/*
- * $PostgreSQL: pgsql/contrib/xml2/xpath.c,v 1.25 2010/01/17 12:11:25 mha Exp $
+ * $PostgreSQL: pgsql/contrib/xml2/xpath.c,v 1.26 2010/02/28 19:51:37 tgl Exp $
*
* Parser interface for DOM-based parser (libxml) rather than
* stream-based SAX-type parser
PG_MODULE_MAGIC;
-/* declarations */
+/* externally accessible functions */
+
+Datum xml_is_well_formed(PG_FUNCTION_ARGS);
+Datum xml_encode_special_chars(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);
+
+/* these are exported for use by xslt_proc.c */
+
+void elog_error(const char *explain, bool force);
+void pgxml_parser_init(void);
+
+/* local declarations */
static void *pgxml_palloc(size_t size);
static void *pgxml_repalloc(void *ptr, size_t size);
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,
+static text *pgxml_result_to_text(xmlXPathObjectPtr res, xmlChar *toptag,
xmlChar *septag, xmlChar *plainsep);
-xmlChar *pgxml_texttoxmlchar(text *textstring);
+static xmlChar *pgxml_texttoxmlchar(text *textstring);
static xmlXPathObjectPtr pgxml_xpath(text *document, xmlChar *xpath);
-Datum xml_is_well_formed(PG_FUNCTION_ARGS);
-Datum xml_encode_special_chars(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 */
+static char *pgxml_errorMsg = NULL; /* overall error message */
-#define ERRBUF_SIZE 200
/* memory handling passthrough functions (e.g. palloc, pstrdup are
currently macros, and the others might become so...) */
return pstrdup(string);
}
-/* The error handling function. This formats an error message and sets
+/*
+ * 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,...)
{
+ char errbuf[1024]; /* per line error buffer */
va_list args;
+ /* Format the message */
va_start(args, msg);
- vsnprintf(errbuf, ERRBUF_SIZE, msg, args);
+ vsnprintf(errbuf, sizeof(errbuf), msg, args);
va_end(args);
- /* Now copy the argument across */
+ /* Store in, or append to, pgxml_errorMsg */
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';
+ size_t oldsize = strlen(pgxml_errorMsg);
+ size_t newsize = strlen(errbuf);
+ /*
+ * We intentionally discard the last char of the existing message,
+ * which should be a carriage return. (XXX wouldn't it be saner
+ * to keep it?)
+ */
+ pgxml_errorMsg = repalloc(pgxml_errorMsg, oldsize + newsize);
+ memcpy(&pgxml_errorMsg[oldsize - 1], errbuf, newsize);
+ pgxml_errorMsg[oldsize + newsize - 1] = '\0';
}
- memset(errbuf, 0, ERRBUF_SIZE);
}
-/* This function reports the current message at the level specified */
+/*
+ * This function ereports the current message if any. If force is true
+ * then an error is thrown even if pgxml_errorMsg hasn't been set.
+ */
void
-elog_error(int level, char *explain, int force)
+elog_error(const char *explain, bool force)
{
- if (force || (pgxml_errorMsg != NULL))
+ if (force || pgxml_errorMsg != NULL)
{
if (pgxml_errorMsg == NULL)
- {
- ereport(level, (errcode(ERRCODE_EXTERNAL_ROUTINE_EXCEPTION),
- errmsg("%s", explain)));
- }
+ ereport(ERROR,
+ (errcode(ERRCODE_EXTERNAL_ROUTINE_EXCEPTION),
+ errmsg("%s", explain)));
else
- {
- ereport(level, (errcode(ERRCODE_EXTERNAL_ROUTINE_EXCEPTION),
- errmsg("%s:%s", explain, pgxml_errorMsg)));
- pfree(pgxml_errorMsg);
- }
+ ereport(ERROR,
+ (errcode(ERRCODE_EXTERNAL_ROUTINE_EXCEPTION),
+ errmsg("%s: %s", explain, pgxml_errorMsg)));
}
}
+/*
+ * Initialize for xml parsing.
+ */
void
-pgxml_parser_init()
+pgxml_parser_init(void)
{
- /*
- * This code could also set parser settings from user-supplied info.
- * Quite how these settings are made is another matter :)
- */
+ /* Set up error handling */
+ pgxml_errorMsg = NULL;
+ xmlSetGenericErrorFunc(NULL, pgxml_errorHandler);
+ /* Replace libxml memory handling (DANGEROUS) */
xmlMemSetup(pgxml_pfree, pgxml_palloc, pgxml_repalloc, pgxml_pstrdup);
- xmlInitParser();
- xmlSetGenericErrorFunc(NULL, pgxml_errorHandler);
+ /* Initialize libxml */
+ xmlInitParser();
xmlSubstituteEntitiesDefault(1);
xmlLoadExtDtdDefaultValue = 1;
-
- pgxml_errorMsg = NULL;
-
- errbuf = palloc(200);
- memset(errbuf, 0, 200);
-
}
Datum
xml_is_well_formed(PG_FUNCTION_ARGS)
{
- /* called as xml_is_well_formed(document) */
- xmlDocPtr doctree;
text *t = PG_GETARG_TEXT_P(0); /* document buffer */
int32 docsize = VARSIZE(t) - VARHDRSZ;
+ xmlDocPtr doctree;
pgxml_parser_init();
xmlCleanupParser();
PG_RETURN_BOOL(false); /* i.e. not well-formed */
}
- xmlCleanupParser();
xmlFreeDoc(doctree);
+ xmlCleanupParser();
PG_RETURN_BOOL(true);
}
PG_RETURN_TEXT_P(tout);
}
-static xmlChar
- *
+/*
+ * 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) separator - if used, then nodes are
+ * cast to string as output method
+ */
+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;
{
for (i = 0; i < nodeset->nodeNr; i++)
{
-
if (plainsep != NULL)
{
xmlBufferWriteCHAR(buf,
}
else
{
-
-
if ((septagname != NULL) && (xmlStrlen(septagname) > 0))
{
xmlBufferWriteChar(buf, "<");
/* Translate a PostgreSQL "varlena" -i.e. a variable length parameter
* into the libxml2 representation
*/
-
-xmlChar *
+static xmlChar *
pgxml_texttoxmlchar(text *textstring)
{
return (xmlChar *) text_to_cstring(textstring);
}
-/* Public visible XPath functions */
+/* Publicly visible XPath functions */
-/* This is a "raw" xpath function. Check that it returns child elements
+/*
+ * This is a "raw" xpath function. Check that it returns child elements
* properly
*/
-
PG_FUNCTION_INFO_V1(xpath_nodeset);
Datum
*toptag,
*septag;
int32 pathsize;
- text
- *xpathsupp,
+ text *xpathsupp,
*xpres;
/* PG_GETARG_TEXT_P(0) is document buffer */
xpath = pgxml_texttoxmlchar(xpathsupp);
- xpres = pgxml_result_to_text(
- pgxml_xpath(PG_GETARG_TEXT_P(0), xpath),
+ xpres = pgxml_result_to_text(pgxml_xpath(PG_GETARG_TEXT_P(0), xpath),
toptag, septag, NULL);
/* xmlCleanupParser(); done by result_to_text routine */
PG_RETURN_TEXT_P(xpres);
}
-/* The following function is almost identical, but returns the elements in */
-/* a list. */
-
+/*
+ * The following function is almost identical, but returns the elements in
+ * a list.
+ */
PG_FUNCTION_INFO_V1(xpath_list);
Datum
xmlChar *xpath,
*plainsep;
int32 pathsize;
- text
- *xpathsupp,
+ text *xpathsupp,
*xpres;
/* PG_GETARG_TEXT_P(0) is document buffer */
xpath = pgxml_texttoxmlchar(xpathsupp);
- xpres = pgxml_result_to_text(
- pgxml_xpath(PG_GETARG_TEXT_P(0), xpath),
+ xpres = pgxml_result_to_text(pgxml_xpath(PG_GETARG_TEXT_P(0), xpath),
NULL, NULL, plainsep);
/* xmlCleanupParser(); done by result_to_text routine */
{
xmlChar *xpath;
int32 pathsize;
- text
- *xpathsupp,
+ text *xpathsupp,
*xpres;
/* PG_GETARG_TEXT_P(0) is document buffer */
xpath[pathsize + 7] = ')';
xpath[pathsize + 8] = '\0';
- xpres = pgxml_result_to_text(
- pgxml_xpath(PG_GETARG_TEXT_P(0), xpath),
+ xpres = pgxml_result_to_text(pgxml_xpath(PG_GETARG_TEXT_P(0), xpath),
NULL, NULL, NULL);
xmlCleanupParser();
{
xmlChar *xpath;
int32 pathsize;
- text
- *xpathsupp;
-
+ text *xpathsupp;
float4 fRes;
xmlXPathObjectPtr res;
PG_RETURN_NULL();
PG_RETURN_FLOAT4(fRes);
-
}
{
xmlChar *xpath;
int32 pathsize;
- text
- *xpathsupp;
-
+ text *xpathsupp;
int bRes;
xmlXPathObjectPtr res;
bRes = xmlXPathCastToBoolean(res);
xmlCleanupParser();
PG_RETURN_BOOL(bRes);
-
}
/* Core function to evaluate XPath query */
-xmlXPathObjectPtr
+static xmlXPathObjectPtr
pgxml_xpath(text *document, xmlChar *xpath)
{
-
xmlDocPtr doctree;
xmlXPathContextPtr ctxt;
xmlXPathObjectPtr res;
-
xmlXPathCompExprPtr comppath;
-
int32 docsize;
-
docsize = VARSIZE(document) - VARHDRSZ;
pgxml_parser_init();
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;
+ elog_error("XPath Syntax Error", true);
}
/* Now evaluate the path expression. */
return res;
}
-text
- *
+static text *
pgxml_result_to_text(xmlXPathObjectPtr res,
xmlChar *toptag,
xmlChar *septag,
xpresstr = xmlStrdup((const xmlChar *) "<unsupported/>");
}
-
/* Now convert this result back to text */
xpres = cstring_to_text((char *) xpresstr);
xmlFree(xpresstr);
- elog_error(ERROR, "XPath error", 0);
-
+ elog_error("XPath error", false);
return xpres;
}
-/* xpath_table is a table function. It needs some tidying (as do the
+/*
+ * 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 */
+ /* Function parameters */
+ char *pkeyfield = text_to_cstring(PG_GETARG_TEXT_PP(0));
+ char *xmlfield = text_to_cstring(PG_GETARG_TEXT_PP(1));
+ char *relname = text_to_cstring(PG_GETARG_TEXT_PP(2));
+ char *xpathset = text_to_cstring(PG_GETARG_TEXT_PP(3));
+ char *condition = text_to_cstring(PG_GETARG_TEXT_PP(4));
+
+ /* SPI (input tuple) support */
SPITupleTable *tuptable;
HeapTuple spi_tuple;
TupleDesc spi_tupdesc;
-/* Output tuple (tuplestore) support */
+ /* Output tuple (tuplestore) support */
Tuplestorestate *tupstore = NULL;
TupleDesc ret_tupdesc;
HeapTuple ret_tuple;
MemoryContext per_query_ctx;
MemoryContext oldcontext;
-/* Function parameters */
- char *pkeyfield = text_to_cstring(PG_GETARG_TEXT_PP(0));
- char *xmlfield = text_to_cstring(PG_GETARG_TEXT_PP(1));
- char *relname = text_to_cstring(PG_GETARG_TEXT_PP(2));
- char *xpathset = text_to_cstring(PG_GETARG_TEXT_PP(3));
- char *condition = text_to_cstring(PG_GETARG_TEXT_PP(4));
-
char **values;
xmlChar **xpaths;
char *pos;
int rownr; /* For issuing multiple rows from one original
* document */
int had_values; /* To determine end of nodeset results */
-
StringInfoData query_buf;
/* We only have a valid tuple description in table function mode */
pkeyfield,
xmlfield,
relname,
- condition
- );
-
+ condition);
if ((ret = SPI_connect()) < 0)
elog(ERROR, "xpath_table: SPI_connect returned %d", ret);
if ((ret = SPI_exec(query_buf.data, 0)) != SPI_OK_SELECT)
- elog(ERROR, "xpath_table: SPI execution failed for query %s", query_buf.data);
+ elog(ERROR, "xpath_table: SPI execution failed for query %s",
+ query_buf.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 */
+ /* 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.
- */
-
+ /*
+ * 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),
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.
- */
+ /*
+ * 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 */
{
char *pkey;
char *xmldoc;
-
xmlDocPtr doctree;
xmlXPathContextPtr ctxt;
xmlXPathObjectPtr res;
xmlChar *resstr;
-
-
xmlXPathCompExprPtr comppath;
/* Extract the row data as C Strings */
/*
* Clear the values array, so that not-well-formed documents return
- * NULL in all columns.
+ * NULL in all columns. Note that this also means that spare columns
+ * will be NULL.
*/
-
- /* Note that this also means that spare columns will be NULL. */
for (j = 0; j < ret_tupdesc->natts; j++)
values[j] = NULL;
had_values = 0;
for (j = 0; j < numpaths; j++)
{
-
ctxt = xmlXPathNewContext(doctree);
ctxt->node = xmlDocGetRootElement(doctree);
xmlSetGenericErrorFunc(ctxt, pgxml_errorHandler);
xmlCleanupParser();
xmlFreeDoc(doctree);
- elog_error(ERROR, "XPath Syntax Error", 1);
+ elog_error("XPath Syntax Error", true);
PG_RETURN_NULL(); /* Keep compiler happy */
}
resstr = xmlStrdup((const xmlChar *) "<unsupported/>");
}
-
/*
* Insert this into the appropriate column in the
* result tuple.
}
xmlXPathFreeContext(ctxt);
}
+
/* Now add the tuple to the output, if there is one. */
if (had_values)
{
}
rownr++;
-
} while (had_values);
-
}
xmlFreeDoc(doctree);
}
xmlCleanupParser();
-/* Needed to flag completeness in 7.3.1. 7.4 defines it as a no-op. */
+
tuplestore_donestoring(tupstore);
SPI_finish();
* expecting.
*/
return (Datum) 0;
-
}
projects / postgresql / commitdiff