#include "access/htup_details.h"
#include "funcapi.h"
-#include "utils/builtins.h"
+#include "catalog/pg_type.h"
#include "miscadmin.h"
+#include "utils/array.h"
+#include "utils/builtins.h"
+#include "utils/rel.h"
/*
}
+/*
+ * text_to_bits
+ *
+ * Converts a c-string representation of bits into a bits8-array. This is
+ * the reverse operation of previous routine.
+ */
+static bits8 *
+text_to_bits(char *str, int len)
+{
+ bits8 *bits;
+ int off = 0;
+ char byte = 0;
+
+ bits = palloc(len + 1);
+
+ while (off < len)
+ {
+ if (off % 8 == 0)
+ byte = 0;
+
+ if ((str[off] == '0') || (str[off] == '1'))
+ byte = byte | ((str[off] - '0') << off % 8);
+ else
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("illegal character '%c' in t_bits string", str[off])));
+
+ if (off % 8 == 7)
+ bits[off / 8] = byte;
+
+ off++;
+ }
+
+ return bits;
+}
+
/*
* heap_page_items
*
HeapTuple resultTuple;
Datum result;
ItemId id;
- Datum values[13];
- bool nulls[13];
+ Datum values[14];
+ bool nulls[14];
uint16 lp_offset;
uint16 lp_flags;
uint16 lp_len;
lp_offset == MAXALIGN(lp_offset) &&
lp_offset + lp_len <= raw_page_size)
{
- HeapTupleHeader tuphdr;
- int bits_len;
+ HeapTupleHeader tuphdr;
+ bytea *tuple_data_bytea;
+ int tuple_data_len;
/* Extract information from the tuple header */
values[4] = UInt32GetDatum(HeapTupleHeaderGetRawXmin(tuphdr));
values[5] = UInt32GetDatum(HeapTupleHeaderGetRawXmax(tuphdr));
- values[6] = UInt32GetDatum(HeapTupleHeaderGetRawCommandId(tuphdr)); /* shared with xvac */
+ /* shared with xvac */
+ values[6] = UInt32GetDatum(HeapTupleHeaderGetRawCommandId(tuphdr));
values[7] = PointerGetDatum(&tuphdr->t_ctid);
values[8] = UInt32GetDatum(tuphdr->t_infomask2);
values[9] = UInt32GetDatum(tuphdr->t_infomask);
values[10] = UInt8GetDatum(tuphdr->t_hoff);
+ /* Copy raw tuple data into bytea attribute */
+ tuple_data_len = lp_len - tuphdr->t_hoff;
+ tuple_data_bytea = (bytea *) palloc(tuple_data_len + VARHDRSZ);
+ SET_VARSIZE(tuple_data_bytea, tuple_data_len + VARHDRSZ);
+ memcpy(VARDATA(tuple_data_bytea), (char *) tuphdr + tuphdr->t_hoff,
+ tuple_data_len);
+ values[13] = PointerGetDatum(tuple_data_bytea);
+
/*
* We already checked that the item is completely within the raw
* page passed to us, with the length given in the line pointer.
{
if (tuphdr->t_infomask & HEAP_HASNULL)
{
- bits_len = tuphdr->t_hoff -
- offsetof(HeapTupleHeaderData, t_bits);
+ int bits_len =
+ ((tuphdr->t_infomask2 & HEAP_NATTS_MASK) / 8 + 1) * 8;
values[11] = CStringGetTextDatum(
- bits_to_text(tuphdr->t_bits, bits_len * 8));
+ bits_to_text(tuphdr->t_bits, bits_len));
}
else
nulls[11] = true;
*/
int i;
- for (i = 4; i <= 12; i++)
+ for (i = 4; i <= 13; i++)
nulls[i] = true;
}
else
SRF_RETURN_DONE(fctx);
}
+
+/*
+ * tuple_data_split_internal
+ *
+ * Split raw tuple data taken directly from a page into an array of bytea
+ * elements. This routine does a lookup on NULL values and creates array
+ * elements accordindly. This is a reimplementation of nocachegetattr()
+ * in heaptuple.c simplified for educational purposes.
+ */
+static Datum
+tuple_data_split_internal(Oid relid, char *tupdata,
+ uint16 tupdata_len, uint16 t_infomask,
+ uint16 t_infomask2, bits8 *t_bits,
+ bool do_detoast)
+{
+ ArrayBuildState *raw_attrs;
+ int nattrs;
+ int i;
+ int off = 0;
+ Relation rel;
+ TupleDesc tupdesc;
+
+ /* Get tuple descriptor from relation OID */
+ rel = relation_open(relid, NoLock);
+ tupdesc = CreateTupleDescCopyConstr(rel->rd_att);
+ relation_close(rel, NoLock);
+
+ raw_attrs = initArrayResult(BYTEAOID, CurrentMemoryContext, false);
+ nattrs = tupdesc->natts;
+
+ if (nattrs < (t_infomask2 & HEAP_NATTS_MASK))
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("number of attributes in tuple header is greater than number of attributes in tuple descriptor")));
+
+ for (i = 0; i < nattrs; i++)
+ {
+ Form_pg_attribute attr;
+ bool is_null;
+ bytea *attr_data = NULL;
+
+ attr = tupdesc->attrs[i];
+ is_null = (t_infomask & HEAP_HASNULL) && att_isnull(i, t_bits);
+
+ /*
+ * Tuple header can specify less attributes than tuple descriptor
+ * as ALTER TABLE ADD COLUMN without DEFAULT keyword does not
+ * actually change tuples in pages, so attributes with numbers greater
+ * than (t_infomask2 & HEAP_NATTS_MASK) should be treated as NULL.
+ */
+ if (i >= (t_infomask2 & HEAP_NATTS_MASK))
+ is_null = true;
+
+ if (!is_null)
+ {
+ int len;
+
+ if (attr->attlen == -1)
+ {
+ off = att_align_pointer(off, tupdesc->attrs[i]->attalign, -1,
+ tupdata + off);
+ /*
+ * As VARSIZE_ANY throws an exception if it can't properly
+ * detect the type of external storage in macros VARTAG_SIZE,
+ * this check is repeated to have a nicer error handling.
+ */
+ if (VARATT_IS_EXTERNAL(tupdata + off) &&
+ !VARATT_IS_EXTERNAL_ONDISK(tupdata + off) &&
+ !VARATT_IS_EXTERNAL_INDIRECT(tupdata + off))
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("first byte of varlena attribute is incorrect for attribute %d", i)));
+
+ len = VARSIZE_ANY(tupdata + off);
+ }
+ else
+ {
+ off = att_align_nominal(off, tupdesc->attrs[i]->attalign);
+ len = attr->attlen;
+ }
+
+ if (tupdata_len < off + len)
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("unexpected end of tuple data")));
+
+ if (attr->attlen == -1 && do_detoast)
+ attr_data = DatumGetByteaPCopy(tupdata + off);
+ else
+ {
+ attr_data = (bytea *) palloc(len + VARHDRSZ);
+ SET_VARSIZE(attr_data, len + VARHDRSZ);
+ memcpy(VARDATA(attr_data), tupdata + off, len);
+ }
+
+ off = att_addlength_pointer(off, tupdesc->attrs[i]->attlen,
+ tupdata + off);
+ }
+
+ raw_attrs = accumArrayResult(raw_attrs, PointerGetDatum(attr_data),
+ is_null, BYTEAOID, CurrentMemoryContext);
+ if (attr_data)
+ pfree(attr_data);
+ }
+
+ if (tupdata_len != off)
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("end of tuple reached without looking at all its data")));
+
+ return makeArrayResult(raw_attrs, CurrentMemoryContext);
+}
+
+/*
+ * tuple_data_split
+ *
+ * Split raw tuple data taken directly from page into distinct elements
+ * taking into account null values.
+ */
+PG_FUNCTION_INFO_V1(tuple_data_split);
+
+Datum
+tuple_data_split(PG_FUNCTION_ARGS)
+{
+ Oid relid;
+ bytea *raw_data;
+ uint16 t_infomask;
+ uint16 t_infomask2;
+ char *t_bits_str;
+ bool do_detoast = false;
+ bits8 *t_bits = NULL;
+ Datum res;
+
+ relid = PG_GETARG_OID(0);
+ raw_data = PG_ARGISNULL(1) ? NULL : PG_GETARG_BYTEA_P(1);
+ t_infomask = PG_GETARG_INT16(2);
+ t_infomask2 = PG_GETARG_INT16(3);
+ t_bits_str = PG_ARGISNULL(4) ? NULL :
+ text_to_cstring(PG_GETARG_TEXT_PP(4));
+
+ if (PG_NARGS() >= 6)
+ do_detoast = PG_GETARG_BOOL(5);
+
+ if (!superuser())
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
+ errmsg("must be superuser to use raw page functions")));
+
+ if (!raw_data)
+ PG_RETURN_NULL();
+
+ /*
+ * Convert t_bits string back to the bits8 array as represented in the
+ * tuple header.
+ */
+ if (t_infomask & HEAP_HASNULL)
+ {
+ int bits_str_len;
+ int bits_len;
+
+ bits_len = (t_infomask2 & HEAP_NATTS_MASK) / 8 + 1;
+ if (!t_bits_str)
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("argument of t_bits is null, but it is expected to be null and %i character long",
+ bits_len * 8)));
+
+ bits_str_len = strlen(t_bits_str);
+ if ((bits_str_len % 8) != 0)
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("length of t_bits is not a multiple of eight")));
+
+ if (bits_len * 8 != bits_str_len)
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("unexpected length of t_bits %u, expected %i",
+ bits_str_len, bits_len * 8)));
+
+ /* do the conversion */
+ t_bits = text_to_bits(t_bits_str, bits_str_len);
+ }
+ else
+ {
+ if (t_bits_str)
+ ereport(ERROR,
+ (errcode(ERRCODE_DATA_CORRUPTED),
+ errmsg("t_bits string is expected to be NULL, but instead it is %lu bytes length",
+ strlen(t_bits_str))));
+ }
+
+ /* Split tuple data */
+ res = tuple_data_split_internal(relid, (char *) raw_data + VARHDRSZ,
+ VARSIZE(raw_data) - VARHDRSZ,
+ t_infomask, t_infomask2, t_bits,
+ do_detoast);
+
+ if (t_bits)
+ pfree(t_bits);
+
+ PG_RETURN_ARRAYTYPE_P(res);
+}
<listitem>
<para>
<function>heap_page_items</function> shows all line pointers on a heap
- page. For those line pointers that are in use, tuple headers are also
- shown. All tuples are shown, whether or not the tuples were visible to
- an MVCC snapshot at the time the raw page was copied.
+ page. For those line pointers that are in use, tuple headers as well
+ as tuple raw data are also shown. All tuples are shown, whether or not
+ the tuples were visible to an MVCC snapshot at the time the raw page
+ was copied.
</para>
<para>
A heap page image obtained with <function>get_raw_page</function> should
</listitem>
</varlistentry>
+ <varlistentry>
+ <term>
+ <function>tuple_data_split(rel_oid, t_data bytea, t_infomask integer, t_infomask2 integer, t_bits text [, do_detoast bool]) returns bytea[]</function>
+ <indexterm>
+ <primary>tuple_data_split</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ <function>tuple_data_split</function> splits tuple data into attributes
+ in the same way as backend internals.
+<screen>
+test=# SELECT tuple_data_split('pg_class'::regclass, t_data, t_infomask, t_infomask2, t_bits) FROM heap_page_items(get_raw_page('pg_class', 0));
+</screen>
+ This function should be called with the same arguments as the return
+ attributes of <function>heap_page_items</function>.
+ </para>
+ <para>
+ If <parameter>do_detoast</parameter> is <literal>true</literal>,
+ attribute that will be detoasted as needed. Default value is
+ <literal>false</literal>.
+ </para>
+ </listitem>
+ </varlistentry>
+
+ <varlistentry>
+ <term>
+ <function>heap_page_item_attrs(rel_oid, t_data bytea, [, do_detoast bool]) returns bytea[]</function>
+ <indexterm>
+ <primary>heap_page_item_attrs</primary>
+ </indexterm>
+ </term>
+ <listitem>
+ <para>
+ <function>heap_page_item_attrs</function> is equivalent to
+ <function>heap_page_items</function> except that it returns
+ tuple raw data as an array of attributes that can optionally
+ be detoasted by <parameter>do_detoast</parameter> which is
+ <literal>false</literal> by default.
+ </para>
+ <para>
+ A heap page image obtained with <function>get_raw_page</function> should
+ be passed as argument. For example:
+<screen>
+test=# SELECT * FROM heap_page_item_attrs(get_raw_page('pg_class', 0), 'pg_class'::regclass);
+</screen>
+ </para>
+ </listitem>
+ </varlistentry>
+
<varlistentry>
<term>
<function>bt_metap(relname text) returns record</function>
projects / postgresql / commitdiff