== ANAPSID ==


* source:

  https://github.com/anapsid/anapsid

* execute:
 
 1. get_predicates endpoint.txt endpoint.pre get the $ENDPOINTS file in step2.

 2. 
run_anapsid -e $ENDPOINTS -q $query -p <planType> -s False -o False -d <TypeofDecompostion> -a True -w False [-k <special>]  [-V <typeOfEndpoint>] -r False


$ENDPOINTS: path and name of the file where the description of the endpoints is stored.

$query: path and name of the filw where the query is stored.

<planType>: can be b if the plan is bushy, ll is the plan is left linear, and naive for naive binary tree plan.

-o: can be True or False. True indicates that the input query is in SPARQL1-1 and no decomposition is needed; False, otherwise.

-d: indicates the type of Decomposition. can be SSGM (Star Shaped Group Multiple Endpoints), SSGS (Star Shaped Group Single Endpoint), EG (Exclusive Groups), Recommended SSGM.

-a: indicates if the adaptive operators will be used. Recommended value True.

-w: can be True or False. Indicates if the cardinality of the queries will be estimated by contacting the sources (True) or by using a cost model (False). Turning True this feature may affect execution time.

-k: can be y or c. The value y indicates that the plan will be produced, while c asks that decomposition. This parameter is optional, and should be set up only if the plan of the query wants to be produced.

-r: can be True or False. Use True if the answer of the query will be output and False if only a summary of the execution will be produced.

-V: can be True or False. True indicates if the endpoints to contact are Virtuoso, False is of any other type, e.g., OWLIM.


'''please notice that the Usage information printed when running "./run_anapsid" is wrong!'''



run_anapsid -e endpoint.pre  -q query/query1.txt -p ll -s False -o False -d SSGM -a True -w False -r True >result

* Result:

query1: return 5 results.

||Query\time(ms) ||time 1 || time 2 || time 3 ||time 4||time 5 ||avg||
||case1 ||8995	||9360	||8971	||8894||	8760||8905||

with bush plan, the time took for query 1 is as follows(milsecs):
8557 8330 8533 8692 11943

for case2, no result. "query2 0.00762486457825        0.00762581825256        0.007625818252560.007625 81825256        0       0"

for case3, returned unidentified result: ' rdf'

for case4, no result. returned "query4  0.00775194168091        0.00775289535522        0.007752895355220.007752 89535522        0       0"

for case5: returned error messenge "LexToken(POINT,'.',1,736) unknown text at '.'".


Result: the query parser of ANAPSID is weak: cannot deal with triple pattern like "?drug_kegg kegg:x-pubchem.compound ?cpd." and "?drug sider_vocab:generic-name ?generic;
dcterms:title ?drug_name;" (unknown text at ';')