A short SPARQL query can be compiled into a long SQL statement, especially if data comes from many quad map patterns. A moderately sized application with 50 tables and 10 columns per table may create thousands of quad map patterns for subjects spanning hundreds of different types. An attempt to "select everything" from Linked Data View of that complexity may easily create 5000 lines of SQL code. Thus it is to be expected that some queries will be rejected even if the same queries would work fine if the RDF data were held as physical quads in default storage, rather than synthesized through an Linked Data View.

In addition, the SQL compiler catches typos efficiently, signalling an error if a table or column name is unknown, efficiently catching typos. SPARQL uses IRIs that are long and sometimes unreadable, but there is no "closed world" schema of the data so a typo in an IRI is not an error; it is simply some other IRI. So a typo in an IRI or in a namespace prefix causes missing bindings of some triple patterns of the query and an incomplete result, but usually no errors are reported. A typo in graph or predicate IRI may cause the SPARQL compiler to generate code that accesses default (quad) storage instead of a relational source or generate empty code that accesses nothing.

The SQL compiler does not signal casting errors when it runs the statement generated from SPARQL, because the generated SQL code contains option (QUIETCAST) . This means that mismatches between expected and actual datatypes of values stay invisible and may cause rounding errors (e.g. integer division instead of floating-point) and even empty joins (due to join conditions that silently return NULL instead of returning a comparison error).

In other words, SPARQL queries are so laconic that there is no room for details that let the compiler distinguish between intent and a bug. This masks query complexity, misuse of names and type mismatches. One may make debugging easier by making queries longer.

Two very helpful debugging tools are automatic void variable recognition and plain old code inspection. "Automatic" means "cheap" so the very first step of debugging is to ensure that every triple pattern of the query may in principle return something. This helps in finding typos when the query gets data from Linked Data Views. It also helps when a query tries to join two disjoint sorts of subjects. If the define sql:signal-void-variables 1 directive is placed in the preamble of the SPARQL query, the compiler will signal an error if it finds any triple pattern that cannot bind variables or any variable that is proved to be always unbound. This is especially useful when data are supposed to come from an option (exclusive) or option (soft exclusive) quad map. Without one of these options, the SPARQL compiler will usually bind variables using "physical quads"; the table of physical quads may contain any rows that match any given triple pattern; thus many errors will remain undiscovered. If the name of a quad map pattern is known then it is possible to force the SPARQL compiler to use only that quad map for the whole query or a part of the query. This is possible by using the following syntax:

QUAD MAP quad-map-name { group-pattern }

If some triple pattern inside group-pattern cannot be bound using quad-map-name or one of its descendants then define sql:signal-void-variables 1 will force the compiler to signal the error.

[Note] Note

Although it is technically possible to use QUAD MAP to improve the performance of a query that tries to access redundant Linked Data Views, it is much better to achieve the same effect by providing a more restrictive query or by changing/extending the Linked Data View. If an application relies on this trick then interoperable third-party SPARQL clients may experience problems because they cannot use Virtuoso-specific extensions.

If the automated query checking gives nothing, function sparql_to_sql_text can be used in order to get the SQL text generated from the given query. Its only argument is the text of the SPARQL query to compile (without any leading SPARQL keyword or semicolon at the end). The returned value is the SQL text. The output may be long but it is the most authoritative source of diagnostic data.

When called from ISQL or an ODBC client, the return value of sparql_to_sql_text may be transferred as a BLOB so ISQL requires the "set blobs on" instruction to avoid data truncation. Even better, the SQL text can be saved to a file:

string_to_file ('debug.sql', sparql_to_sql_text ('SELECT * WHERE { graph ?g { ?s a ?type }}'), -2);

(The -2 is to overwrite the previous version of the file, as this function may be called many times).

[Note] Note

When passing the query text to sparql_to_sql_text, if the query contains single quotes, each embedded single quote must be doubled up. Use double quotes in SPARQL queries to avoid this inconvenience.

As an example, let's find out why the query

PREFIX northwind: <http://demo.openlinksw.com/schemas/northwind#>
FROM <http://myhost.example.com/Northwind>
    ?order1 northwind:has_salesrep ?emp ; northwind:shipCountry ?country1 .
    ?order2 northwind:has_salesrep ?emp ; northwind:shipCountry ?country2 .
    filter (?country1 != ?country2) }

is much slower than a similar SQL statement. The call of sparql_to_sql_text returns the equivalent SQL statement:

SELECT DISTINCT sprintf_iri ( 'http://myhost.example.com/Northwind/Employee/%U%U%d#this' ,
    /*retval[*/ "s-6-1-t0"."b067b7d~FirstName~0" /* emp */ /*]retval*/ ,
    /*retval[*/  "s-6-1-t0"."b067b7d~FirstName~1" /*]retval*/ ,
    /*retval[*/  "s-6-1-t0"."b067b7d~FirstName~2" /*]retval*/ ) AS /*tmpl*/ "emp"
FROM (SELECT "s-6-1-t0-int~orders"."OrderID" AS /*tmpl*/ "20ffecc~OrderID",
         "s-6-1-t0-int~employees"."FirstName" AS /*as-name-N*/ "b067b7d~FirstName~0",
         "s-6-1-t0-int~employees"."LastName" AS /*as-name-N*/ "b067b7d~FirstName~1",
         "s-6-1-t0-int~employees"."EmployeeID" AS /*as-name-N*/ "b067b7d~FirstName~2"
         FROM Demo.demo.Employees AS "s-6-1-t0-int~employees", Demo.demo.Orders AS "s-6-1-t0-int~orders"
         WHERE /* inter-alias join cond */
       "s-6-1-t0-int~orders".EmployeeID = "s-6-1-t0-int~employees".EmployeeID) AS "s-6-1-t0",
    (SELECT "s-6-1-t1-int~orders"."OrderID" AS /*tmpl*/ "20ffecc~OrderID",
        "s-6-1-t1-int~orders"."ShipCountry" AS /*tmpl*/ "e45a7f~ShipCountry"
        FROM Demo.demo.Orders AS "s-6-1-t1-int~orders") AS "s-6-1-t1",
    (SELECT "s-6-1-t2-int~orders"."OrderID" AS /*tmpl*/ "20ffecc~OrderID",
        "s-6-1-t2-int~employees"."FirstName" AS /*as-name-N*/ "b067b7d~FirstName~0",
        "s-6-1-t2-int~employees"."LastName" AS /*as-name-N*/ "b067b7d~FirstName~1",
        "s-6-1-t2-int~employees"."EmployeeID" AS /*as-name-N*/ "b067b7d~FirstName~2"
        FROM Demo.demo.Employees AS "s-6-1-t2-int~employees", Demo.demo.Orders AS "s-6-1-t2-int~orders"
    WHERE /* inter-alias join cond */
       "s-6-1-t2-int~orders".EmployeeID = "s-6-1-t2-int~employees".EmployeeID) AS "s-6-1-t2",
    (SELECT "s-6-1-t3-int~orders"."OrderID" AS /*tmpl*/ "20ffecc~OrderID",
        "s-6-1-t3-int~orders"."ShipCountry" AS /*tmpl*/ "e45a7f~ShipCountry"
    FROM Demo.demo.Orders AS "s-6-1-t3-int~orders") AS "s-6-1-t3"
WHERE /* two fields belong to same equiv */
    /*retval[*/  "s-6-1-t0"."20ffecc~OrderID" /* order1 */ /*]retval*/  =
    /*retval[*/  "s-6-1-t1"."20ffecc~OrderID" /* order1 */ /*]retval*/
    AND /* two fields belong to same equiv */
    sprintf_iri ( 'http://myhost.example.com/Northwind/Employee/%U%U%d#this' ,
        /*retval[*/  "s-6-1-t0"."b067b7d~FirstName~0" /* emp */ /*]retval*/ ,
        /*retval[*/  "s-6-1-t0"."b067b7d~FirstName~1" /*]retval*/ ,
        /*retval[*/  "s-6-1-t0"."b067b7d~FirstName~2" /*]retval*/ ) =
    sprintf_iri ( 'http://myhost.example.com/Northwind/Employee/%U%U%d#this' ,
        /*retval[*/  "s-6-1-t2"."b067b7d~FirstName~0" /* emp */ /*]retval*/ ,
        /*retval[*/  "s-6-1-t2"."b067b7d~FirstName~1" /*]retval*/ ,
        /*retval[*/  "s-6-1-t2"."b067b7d~FirstName~2" /*]retval*/ )
    AND /* two fields belong to same equiv */
    /*retval[*/  "s-6-1-t2"."20ffecc~OrderID" /* order2 */ /*]retval*/  =
    /*retval[*/  "s-6-1-t3"."20ffecc~OrderID" /* order2 */ /*]retval*/
    AND /* filter */
   ( /*retval[*/  "s-6-1-t1"."e45a7f~ShipCountry" /* country1 */ /*]retval*/  <>
        /*retval[*/  "s-6-1-t3"."e45a7f~ShipCountry" /* country2 */ /*]retval*/ )

The query is next to unreadable but some comments split it into meaningful expressions. Every triple (or list of similar triples) becomes a subquery that returns fields needed to build the values of bound variables. The fields are printed wrapped by comments like /*retval[*/ expression /* original variable name */ /*]retval*/ . Names like"s-6-1-t0" contain the source line number where a group pattern begins (6) and the serial number of the triple (0). Comment /* inter-alias join cond */ means that the expression which follows is the condition as written in the declaration of the quad map pattern. Comment /* filter */ precedes expressions for FILTER expressions in the source SPARQL. The word "equiv" means "equivalence class", i.e. a group of occurrences of variables in the source query such that all occurrences are bound to the same value. E.g. when a name repeats in many triples of a group, all its occurrences form an equivalence class. In some cases the compiler can prove that two variables are always equal even if the names differ - these variables are also placed into an "equiv".

Looking at this query, you may notice equalities like sprintf_iri (...) = sprintf_iri (...) . That is sub-optimal because it indicates that no index will be used to optimize the join and that there will be one function call per row. When the variable ?emp appears in two different triples, it means that the value of the variable is the same in both triples. The query compares IRIs instead of comparing the arguments of sprintf_iri because the format string is not proven to be a bijection. Indeed it cannot be a bijection for arbitrary strings, but the database must reflect the real world. If it is assumed that the real names of persons never start with a digit, within the %d%U format fragment, the digits will always be distinguishable from the name; so the IRI class can be declared as a bijection even if it is not true for arbitrary strings. The script can then include "suspicious" option (bijection) as follows:

create iri class sample:Employee "http://example.com/Employee/%d%U#this"
  (in employee_id integer not null, in employee_lastname varchar not null)
  option (bijection) .

Unfortunately, attempts to use the same trick with the declaration from the Northwind example will fail:

create iri class northwind:Employee "http://^{URIQADefaultHost}^/Northwind/Employee/%U%U%d#this"
  (in employee_firstname varchar not null, in employee_lastname varchar not null, in employee_id integer not null)
  option (bijection) .

Bijection will allow the parsing, but it will never give the proper result, because the first %U will read the whole concatenation of %U%U%d , leaving nothing before the#this for the second %U (this is an error) and leaving nothing for the %d (that is an explicit parse error, becauses the integer field cannot be empty).

The string parser will process the string from left to right so it will be unable to parse the string. The compiler might sometimes report an error if it can prove that the format string is not appropriate for bijection.

The correct way of improving the Northwind example is to enable reliable bijection by adding strong delimiters:

create iri class northwind:Employee "http://^{URIQADefaultHost}^/Northwind/Employee/%U/%U/%d#this"
  (in employee_firstname varchar not null, in employee_lastname varchar not null, in employee_id integer not null)
  option (bijection) .

After running the updated script, the query contains three comparisons of fields that were arguments of sprintf_iri in the previous version.

Example for casting string as IRI type

create function DB.DBA.RDF_DF_GRANTEE_ID_URI (in id integer)
  declare isrole integer;
  isrole := coalesce ((SELECT top 1 U_IS_ROLE FROM DB.DBA.SYS_USERS WHERE U_ID = id));
  if (isrole is null)
    return NULL;
  else if (isrole)
    return sprintf ('http://%s/sys/group?id=%d', registry_get ('URIQADefaultHost'), id);
    return sprintf ('http://%s/sys/user?id=%d', registry_get ('URIQADefaultHost'), id);


create function DB.DBA.RDF_DF_GRANTEE_ID_URI_INVERSE (in id_iri varchar)
  declare parts any;
  parts := sprintf_inverse (id_iri, sprintf ('http://%s/sys/user?id=%%d', registry_get ('URIQADefaultHost')), 1);
  if (parts is not null)
      if (exists (SELECT TOP 1 1 FROM DB.DBA.SYS_USERS WHERE U_ID = parts[0] and not U_IS_ROLE))
        return parts[0];
  parts := sprintf_inverse (id_iri, sprintf ('http://%s/sys/group?id=%%d', registry_get ('URIQADefaultHost')), 1);
  if (parts is not null)
      if (exists (SELECT TOP 1 1 FROM DB.DBA.SYS_USERS WHERE U_ID = parts[0] and U_IS_ROLE))
        return parts[0];
  return NULL;


create iri class oplsioc:grantee_iri using
  function DB.DBA.RDF_DF_GRANTEE_ID_URI (in id integer) returns varchar ,
  function DB.DBA.RDF_DF_GRANTEE_ID_URI_INVERSE (in id_iri varchar) returns integer
  option ( bijection ,
    returns     "http://^{URIQADefaultHost}^/sys/group?id=%d"
    union       "http://^{URIQADefaultHost}^/sys/user?id=%d" ) .