Multi CPU testrun

Author: mathiasrw

README.md

@@ -6,37 +6,27 @@ Testing SQL compabillity for [AlaSQL](https://github.com/agershun/alasql)
 How?
 ----
 
-Go get markdown summery run npm test (will include starting node with garbagecollection set to true):
+Go get markdown summery run npm test :
 
     npm test
 
-Result will output to results.md
+Result will output to `results/results.md`
 
 
-----
-
-**==> The rest of this file is probably outdated <==**
-
-----
-
-**To get more details** execute `run_raw.js` in the sqllogic folder. As the output is very big its suggested to direct output to a file and let the garbagecollection be initiated between the run of each file. To execute `run_raw.js` run node with the --expose-gc flag:
-
-    cd sqllogic
-    node --expose-gc run_raw > results.log
 
 
 What?
 -----
 
-The output of `run_raw.js` is the output of each test file followed by the total score. (so you can keep track on overall % of passed tests why its running)
+The output of `run.js` is the output of each test file followed by the total score. (so you can keep track on overall % of passed tests why its running)
 
 There are about 2 million tests in the full test suite. Many of the tests will trigger the same kind of error - this is why the output will only print the same error the first time it comes forward. To have more examples it will randomly print errors even if it has been printes before. You can also choose to print all errors. Please see the config section in the run.js file.
 
 Please note that the current version of the script **ONLY TESTS IF THE SQL CAN COMPILE** - not if the returned values are correct.
 
 To make sure you are testing the lates stable release please run the following:
 
-    cd sqllogic && npm install alasql
+    npm install alasql
 
 Please see the config section of the `run` files to run tests on local version instead of npm version.
 
@@ -62,6 +52,7 @@ Supported
 
 - NULL values are rendered as "NULL". Empty strings are rendered as "(empty)". Within non-empty strings, all control characters and unprintable characters are rendered as "@".
 
+- Implement verification of returned valued with sortorder set
 
 
 
@@ -77,9 +68,6 @@ ToDo
 
 
 
-- Implement verification of returned valued with sortorder set
-> The <sort-mode> argument is optional. If included, it must be one of "nosort", "rowsort", or "valuesort". The default is "nosort". In nosort mode, the results appear in exactly the order in which they were received from the database engine. The nosort mode should only be used on queries that have an ORDER BY clause or which only have a single row of result, since otherwise the order of results is undefined and might vary from one database engine to another. The "rowsort" mode gathers all output from the database engine then sorts it by rows on the client side. Sort comparisons use strcmp() on the rendered ASCII text representation of the values. Hence, "9" sorts after "10", not before. The "valuesort" mode works like rowsort except that it does not honor row groupings. Each individual result value is sorted on its own.
-
 **Not ToDo**
 - Impelment mocha testframework (testresults does not get printed before all are done - so fills the memory)
 
@@ -102,26 +90,26 @@ The output from run.js is structured the following way:
 
 The output from a single testfile could look like this
 
-	-----------------------------
-	Looking at ./test/index/between/1/slt_good_0.test
-
-	Ran 22 tests
-	100% was OK
-	0 was not
-	./test/index/between/1/slt_good_0.test: 47ms
-
-	-----------------------------
-
+> ### 11/32 [`./test/index/delete/1000/slt_good_0.test`](https://github.com/mathiasrw/alasql-logictest/blob/master/sqllogic/./test/index/delete/1000/slt_good_0.test)
+> 
+> _Mimic sqlite_
+> 
+> #### ★ Ran 11,924 tests as _sqlite_
+> 
+> * 100% was OK
+> 
+> `sqlite 11924 OK: ./test/index/delete/1000/slt_good_0.test`
+ 
 It makes sense, as no errors was found. But it can also look like this:
 
-	-----------------------------
-	Looking at ./test/evidence/slt_lang_update.test
-
-	Ran 25 tests
-	96% was OK
-	1 was not
-	./test/evidence/slt_lang_update.test: 98ms
-
-	-----------------------------
-
-making less sense as an error was found but was not printed. Because there are about 2 million tests in the full testset the same errors will be found many times. To limit noise an error only gets printed first time its seen. Settings for printing errors can be found in the config section of run_x.js
+> ### 6/32 [`./test/index/commute/100/slt_good_0.test`](https://github.com/mathiasrw/alasql-logictest/blob/master/sqllogic/./test/index/commute/100/slt_good_0.test)
+> 
+> _Mimic sqlite_
+> 
+> #### ☓ Ran 10,122 tests as _sqlite_
+> 
+> * 20 failed
+> * 99% was OK
+> 
+
+making less sense as an error was found but was not printed. Because there are about 2 million tests in the full testset the same errors will be found many times. To limit noise an error only gets printed first time its seen. Settings for printing errors can be found in the config section of run.js

package.json

@@ -26,6 +26,7 @@
     "comparray": "^1.1.2",
     "md5": "^2.2.1",
     "numeral": "^1.5.3",
+    "paralleljs": "^0.2.1",
     "pegjs": "^0.9.0",
     "sql.js": "^0.2.21",
     "yargs": "^3.31.0"