default behavior should be satisfactory for all other standard231 properties.232 </p><p>233</p><pre class="programlisting">234char *235ambuildphasename (int64 phasenum);236</pre><p>237 Return the textual name of the given build phase number.238 The phase numbers are those reported during an index build via the239 <code class="function">pgstat_progress_update_param</code> interface.240 The phase names are then exposed in the241 <code class="structname">pg_stat_progress_create_index</code> view.242 </p><p>243</p><pre class="programlisting">244bool245amvalidate (Oid opclassoid);246</pre><p>247 Validate the catalog entries for the specified operator class, so far as248 the access method can reasonably do that. For example, this might include249 testing that all required support functions are provided.250 The <code class="function">amvalidate</code> function must return false if the opclass is251 invalid. Problems should be reported with <code class="function">ereport</code>252 messages, typically at <code class="literal">INFO</code> level.253 </p><p>254</p><pre class="programlisting">255void256amadjustmembers (Oid opfamilyoid,257 Oid opclassoid,258 List *operators,259 List *functions);260</pre><p>261 Validate proposed new operator and function members of an operator family,262 so far as the access method can reasonably do that, and set their263 dependency types if the default is not satisfactory. This is called264 during <code class="command">CREATE OPERATOR CLASS</code> and during265 <code class="command">ALTER OPERATOR FAMILY ADD</code>; in the latter266 case <em class="parameter"><code>opclassoid</code></em> is <code class="literal">InvalidOid</code>.267 The <code class="type">List</code> arguments are lists268 of <code class="structname">OpFamilyMember</code> structs, as defined269 in <code class="filename">amapi.h</code>.270 271 Tests done by this function will typically be a subset of those272 performed by <code class="function">amvalidate</code>,273 since <code class="function">amadjustmembers</code> cannot assume that it is274 seeing a complete set of members. For example, it would be reasonable275 to check the signature of a support function, but not to check whether276 all required support functions are provided. Any problems can be277 reported by throwing an error.278 279 The dependency-related fields of280 the <code class="structname">OpFamilyMember</code> structs are initialized by281 the core code to create hard dependencies on the opclass if this282 is <code class="command">CREATE OPERATOR CLASS</code>, or soft dependencies on the283 opfamily if this is <code class="command">ALTER OPERATOR FAMILY ADD</code>.284 <code class="function">amadjustmembers</code> can adjust these fields if some other285 behavior is more appropriate. For example, GIN, GiST, and SP-GiST286 always set operator members to have soft dependencies on the opfamily,287 since the connection between an operator and an opclass is relatively288 weak in these index types; so it is reasonable to allow operator members289 to be added and removed freely. Optional support functions are typically290 also given soft dependencies, so that they can be removed if necessary.291 </p><p>292 The purpose of an index, of course, is to support scans for tuples matching293 an indexable <code class="literal">WHERE</code> condition, often called a294 <em class="firstterm">qualifier</em> or <em class="firstterm">scan key</em>. The semantics of295 index scanning are described more fully in <a class="xref" href="index-scanning.html" title="64.3. Index Scanning">Section 64.3</a>,296 below. An index access method can support <span class="quote">“<span class="quote">plain</span>”</span> index scans,297 <span class="quote">“<span class="quote">bitmap</span>”</span> index scans, or both. The scan-related functions that an298 index access method must or may provide are:299 </p><p>300</p><pre class="programlisting">301IndexScanDesc302ambeginscan (Relation indexRelation,303 int nkeys,304 int norderbys);305</pre><p>306 Prepare for an index scan. The <code class="literal">nkeys</code> and <code class="literal">norderbys</code>307 parameters indicate the number of quals and ordering operators that will be308 used in the scan; these may be useful for space allocation purposes.309 Note that the actual values of the scan keys aren't provided yet.310 The result must be a palloc'd struct.311 For implementation reasons the index access method312 <span class="emphasis"><em>must</em></span> create this struct by calling313 <code class="function">RelationGetIndexScan()</code>. In most cases314 <code class="function">ambeginscan</code> does little beyond making that call and perhaps315 acquiring locks;316 the interesting parts of index-scan startup are in <code class="function">amrescan</code>.317 </p><p>318</p><pre class="programlisting">319void320amrescan (IndexScanDesc scan,321 ScanKey keys,322 int nkeys,323 ScanKey orderbys,324 int norderbys);325</pre><p>326 Start or restart an index scan, possibly with new scan keys. (To restart327 using previously-passed keys, NULL is passed for <code class="literal">keys</code> and/or328 <code class="literal">orderbys</code>.) Note that it is not allowed for329 the number of keys or order-by operators to be larger than330 what was passed to <code class="function">ambeginscan</code>. In practice the restart331 feature is used when a new outer tuple is selected by a nested-loop join332 and so a new key comparison value is needed, but the scan key structure333 remains the same.334 </p><p>335</p><pre class="programlisting">336bool337amgettuple (IndexScanDesc scan,338 ScanDirection direction);339</pre><p>340 Fetch the next tuple in the given scan, moving in the given341 direction (forward or backward in the index). Returns true if a tuple was342 obtained, false if no matching tuples remain. In the true case the tuple343 TID is stored into the <code class="literal">scan</code> structure. Note that344 <span class="quote">“<span class="quote">success</span>”</span> means only that the index contains an entry that matches345 the scan keys, not that the tuple necessarily still exists in the heap or346 will pass the caller's snapshot test. On success, <code class="function">amgettuple</code>347 must also set <code class="literal">scan->xs_recheck</code> to true or false.348 False means it is certain that the index entry matches the scan keys.349 True means this is not certain, and the conditions represented by the350 scan keys must be rechecked against the heap tuple after fetching it.351 This provision supports <span class="quote">“<span class="quote">lossy</span>”</span> index operators.352 Note that rechecking will extend only to the scan conditions; a partial353 index predicate (if any) is never rechecked by <code class="function">amgettuple</code>354 callers.355 </p><p>356 If the index supports <a class="link" href="indexes-index-only-scans.html" title="11.9. Index-Only Scans and Covering Indexes">index-only357 scans</a> (i.e., <code class="function">amcanreturn</code> returns true for any358 of its columns),359 then on success the AM must also check <code class="literal">scan->xs_want_itup</code>,360 and if that is true it must return the originally indexed data for the361 index entry. Columns for which <code class="function">amcanreturn</code> returns362 false can be returned as nulls.363 The data can be returned in the form of an364 <code class="structname">IndexTuple</code> pointer stored at <code class="literal">scan->xs_itup</code>,365 with tuple descriptor <code class="literal">scan->xs_itupdesc</code>; or in the form of366 a <code class="structname">HeapTuple</code> pointer stored at <code class="literal">scan->xs_hitup</code>,367 with tuple descriptor <code class="literal">scan->xs_hitupdesc</code>. (The latter368 format should be used when reconstructing data that might possibly not fit369 into an <code class="structname">IndexTuple</code>.) In either case,370 management of the data referenced by the pointer is the access method's371 responsibility. The data must remain good at least until the next372 <code class="function">amgettuple</code>, <code class="function">amrescan</code>, or <code class="function">amendscan</code>373 call for the scan.374 </p><p>375 The <code class="function">amgettuple</code> function need only be provided if the access376 method supports <span class="quote">“<span class="quote">plain</span>”</span> index scans. If it doesn't, the377 <code class="structfield">amgettuple</code> field in its <code class="structname">IndexAmRoutine</code>378 struct must be set to NULL.379 </p><p>380</p><pre class="programlisting">381int64382amgetbitmap (IndexScanDesc scan,383 TIDBitmap *tbm);384</pre><p>385 Fetch all tuples in the given scan and add them to the caller-supplied386 <code class="type">TIDBitmap</code> (that is, OR the set of tuple IDs into whatever set is already387 in the bitmap). The number of tuples fetched is returned (this might be388 just an approximate count, for instance some AMs do not detect duplicates).389 While inserting tuple IDs into the bitmap, <code class="function">amgetbitmap</code> can390 indicate that rechecking of the scan conditions is required for specific391 tuple IDs. This is analogous to the <code class="literal">xs_recheck</code> output parameter392 of <code class="function">amgettuple</code>. Note: in the current implementation, support393 for this feature is conflated with support for lossy storage of the bitmap394 itself, and therefore callers recheck both the scan conditions and the395 partial index predicate (if any) for recheckable tuples. That might not396 always be true, however.397 <code class="function">amgetbitmap</code> and398 <code class="function">amgettuple</code> cannot be used in the same index scan; there399 are other restrictions too when using <code class="function">amgetbitmap</code>, as explained400 in <a class="xref" href="index-scanning.html" title="64.3. Index Scanning">Section 64.3</a>.401 </p><p>402 The <code class="function">amgetbitmap</code> function need only be provided if the access403 method supports <span class="quote">“<span class="quote">bitmap</span>”</span> index scans. If it doesn't, the404 <code class="structfield">amgetbitmap</code> field in its <code class="structname">IndexAmRoutine</code>405 struct must be set to NULL.406 </p><p>407</p><pre class="programlisting">408void409amendscan (IndexScanDesc scan);410</pre><p>411 End a scan and release resources. The <code class="literal">scan</code> struct itself412 should not be freed, but any locks or pins taken internally by the413 access method must be released, as well as any other memory allocated414 by <code class="function">ambeginscan</code> and other scan-related functions.415 </p><p>416</p><pre class="programlisting">417void418ammarkpos (IndexScanDesc scan);419</pre><p>420 Mark current scan position. The access method need only support one421 remembered scan position per scan.422 </p><p>423 The <code class="function">ammarkpos</code> function need only be provided if the access424 method supports ordered scans. If it doesn't,425 the <code class="structfield">ammarkpos</code> field in its <code class="structname">IndexAmRoutine</code>426 struct may be set to NULL.427 </p><p>428</p><pre class="programlisting">429void430amrestrpos (IndexScanDesc scan);431</pre><p>432 Restore the scan to the most recently marked position.433 </p><p>434 The <code class="function">amrestrpos</code> function need only be provided if the access435 method supports ordered scans. If it doesn't,436 the <code class="structfield">amrestrpos</code> field in its <code class="structname">IndexAmRoutine</code>437 struct may be set to NULL.438 </p><p>439 In addition to supporting ordinary index scans, some types of index440 may wish to support <em class="firstterm">parallel index scans</em>, which allow441 multiple backends to cooperate in performing an index scan. The442 index access method should arrange things so that each cooperating443 process returns a subset of the tuples that would be performed by444 an ordinary, non-parallel index scan, but in such a way that the445 union of those subsets is equal to the set of tuples that would be446 returned by an ordinary, non-parallel index scan. Furthermore, while447 there need not be any global ordering of tuples returned by a parallel448 scan, the ordering of that subset of tuples returned within each449 cooperating backend must match the requested ordering. The following450 functions may be implemented to support parallel index scans:451 </p><p>452</p><pre class="programlisting">453Size454amestimateparallelscan (void);455</pre><p>456 Estimate and return the number of bytes of dynamic shared memory which457 the access method will be needed to perform a parallel scan. (This number458 is in addition to, not in lieu of, the amount of space needed for459 AM-independent data in <code class="structname">ParallelIndexScanDescData</code>.)460 </p><p>461 It is not necessary to implement this function for access methods which462 do not support parallel scans or for which the number of additional bytes463 of storage required is zero.464 </p><p>465</p><pre class="programlisting">466void467aminitparallelscan (void *target);468</pre><p>469 This function will be called to initialize dynamic shared memory at the470 beginning of a parallel scan. <em class="parameter"><code>target</code></em> will point to at least471 the number of bytes previously returned by472 <code class="function">amestimateparallelscan</code>, and this function may use that473 amount of space to store whatever data it wishes.474 </p><p>475 It is not necessary to implement this function for access methods which476 do not support parallel scans or in cases where the shared memory space477 required needs no initialization.478 </p><p>479</p><pre class="programlisting">480void481amparallelrescan (IndexScanDesc scan);482</pre><p>483 This function, if implemented, will be called when a parallel index scan484 must be restarted. It should reset any shared state set up by485 <code class="function">aminitparallelscan</code> such that the scan will be restarted from486 the beginning.487 </p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="index-api.html" title="64.1. Basic API Structure for Indexes">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="indexam.html" title="Chapter 64. Index Access Method Interface Definition">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="index-scanning.html" title="64.3. Index Scanning">Next</a></td></tr><tr><td width="40%" align="left" valign="top">64.1. Basic API Structure for Indexes </td><td width="20%" align="center"><a accesskey="h" href="index.html" title="PostgreSQL 16.3 Documentation">Home</a></td><td width="40%" align="right" valign="top"> 64.3. Index Scanning</td></tr></table></div></body></html>