postgresql/contrib/xml2/xpath.c

894 lines
21 KiB
C

/* Parser interface for DOM-based parser (libxml) rather than
stream-based SAX-type parser */
#include "postgres.h"
#include "fmgr.h"
#include "executor/spi.h"
#include "funcapi.h"
#include "miscadmin.h"
#include "lib/stringinfo.h"
/* libxml includes */
#include <libxml/xpath.h>
#include <libxml/tree.h>
#include <libxml/xmlmemory.h>
#include <libxml/xmlerror.h>
#include <libxml/parserInternals.h>
/* declarations */
static void *pgxml_palloc(size_t size);
static void *pgxml_repalloc(void *ptr, size_t size);
static void pgxml_pfree(void *ptr);
static char *pgxml_pstrdup(const char *string);
static void pgxml_errorHandler (void * ctxt, const char *msg, ...);
void elog_error(int level, char *explain, int force);
void pgxml_parser_init(void);
static xmlChar *pgxmlNodeSetToText(xmlNodeSetPtr nodeset,
xmlChar * toptagname, xmlChar * septagname,
xmlChar * plainsep);
text *pgxml_result_to_text(xmlXPathObjectPtr res, xmlChar *toptag,
xmlChar *septag, xmlChar *plainsep);
xmlChar *pgxml_texttoxmlchar(text *textstring);
static xmlXPathObjectPtr pgxml_xpath(text *document, xmlChar* xpath);
Datum pgxml_parse(PG_FUNCTION_ARGS);
Datum xpath_nodeset(PG_FUNCTION_ARGS);
Datum xpath_string(PG_FUNCTION_ARGS);
Datum xpath_number(PG_FUNCTION_ARGS);
Datum xpath_bool(PG_FUNCTION_ARGS);
Datum xpath_list(PG_FUNCTION_ARGS);
Datum xpath_table(PG_FUNCTION_ARGS);
/* Global variables */
char *errbuf; /* per line error buffer */
char *pgxml_errorMsg = NULL; /* overall error message */
/* Convenience macros */
#define GET_TEXT(cstrp) DatumGetTextP(DirectFunctionCall1(textin, CStringGetDatum(cstrp)))
#define GET_STR(textp) DatumGetCString(DirectFunctionCall1(textout, PointerGetDatum(textp)))
#define ERRBUF_SIZE 200
/* memory handling passthrough functions (e.g. palloc, pstrdup are
currently macros, and the others might become so...) */
static void *
pgxml_palloc(size_t size)
{
/* elog(DEBUG1,"Alloc %d in CMC %x",size,CurrentMemoryContext); */
return palloc(size);
}
static void *
pgxml_repalloc(void *ptr, size_t size)
{
/* elog(DEBUG1,"ReAlloc in CMC %x",CurrentMemoryContext);*/
return repalloc(ptr, size);
}
static void
pgxml_pfree(void *ptr)
{
/* elog(DEBUG1,"Free in CMC %x",CurrentMemoryContext); */
return pfree(ptr);
}
static char *
pgxml_pstrdup(const char *string)
{
return pstrdup(string);
}
/* The error handling function. This formats an error message and sets
* a flag - an ereport will be issued prior to return
*/
static void
pgxml_errorHandler (void * ctxt, const char *msg, ...)
{
va_list args;
va_start(args, msg);
vsnprintf(errbuf, ERRBUF_SIZE, msg, args);
va_end(args);
/* Now copy the argument across */
if (pgxml_errorMsg == NULL)
{
pgxml_errorMsg = pstrdup(errbuf);
}
else
{
int32 xsize = strlen(pgxml_errorMsg);
pgxml_errorMsg = repalloc(pgxml_errorMsg,
(size_t) (xsize + strlen(errbuf) + 1));
strncpy(&pgxml_errorMsg[xsize-1],errbuf,strlen(errbuf));
pgxml_errorMsg[xsize+strlen(errbuf)-1]='\0';
}
memset(errbuf,0,ERRBUF_SIZE);
}
/* This function reports the current message at the level specified */
void elog_error(int level, char *explain, int force)
{
if (force || (pgxml_errorMsg != NULL))
{
if (pgxml_errorMsg == NULL)
{
ereport(level,(errcode(ERRCODE_EXTERNAL_ROUTINE_EXCEPTION),
errmsg(explain)));
}
else
{
ereport(level,(errcode(ERRCODE_EXTERNAL_ROUTINE_EXCEPTION),
errmsg("%s:%s",explain,pgxml_errorMsg)));
pfree(pgxml_errorMsg);
}
}
}
void
pgxml_parser_init()
{
/*
* This code could also set parser settings from user-supplied info.
* Quite how these settings are made is another matter :)
*/
xmlMemSetup(pgxml_pfree, pgxml_palloc, pgxml_repalloc, pgxml_pstrdup);
xmlInitParser();
xmlSetGenericErrorFunc(NULL, pgxml_errorHandler);
xmlSubstituteEntitiesDefault(1);
xmlLoadExtDtdDefaultValue = 1;
pgxml_errorMsg = NULL;
errbuf = palloc(200);
memset(errbuf,0,200);
}
/* Returns true if document is well-formed */
PG_FUNCTION_INFO_V1(pgxml_parse);
Datum
pgxml_parse(PG_FUNCTION_ARGS)
{
/* called as pgxml_parse(document) */
xmlDocPtr doctree;
text *t = PG_GETARG_TEXT_P(0); /* document buffer */
int32 docsize = VARSIZE(t) - VARHDRSZ;
pgxml_parser_init();
doctree = xmlParseMemory((char *) VARDATA(t), docsize);
if (doctree == NULL)
{
xmlCleanupParser();
PG_RETURN_BOOL(false); /* i.e. not well-formed */
}
xmlCleanupParser();
xmlFreeDoc(doctree);
PG_RETURN_BOOL(true);
}
static xmlChar
*
pgxmlNodeSetToText(xmlNodeSetPtr nodeset,
xmlChar * toptagname,
xmlChar * septagname,
xmlChar * plainsep)
{
/* Function translates a nodeset into a text representation */
/*
* iterates over each node in the set and calls xmlNodeDump to write
* it to an xmlBuffer -from which an xmlChar * string is returned.
*/
/* each representation is surrounded by <tagname> ... </tagname> */
/* plainsep is an ordinary (not tag) seperator - if used, then
* nodes are cast to string as output method */
xmlBufferPtr buf;
xmlChar *result;
int i;
buf = xmlBufferCreate();
if ((toptagname != NULL) && (xmlStrlen(toptagname) > 0))
{
xmlBufferWriteChar(buf, "<");
xmlBufferWriteCHAR(buf, toptagname);
xmlBufferWriteChar(buf, ">");
}
if (nodeset != NULL)
{
for (i = 0; i < nodeset->nodeNr; i++)
{
if (plainsep != NULL) {
xmlBufferWriteCHAR(buf,
xmlXPathCastNodeToString(nodeset->nodeTab[i]));
/* If this isn't the last entry, write the plain sep. */
if (i < (nodeset->nodeNr)-1) {
xmlBufferWriteChar(buf, plainsep);
}
} else {
if ((septagname != NULL) && (xmlStrlen(septagname) > 0))
{
xmlBufferWriteChar(buf, "<");
xmlBufferWriteCHAR(buf, septagname);
xmlBufferWriteChar(buf, ">");
}
xmlNodeDump(buf,
nodeset->nodeTab[i]->doc,
nodeset->nodeTab[i],
1, 0);
if ((septagname != NULL) && (xmlStrlen(septagname) > 0))
{
xmlBufferWriteChar(buf, "</");
xmlBufferWriteCHAR(buf, septagname);
xmlBufferWriteChar(buf, ">");
}
}
}
}
if ((toptagname != NULL) && (xmlStrlen(toptagname) > 0))
{
xmlBufferWriteChar(buf, "</");
xmlBufferWriteCHAR(buf, toptagname);
xmlBufferWriteChar(buf, ">");
}
result = xmlStrdup(buf->content);
xmlBufferFree(buf);
return result;
}
/* Translate a PostgreSQL "varlena" -i.e. a variable length parameter
* into the libxml2 representation
*/
xmlChar *
pgxml_texttoxmlchar(text *textstring)
{
xmlChar *res;
int32 txsize;
txsize = VARSIZE(textstring) - VARHDRSZ;
res = (xmlChar *) palloc(txsize + 1);
memcpy((char *) res, VARDATA(textstring), txsize);
res[txsize] = '\0';
return res;
}
/* Public visible XPath functions */
/* This is a "raw" xpath function. Check that it returns child elements
* properly
*/
PG_FUNCTION_INFO_V1(xpath_nodeset);
Datum
xpath_nodeset(PG_FUNCTION_ARGS)
{
xmlChar *xpath, *toptag, *septag;
int32 pathsize;
text
*xpathsupp,
*xpres;
/* PG_GETARG_TEXT_P(0) is document buffer */
xpathsupp = PG_GETARG_TEXT_P(1); /* XPath expression */
toptag = pgxml_texttoxmlchar(PG_GETARG_TEXT_P(2));
septag = pgxml_texttoxmlchar(PG_GETARG_TEXT_P(3));
pathsize = VARSIZE(xpathsupp) - VARHDRSZ;
xpath = pgxml_texttoxmlchar(xpathsupp);
xpres = pgxml_result_to_text(
pgxml_xpath(PG_GETARG_TEXT_P(0),xpath),
toptag,septag,NULL);
/* xmlCleanupParser(); done by result_to_text routine */
pfree((void *) xpath);
if (xpres == NULL)
{
PG_RETURN_NULL();
}
PG_RETURN_TEXT_P(xpres);
}
// The following function is almost identical, but returns the elements in
// a list.
PG_FUNCTION_INFO_V1(xpath_list);
Datum
xpath_list(PG_FUNCTION_ARGS)
{
xmlChar *xpath, *plainsep;
int32 pathsize;
text
*xpathsupp,
*xpres;
/* PG_GETARG_TEXT_P(0) is document buffer */
xpathsupp = PG_GETARG_TEXT_P(1); /* XPath expression */
plainsep = pgxml_texttoxmlchar(PG_GETARG_TEXT_P(2));
pathsize = VARSIZE(xpathsupp) - VARHDRSZ;
xpath = pgxml_texttoxmlchar(xpathsupp);
xpres = pgxml_result_to_text(
pgxml_xpath(PG_GETARG_TEXT_P(0),xpath),
NULL,NULL,plainsep);
/* xmlCleanupParser(); done by result_to_text routine */
pfree((void *) xpath);
if (xpres == NULL)
{
PG_RETURN_NULL();
}
PG_RETURN_TEXT_P(xpres);
}
PG_FUNCTION_INFO_V1(xpath_string);
Datum
xpath_string(PG_FUNCTION_ARGS)
{
xmlChar *xpath;
int32 pathsize;
text
*xpathsupp,
*xpres;
/* PG_GETARG_TEXT_P(0) is document buffer */
xpathsupp = PG_GETARG_TEXT_P(1); /* XPath expression */
pathsize = VARSIZE(xpathsupp) - VARHDRSZ;
/* We encapsulate the supplied path with "string()"
* = 8 chars + 1 for NUL at end */
/* We could try casting to string using the libxml function? */
xpath =(xmlChar *) palloc(pathsize + 9);
memcpy((char *) (xpath+7), VARDATA(xpathsupp), pathsize);
strncpy((char *) xpath, "string(",7);
xpath[pathsize+7] = ')';
xpath[pathsize+8] = '\0';
xpres = pgxml_result_to_text(
pgxml_xpath(PG_GETARG_TEXT_P(0),xpath),
NULL,NULL,NULL);
xmlCleanupParser();
pfree((void *) xpath);
if (xpres == NULL)
{
PG_RETURN_NULL();
}
PG_RETURN_TEXT_P(xpres);
}
PG_FUNCTION_INFO_V1(xpath_number);
Datum
xpath_number(PG_FUNCTION_ARGS)
{
xmlChar *xpath;
int32 pathsize;
text
*xpathsupp;
float4 fRes;
xmlXPathObjectPtr res;
/* PG_GETARG_TEXT_P(0) is document buffer */
xpathsupp = PG_GETARG_TEXT_P(1); /* XPath expression */
pathsize = VARSIZE(xpathsupp) - VARHDRSZ;
xpath = pgxml_texttoxmlchar(xpathsupp);
res = pgxml_xpath(PG_GETARG_TEXT_P(0),xpath);
pfree((void *) xpath);
if (res == NULL)
{
xmlCleanupParser();
PG_RETURN_NULL();
}
fRes = xmlXPathCastToNumber(res);
xmlCleanupParser();
if (xmlXPathIsNaN(fRes))
{
PG_RETURN_NULL();
}
PG_RETURN_FLOAT4(fRes);
}
PG_FUNCTION_INFO_V1(xpath_bool);
Datum
xpath_bool(PG_FUNCTION_ARGS)
{
xmlChar *xpath;
int32 pathsize;
text
*xpathsupp;
int bRes;
xmlXPathObjectPtr res;
/* PG_GETARG_TEXT_P(0) is document buffer */
xpathsupp = PG_GETARG_TEXT_P(1); /* XPath expression */
pathsize = VARSIZE(xpathsupp) - VARHDRSZ;
xpath = pgxml_texttoxmlchar(xpathsupp);
res = pgxml_xpath(PG_GETARG_TEXT_P(0),xpath);
pfree((void *) xpath);
if (res == NULL)
{
xmlCleanupParser();
PG_RETURN_BOOL(false);
}
bRes = xmlXPathCastToBoolean(res);
xmlCleanupParser();
PG_RETURN_BOOL(bRes);
}
/* Core function to evaluate XPath query */
xmlXPathObjectPtr
pgxml_xpath(text *document, xmlChar *xpath)
{
xmlDocPtr doctree;
xmlXPathContextPtr ctxt;
xmlXPathObjectPtr res;
xmlXPathCompExprPtr comppath;
int32 docsize;
docsize = VARSIZE(document) - VARHDRSZ;
pgxml_parser_init();
doctree = xmlParseMemory((char *) VARDATA(document), docsize);
if (doctree == NULL)
{ /* not well-formed */
return NULL;
}
ctxt = xmlXPathNewContext(doctree);
ctxt->node = xmlDocGetRootElement(doctree);
/* compile the path */
comppath = xmlXPathCompile(xpath);
if (comppath == NULL)
{
xmlCleanupParser();
xmlFreeDoc(doctree);
elog_error(ERROR,"XPath Syntax Error",1);
return NULL;
}
/* Now evaluate the path expression. */
res = xmlXPathCompiledEval(comppath, ctxt);
xmlXPathFreeCompExpr(comppath);
if (res == NULL)
{
xmlXPathFreeContext(ctxt);
// xmlCleanupParser();
xmlFreeDoc(doctree);
return NULL;
}
/* xmlFreeDoc(doctree); */
return res;
}
text
*pgxml_result_to_text(xmlXPathObjectPtr res,
xmlChar *toptag,
xmlChar *septag,
xmlChar *plainsep)
{
xmlChar *xpresstr;
int32 ressize;
text *xpres;
if (res == NULL)
{
return NULL;
}
switch (res->type)
{
case XPATH_NODESET:
xpresstr = pgxmlNodeSetToText(res->nodesetval,
toptag,
septag, plainsep);
break;
case XPATH_STRING:
xpresstr = xmlStrdup(res->stringval);
break;
default:
elog(NOTICE, "Unsupported XQuery result: %d", res->type);
xpresstr = xmlStrdup("<unsupported/>");
}
/* Now convert this result back to text */
ressize = strlen(xpresstr);
xpres = (text *) palloc(ressize + VARHDRSZ);
memcpy(VARDATA(xpres), xpresstr, ressize);
VARATT_SIZEP(xpres) = ressize + VARHDRSZ;
/* Free various storage */
xmlCleanupParser();
/* xmlFreeDoc(doctree); -- will die at end of tuple anyway */
xmlFree(xpresstr);
elog_error(ERROR,"XPath error",0);
return xpres;
}
/* xpath_table is a table function. It needs some tidying (as do the
* other functions here!
*/
PG_FUNCTION_INFO_V1(xpath_table);
Datum xpath_table(PG_FUNCTION_ARGS)
{
/* SPI (input tuple) support */
SPITupleTable *tuptable;
HeapTuple spi_tuple;
TupleDesc spi_tupdesc;
/* Output tuple (tuplestore) support */
Tuplestorestate *tupstore = NULL;
TupleDesc ret_tupdesc;
HeapTuple ret_tuple;
ReturnSetInfo *rsinfo = (ReturnSetInfo *) fcinfo->resultinfo;
AttInMetadata *attinmeta;
MemoryContext per_query_ctx;
MemoryContext oldcontext;
/* Function parameters */
char *pkeyfield = GET_STR(PG_GETARG_TEXT_P(0));
char *xmlfield = GET_STR(PG_GETARG_TEXT_P(1));
char *relname = GET_STR(PG_GETARG_TEXT_P(2));
char *xpathset = GET_STR(PG_GETARG_TEXT_P(3));
char *condition = GET_STR(PG_GETARG_TEXT_P(4));
char **values;
xmlChar **xpaths;
xmlChar *pos;
xmlChar *pathsep= "|";
int numpaths;
int ret;
int proc;
int i;
int j;
int rownr; /* For issuing multiple rows from one original document */
int had_values; /* To determine end of nodeset results */
StringInfo querysql;
/* We only have a valid tuple description in table function mode */
if (rsinfo->expectedDesc == NULL) {
ereport(ERROR,(errcode(ERRCODE_SYNTAX_ERROR),
errmsg("xpath_table must be called as a table function")));
}
/* The tuplestore must exist in a higher context than
* this function call (per_query_ctx is used) */
per_query_ctx = rsinfo->econtext->ecxt_per_query_memory;
oldcontext = MemoryContextSwitchTo(per_query_ctx);
/* Create the tuplestore - SortMem is the max in-memory size before it is
* shipped to a disk heap file. Just like ... SortMem!
*/
tupstore = tuplestore_begin_heap(true, false, SortMem);
MemoryContextSwitchTo(oldcontext);
/* get the requested return tuple description */
ret_tupdesc = CreateTupleDescCopy(rsinfo->expectedDesc);
/* At the moment we assume that the returned attributes make sense
* for the XPath specififed (i.e. we trust the caller).
* It's not fatal if they get it wrong - the input function for the
* column type will raise an error if the path result can't be converted
* into the correct binary representation.
*/
attinmeta = TupleDescGetAttInMetadata(ret_tupdesc);
/* We want to materialise because it means that we don't have to
* carry libxml2 parser state between invocations of this function
*/
/* check to see if caller supports us returning a tuplestore */
if (!rsinfo || !(rsinfo->allowedModes & SFRM_Materialize))
ereport(ERROR, (errcode(ERRCODE_SYNTAX_ERROR),
errmsg("xpath_table requires Materialize mode, but it is not "
"allowed in this context")));
// Set return mode and allocate value space.
rsinfo->returnMode = SFRM_Materialize;
rsinfo->setDesc = ret_tupdesc;
values = (char **) palloc(ret_tupdesc->natts * sizeof(char *));
xpaths = (xmlChar **) palloc(ret_tupdesc->natts * sizeof(xmlChar *));
/* Split XPaths. xpathset is a writable CString. */
/* Note that we stop splitting once we've done all needed for tupdesc */
numpaths=0;
pos = xpathset;
do {
xpaths[numpaths] = pos;
pos = strstr(pos,pathsep);
if (pos != NULL) {
*pos = '\0';
pos++;
}
numpaths++;
} while ((pos != NULL) && (numpaths < (ret_tupdesc->natts - 1) ));
/* Now build query */
querysql = makeStringInfo();
/* Build initial sql statement */
appendStringInfo(querysql, "SELECT %s, %s FROM %s WHERE %s",
pkeyfield,
xmlfield,
relname,
condition
);
if ((ret = SPI_connect()) < 0) {
elog(ERROR, "xpath_table: SPI_connect returned %d", ret);
}
if ((ret = SPI_exec(querysql->data,0)) != SPI_OK_SELECT) {
elog(ERROR,"xpath_table: SPI execution failed for query %s",querysql->data);
}
proc= SPI_processed;
/* elog(DEBUG1,"xpath_table: SPI returned %d rows",proc); */
tuptable = SPI_tuptable;
spi_tupdesc = tuptable->tupdesc;
/* Switch out of SPI context */
MemoryContextSwitchTo(oldcontext);
/* Check that SPI returned correct result. If you put a comma into one of
* the function parameters, this will catch it when the SPI query returns
* e.g. 3 columns.
*/
if (spi_tupdesc->natts != 2) {
ereport(ERROR,(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("Expression returning multiple columns is not valid in parameter list"),
errdetail("Expected two columns in SPI result, got %d",spi_tupdesc->natts)));
}
/* Setup the parser. Beware that this must happen in the same context as the
* cleanup - which means that any error from here on must do cleanup to
* ensure that the entity table doesn't get freed by being out of context.
*/
pgxml_parser_init();
/* For each row i.e. document returned from SPI */
for (i=0; i < proc; i++) {
char *pkey;
char *xmldoc;
xmlDocPtr doctree;
xmlXPathContextPtr ctxt;
xmlXPathObjectPtr res;
xmlChar *resstr;
xmlXPathCompExprPtr comppath;
/* Extract the row data as C Strings */
spi_tuple = tuptable->vals[i];
pkey = SPI_getvalue(spi_tuple, spi_tupdesc,1);
xmldoc = SPI_getvalue(spi_tuple, spi_tupdesc,2);
/* Clear the values array, so that not-well-formed documents
* return NULL in all columns.
*/
/* Note that this also means that spare columns will be NULL. */
for (j=0; j < ret_tupdesc->natts; j++) {
values[j]= NULL;
}
/* Insert primary key */
values[0]=pkey;
/* Parse the document */
doctree = xmlParseMemory(xmldoc, strlen(xmldoc));
if (doctree == NULL)
{ /* not well-formed, so output all-NULL tuple */
ret_tuple = BuildTupleFromCStrings(attinmeta, values);
oldcontext = MemoryContextSwitchTo(per_query_ctx);
tuplestore_puttuple(tupstore, ret_tuple);
MemoryContextSwitchTo(oldcontext);
heap_freetuple(ret_tuple);
}
else
{
/* New loop here - we have to deal with nodeset results */
rownr=0;
do {
/* Now evaluate the set of xpaths. */
had_values=0;
for (j=0; j < numpaths; j++) {
ctxt = xmlXPathNewContext(doctree);
ctxt->node = xmlDocGetRootElement(doctree);
xmlSetGenericErrorFunc(ctxt, pgxml_errorHandler);
/* compile the path */
comppath = xmlXPathCompile(xpaths[j]);
if (comppath == NULL)
{
xmlCleanupParser();
xmlFreeDoc(doctree);
elog_error(ERROR,"XPath Syntax Error",1);
PG_RETURN_NULL(); /* Keep compiler happy */
}
/* Now evaluate the path expression. */
res = xmlXPathCompiledEval(comppath, ctxt);
xmlXPathFreeCompExpr(comppath);
if (res != NULL)
{
switch (res->type)
{
case XPATH_NODESET:
/* We see if this nodeset has enough nodes */
if ((res->nodesetval != NULL) && (rownr < res->nodesetval->nodeNr)) {
resstr =
xmlXPathCastNodeToString(res->nodesetval->nodeTab[rownr]);
had_values=1;
} else {
resstr = NULL;
}
break;
case XPATH_STRING:
resstr = xmlStrdup(res->stringval);
break;
default:
elog(NOTICE, "Unsupported XQuery result: %d", res->type);
resstr = xmlStrdup("<unsupported/>");
}
// Insert this into the appropriate column in the result tuple.
values[j+1] = resstr;
}
xmlXPathFreeContext(ctxt);
}
// Now add the tuple to the output, if there is one.
if (had_values) {
ret_tuple = BuildTupleFromCStrings(attinmeta, values);
oldcontext = MemoryContextSwitchTo(per_query_ctx);
tuplestore_puttuple(tupstore, ret_tuple);
MemoryContextSwitchTo(oldcontext);
heap_freetuple(ret_tuple);
}
rownr++;
} while (had_values);
}
xmlFreeDoc(doctree);
pfree(pkey);
pfree(xmldoc);
}
xmlCleanupParser();
/* Needed to flag completeness in 7.3.1. 7.4 defines it as a no-op. */
tuplestore_donestoring(tupstore);
SPI_finish();
rsinfo->setResult=tupstore;
/*
* SFRM_Materialize mode expects us to return a NULL Datum. The actual
* tuples are in our tuplestore and passed back through
* rsinfo->setResult. rsinfo->setDesc is set to the tuple description
* that we actually used to build our tuples with, so the caller can
* verify we did what it was expecting.
*/
return (Datum) 0;
}