First class function / callback help

[jondecker76]

There has to be something that I’m not understanding about callbacks.  I can’t find any documentation other than the fact that first class functions are supported, and numerous searches back here have yielded no explanations.

I’m trying to wrap the sqlite3 module into a much easier to use class, but I’ve met a roadblock with how to handle a callback function.

I’m currently trying to do the following (paired down for readability)

Monkey

Function CB:Int(data:Void Ptr, argc:Int, argv:char_t Ptr, azColName:char_t Ptr) Print "Inside callback function!" Return 0 End</p> <p>sqlite3_exec(db,query,CB,Null,Varptr err_msg) 'Fails at CB sqlite3_exec(db,query,Null,Null,Varptr err_msg)' Runs, but then I don't get a callback to store the retrieved data

1 2 3 4 5 6 7 8

Function CB:Int(data:Void Ptr, argc:Int, argv:char_t Ptr, azColName:char_t Ptr) Print "Inside callback function!" Return 0 End   sqlite3_exec(db,query,CB,Null,Varptr err_msg) 'Fails at CB sqlite3_exec(db,query,Null,Null,Varptr err_msg)' Runs, but then I don't get a callback to store the retrieved data

The CB implementation matches exactly what the parameter definition is expecting.  I’ve tried prefixing it with Varptr with no luck either.  So how can I use my function “CB” as a parameter to the sqlite3_exec() function?  If I change it to Null, then the function runs, so I know this is where the issue is, but can’t for the life of me figure out how to get it to work with a callback, and I’ve found no documentation at all to figure it out.

 

Thanks!

[NoOdle]

The only thing I can see that might affect it is the callback char_t params are defined as PTR PTR. Just a guess, only started playing with Monkey2 the other day and this is all new to me.

Monkey

Function sqlite3_exec:Int( sqlite3 Ptr, sql:CString, callback:Int( Void Ptr, Int, libc.char_t Ptr Ptr, libc.char_t Ptr Ptr ), Void Ptr, errmsg:libc.char_t Ptr Ptr )

1	Function sqlite3_exec:Int( sqlite3 Ptr, sql:CString, callback:Int( Void Ptr, Int, libc.char_t Ptr Ptr, libc.char_t Ptr Ptr ), Void Ptr, errmsg:libc.char_t Ptr Ptr )

[jondecker76]

Thanks for the response.  I will give that a try.

I had noticed before that a passing a Ptr with a Varptr preface would show up as ptr ptr when hovered over, so I guess I was assuming that would have been expected behavior.

Thanks again

[abakobo]

The fisrt class functions has never been a problem with callbacks, for me.

To get your ‘char_t Ptr Ptr’ you might want to use ‘VarPtr myChar_tArray.Data’ as ‘Array.Data’ is already a pointer to the first element of the array. Edit doesn’t work. I’ve put some working code on the codebox below.

Monkey

</p> <p>#Import "<libc>"</p> <p>Function Main()</p> <p> Local baseStr:="y" Local cstr:=New libc.char_t[baseStr.Length+1] '1 char larger to put the null termination char baseStr.ToCString(cstr.Data,baseStr.Length+1) 'writes the baseStr to the cstr</p> <p> Local csDat:=cstr.Data 'VarPtr cstr.Data is not accepted by compiler! & won't accept VarPtr twice neither Local chptrptr:=Varptr csDat</p> <p> Printstuff(chptrptr)</p> <p>End</p> <p>Function Printstuff(csptrptr:libc.char_t Ptr Ptr) Print "mx2 callback" Print String.FromCString(csptrptr[0]) End

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

#Import "<libc>"   Function Main() Local baseStr:="y" Local cstr:=New libc.char_t[baseStr.Length+1] '1 char larger to put the null termination char baseStr.ToCString(cstr.Data,baseStr.Length+1) 'writes the baseStr to the cstr   Local csDat:=cstr.Data 'VarPtr cstr.Data is not accepted by compiler! & won't accept VarPtr twice neither Local chptrptr:=Varptr csDat Printstuff(chptrptr) End   Function Printstuff(csptrptr:libc.char_t Ptr Ptr) Print "mx2 callback" Print String.FromCString(csptrptr[0]) End

I’ve joined a little zip/demo code using external function pointers with first class function in various ‘directions’.

Attachments:

1. C_callback_fn_ptr.zip

[jondecker76]

Thanks to both of you!

I finally have it figured out.

I do a lot of database work, and have been evaluating Monkey2 for a large project that I would like to make cross-platform.  I find the current sqlite3 library very ugly and difficult to work with, so I created a very simple wrapper class that makes things so much easier for me.  Here is the current state, though I do plan on making this more complete and turning this into a module.  Thanks again for the help!

Monkey

#Import "<sqlite>" #Import "<std>" #Import "<libc>"</p> <p>Using sqlite.. Using libc.. Using std..</p> <p>#Rem Monkeydoc SQLiteResult (convenience class) I found no good reason to create a separate SQLiteResult class when a List of String Maps contains all the functionality we should need (for now anyways). See @List for more information... TODO: Should I implement an Interface that accepts Map<String,String>,Map<String,Int>, Map<String,Bool> etc? #End Alias SQLiteResult:List<Map<String,String>></p> <p>#rem monkeydoc SQLite: A sqlite3 wrapper class for easy database interaction The SQLite class wraps the monkey2 sqlite3 module into an easy-to-use interface for interacting with sqlite3 databases. #end Class SQLite Private Field dbPath:String Field db:sqlite3 Ptr</p> <p> ' Callback function for building a @SQLiteResult (List of String Maps) Function BuildResult:Int(data:Void Ptr, argc:Int, argv:char_t Ptr Ptr, azColName:char_t Ptr Ptr) Local M:=New Map<String,String> For Local i:= 0 To argc-1 Local col:String=String.FromCString( azColName[i]) Local val:String=String.FromCString( argv[i] ) M.Add(col,val) Next Cast<SQLiteResult>(data).AddLast(M) Return 0 ' 0=continue, 1= stop after this callback... End</p> <p> Public</p> <p> #rem monkeydoc Create a new SQLite database instance Defaults to ":memory:", which is a special in-memory database Note that we must connect ceparately using @Connect() #end Method New(_dbPath:String=":memory:") dbPath=_dbPath End</p> <p> #rem monkeydoc Connect to/Create a new database Connects to the database specified in the constructor. If the database does not exist, create a new one. #end Method Connect:Bool() Local rc:=sqlite3_open( dbPath,Varptr db ) If rc<>SQLITE_OK Print "Failed to open DB: "+sqlite3_errmsg( db ) Disconnect() Return False Endif Return True End</p> <p> #rem monkeydoc Disconnect from the database Disconnect from the database specified in the constructor. You can always reconnect at any time... #end Method Disconnect() sqlite3_close(db) End</p> <p> #rem monkeydoc Run a SQL Query and returl a result Most commonly used for true "SELECT" queries. Returns a result that is a List<Map<String,String>>, Aliased as a "SQLite Result" #end Method Query:SQLiteResult(query:String) Local errMsg:char_t Ptr Local R:=New SQLiteResult</p> <p> Local rc:=sqlite3_exec(db,query,BuildResult,Cast<Void Ptr>(R),Varptr errMsg) If rc<>SQLITE_OK Print "Failed to fetch data: "+String.FromCString(errMsg) Return Null Endif</p> <p> Return R End</p> <p> #rem monkeydoc Execute a SQL statement (no result is returned) Very similar to @Query, but best suited for UPDATE, INSERT and other statments that have no need to return a result #end Method Execute:Bool(query:String) Local errMsg:char_t Ptr</p> <p> Local rc:=sqlite3_exec(db,query, Null, Null, Varptr errMsg) If rc<>SQLITE_OK Print "Failed to execute: " + String.FromCString(errMsg) Return False Endif Return True End End</p> <p>' TEST NEW SQLite Wrapper Class!</p> <p>Function Main() Local db:=New SQLite("/home/viavacavi/Dropbox/monkey2/test.sql3") db.Connect() Local q:String q = "DROP TABLE IF EXISTS Cars;" q += "CREATE TABLE Cars(Id INT, Name TEXT, Price INT);" q += "INSERT INTO Cars VALUES(1, 'Audi', 52642);" q += "INSERT INTO Cars VALUES(2, 'Mercedes', 57127);" q += "INSERT INTO Cars VALUES(3, 'Skoda', 9000);" q += "INSERT INTO Cars VALUES(4, 'Volvo', 29000);" q += "INSERT INTO Cars VALUES(5, 'Bentley', 350000);" q += "INSERT INTO Cars VALUES(6, 'Citroen', 21000);" q += "INSERT INTO Cars VALUES(7, 'Hummer', 41400);" q += "INSERT INTO Cars VALUES(8, 'Volkswagen', 21600);" db.Execute(q) db.Disconnect() ' Just testing the Disconnect() method here...</p> <p> db.Connect() ' Testing reconnecting... q="SELECT * FROM Cars" Local result:= db.Query(q)</p> <p> Print "Number of records: " + result.Count() For Local row:=Eachin result</p> <p> Print row.Get("Id") + ") " + row.Get("Name") Print "~t" + row.Get("Price") Next</p> <p> db.Disconnect() ' All done! End

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138

#Import "<sqlite>" #Import "<std>" #Import "<libc>"   Using sqlite.. Using libc.. Using std..     #Rem Monkeydoc SQLiteResult (convenience class) I found no good reason to create a separate SQLiteResult class when a List of String Maps contains all the functionality we should need (for now anyways). See @List for more information... TODO: Should I implement an Interface that accepts Map<String,String>,Map<String,Int>, Map<String,Bool> etc? #End Alias SQLiteResult:List<Map<String,String>> #rem monkeydoc SQLite: A sqlite3 wrapper class for easy database interaction The SQLite class wraps the monkey2 sqlite3 module into an easy-to-use interface for interacting with sqlite3 databases. #end Class SQLite Private Field dbPath:String Field db:sqlite3 Ptr ' Callback function for building a @SQLiteResult (List of String Maps) Function BuildResult:Int(data:Void Ptr, argc:Int, argv:char_t Ptr Ptr, azColName:char_t Ptr Ptr) Local M:=New Map<String,String> For Local i:= 0 To argc-1 Local col:String=String.FromCString( azColName[i]) Local val:String=String.FromCString( argv[i] ) M.Add(col,val) Next Cast<SQLiteResult>(data).AddLast(M) Return 0 ' 0=continue, 1= stop after this callback... End Public #rem monkeydoc Create a new SQLite database instance Defaults to ":memory:", which is a special in-memory database Note that we must connect ceparately using @Connect() #end Method New(_dbPath:String=":memory:") dbPath=_dbPath End #rem monkeydoc Connect to/Create a new database Connects to the database specified in the constructor. If the database does not exist, create a new one. #end Method Connect:Bool() Local rc:=sqlite3_open( dbPath,Varptr db ) If rc<>SQLITE_OK Print "Failed to open DB: "+sqlite3_errmsg( db ) Disconnect() Return False Endif Return True End #rem monkeydoc Disconnect from the database Disconnect from the database specified in the constructor. You can always reconnect at any time... #end Method Disconnect() sqlite3_close(db) End #rem monkeydoc Run a SQL Query and returl a result Most commonly used for true "SELECT" queries. Returns a result that is a List<Map<String,String>>, Aliased as a "SQLite Result" #end Method Query:SQLiteResult(query:String) Local errMsg:char_t Ptr Local R:=New SQLiteResult Local rc:=sqlite3_exec(db,query,BuildResult,Cast<Void Ptr>(R),Varptr errMsg) If rc<>SQLITE_OK Print "Failed to fetch data: "+String.FromCString(errMsg) Return Null Endif Return R End #rem monkeydoc Execute a SQL statement (no result is returned) Very similar to @Query, but best suited for UPDATE, INSERT and other statments that have no need to return a result #end Method Execute:Bool(query:String) Local errMsg:char_t Ptr Local rc:=sqlite3_exec(db,query, Null, Null, Varptr errMsg) If rc<>SQLITE_OK Print "Failed to execute: " + String.FromCString(errMsg) Return False Endif Return True End End           ' TEST NEW SQLite Wrapper Class!   Function Main() Local db:=New SQLite("/home/viavacavi/Dropbox/monkey2/test.sql3") db.Connect() Local q:String q = "DROP TABLE IF EXISTS Cars;" q += "CREATE TABLE Cars(Id INT, Name TEXT, Price INT);" q += "INSERT INTO Cars VALUES(1, 'Audi', 52642);" q += "INSERT INTO Cars VALUES(2, 'Mercedes', 57127);" q += "INSERT INTO Cars VALUES(3, 'Skoda', 9000);" q += "INSERT INTO Cars VALUES(4, 'Volvo', 29000);" q += "INSERT INTO Cars VALUES(5, 'Bentley', 350000);" q += "INSERT INTO Cars VALUES(6, 'Citroen', 21000);" q += "INSERT INTO Cars VALUES(7, 'Hummer', 41400);" q += "INSERT INTO Cars VALUES(8, 'Volkswagen', 21600);" db.Execute(q) db.Disconnect() ' Just testing the Disconnect() method here... db.Connect() ' Testing reconnecting... q="SELECT * FROM Cars" Local result:= db.Query(q) Print "Number of records: " + result.Count() For Local row:=Eachin result   Print row.Get("Id") + ") " + row.Get("Name") Print "~t" + row.Get("Price") Next   db.Disconnect() ' All done! End

[abakobo]

Nice!

Note that you can assign a class instance’s method to a first class function too.

[Author]

Posts