Hi Iñigo,
happy you made quite some progress.
If you were asking yourself why for BYTE, WORD and DWORD the driver is
returning a List ... these are considered Bit-Strings. If you want to
read/write a single byte as an integer value, please use SINT instead.
So regarding one of your questions ... could you instead of doing:
builderWriter.addItem("mivariable", "%DB20:DBB06:BYTE", 127);
Please try:
builderWriter.addItem("mivariable", "%DB20:DBB06:SINT", 127);
And if you're getting an INTERNAL_ERROR that's probably our fault. As I
mentioned, we haven't done too much writing as no one ever uses that ;-)
Chris
Am 02.07.20, 15:56 schrieb "Iñigo Angulo" <[email protected]>:
Hi,
I have been doing further tests with the S7 driver. I would like to ask you
some doubts, which i am not sure if relate to the PLC4X driver or the actual
hardware (Siemens S7-300) i am using (pretty novice in PLC data access im
afraid...)
The "Read value" tests I have done work correctly for different datatypes,
for example:
builderReader.addItem("mivariable", "%DB20:DBX5.0:BOOL"); //Single
boolean value
builderReader.addItem("mivariable", "%DB20:DBB06:BYTE"); //returns a
PlcList (Array of 8 boolean values)
builderReader.addItem("mivariable", "%DB20:DBW06:WORD"); //returns a
PlcList (Array of 16 boolean values)
builderReader.addItem("mivariable", "%DB20:DBD06:DWORD"); //returns a
PlcList (Array of 32 boolean values)
builderReader.addItem("mivariable", "%DB20:DBW06:INT"); //returns a
PlcInteger (16 bit integer (signed))
builderReader.addItem("mivariable", "%DB20:DBD06:DINT"); //returns a
PlcInteger (32 bit integer (signed))
builderReader.addItem("mivariable", "%DB20:DBD06:REAL"); //returns a
PlcFloat (32 bit IEEE 754 full precision floating point value (signed))
In the "Write value" tests, write request for all datatype seem to work
fine:
builderWriter.addItem("mivariable", "%DB20:DBX6.1:BOOL", true); //write
response code "OK"
builderWriter.addItem("mivariable", "%DB20:DBB06:BYTE", 127); //write
response code "OK"
builderWriter.addItem("mivariable", "%DB20:DBW06:WORD", 7); //write
response code "OK"
builderWriter.addItem("mivariable", "%DB20:DBD06:DWORD", 1); //write
response code "OK"
Except for the 'INT' and 'DINT' datatypes, which return an INTERNAL_ERROR
response code
builderWriter.addItem("mivariable", "%DB20:DBW06:INT", 1); //write
response code "INTERNAL_ERROR"
builderWriter.addItem("mivariable", "%DB20:DBD06:DINT", 1); //write
response code "INTERNAL_ERROR"
Is the INTERNAL_ERROR code a hardware issue? (Maybe a device configuration
problem?)
---
Also, the other thing I found in the test:
When I write a value different than BOOL, and then read the same memory
address, I dont get the actual value but a 'false' filled array. TO give some
examples,
with BOOL values:
builderWriter.addItem("mivariable", "%DB20:DBX6.1:BOOL", true);
builderReader.addItem("mivariable", "%DB20:DBX6.1:BOOL"); //return true
builderWriter.addItem("mivariable", "%DB20:DBX6.1:BOOL", false);
builderReader.addItem("mivariable", "%DB20:DBX6.1:BOOL"); //return false
The written value is "correctly updated" and read fine.
with other datatypes:
builderWriter.addItem("mivariable", "%DB20:DBB06:BYTE", 1);
builderReader.addItem("mivariable", "%DB20:DBB06:BYTE"); //returns
[false, false, false, false, false, false, false, false]
builderWriter.addItem("mivariable", "%DB20:DBB06:BYTE", 127);
builderReader.addItem("mivariable", "%DB20:DBB06:BYTE"); //returns
[false, false, false, false, false, false, false, false]
I have tried to combine write single bits and read other datatypes, which
seem to work fine:
builderWriter.addItem("mivariable", "%DB20:DBX6.1:BOOL", true);
builderReader.addItem("mivariable", "%DB20:DBB06:BYTE"); //returns
[false, false, false, false, false, false, true, false]
builderReader.addItem("mivariable", "%DB20:DBW06:INT"); //return 512
And the other way around, write a byte and read single bits
builderWriter.addItem("mivariable", "%DB20:DBB06:BYTE", 127);
builderReader.addItem("mivariable-0", "%DB20:DBX6.0:BOOL"); //returns
false
builderReader.addItem("mivariable-1", "%DB20:DBX6.1:BOOL"); //returns
false
builderReader.addItem("mivariable-2", "%DB20:DBX6.2:BOOL"); //returns
false
builderReader.addItem("mivariable-3", "%DB20:DBX6.3:BOOL"); //returns
false
builderReader.addItem("mivariable-4", "%DB20:DBX6.4:BOOL"); //returns
false
builderReader.addItem("mivariable-5", "%DB20:DBX6.5:BOOL"); //returns
false
builderReader.addItem("mivariable-6", "%DB20:DBX6.6:BOOL"); //returns
false
builderReader.addItem("mivariable-7", "%DB20:DBX6.7:BOOL"); //returns
false
Which leads me to think that could be a write issue, right?
Again I understand that these issues could be an actual hardware
configuration problem.. but any help or advice would be much appreciated
Thank you in advance!
iñigo
P.S. I attached the test classes if anyone wants to take a look.
-----------------------------------------
Iñigo Angulo
ZYLK.net :: consultoría.openSource
Ribera de Axpe, 11
Edificio A, modulo 201-203
48950 Erandio (Bizkaia)
+34 944272119
-----------------------------------------
----- Mensaje original -----
De: "Iñigo Angulo Otegui" <[email protected]>
Para: "dev" <[email protected]>
Enviados: Jueves, 2 de Julio 2020 12:45:49
Asunto: Re: S7 write test doubt
Hi Julian,
Yes, we saw the adopters page and were thinking to contribute there. We
will prepare a short text describing the work we are doing with the project and
send it to you soon.
Thanks
iñigo
-----------------------------------------
Iñigo Angulo
ZYLK.net :: consultoría.openSource
Ribera de Axpe, 11
Edificio A, modulo 201-203
48950 Erandio (Bizkaia)
+34 944272119
-----------------------------------------
----- Mensaje original -----
De: "Julian Feinauer" <[email protected]>
Para: "dev" <[email protected]>
Enviados: Jueves, 2 de Julio 2020 12:00:53
Asunto: Re: S7 write test doubt
And, as i shortly checked your company profile which looks pretty cool...
if you like it, or use it, dont forget to enter your company on the adopters
page : )
https://plc4x.apache.org/users/adopters.html
Julian
Am 02.07.20, 10:31 schrieb "Iñigo Angulo" <[email protected]>:
Hi Chris,
Yes, using the 0.8.0-SNAPSHOT with the update you did the problem is
solved now.
Thank you for the help and the fast solution!
Iñigo
-----------------------------------------
Iñigo Angulo
ZYLK.net :: consultoría.openSource
Ribera de Axpe, 11
Edificio A, modulo 201-203
48950 Erandio (Bizkaia)
+34 944272119
-----------------------------------------
----- Mensaje original -----
De: "Christofer Dutz" <[email protected]>
Para: "dev" <[email protected]>
Enviados: Miércoles, 1 de Julio 2020 13:18:34
Asunto: Re: S7 write test doubt
Hi all,
build passed locall, pushed the changes, build on ci passed .... so you
should be ready to give things a new try.
However keep in mind, if you built with 0.8.0-SNAPSHOT once today,
please be sure to do a "mvn -U install" build or you won't get the update today.
Would be cool if you could reply if this helped.
Chris
Am 01.07.20, 12:30 schrieb "Julian Feinauer"
<[email protected]>:
Thanks both of you!
As I just wrote in Slack we can only grow and improve if we get
feedback, Bug reports and usage scenarios.
And of course thanks Chris for that fast reply and fix <3
Julian
Am 01.07.20, 12:24 schrieb "Iñigo Angulo" <[email protected]>:
Hi Chris,
No problem. I will keep track of the issue in Jira, and use the
0.8.0-SNAPSHOT version.
Thank you for your help
iñigo
-----------------------------------------
Iñigo Angulo
ZYLK.net :: consultoría.openSource
Ribera de Axpe, 11
Edificio A, modulo 201-203
48950 Erandio (Bizkaia)
+34 944272119
-----------------------------------------
----- Mensaje original -----
De: "Christofer Dutz" <[email protected]>
Para: "dev" <[email protected]>
Enviados: Miércoles, 1 de Julio 2020 12:06:28
Asunto: Re: S7 write test doubt
Hi Iñigo,
I will have to ask you to wait for a little while as you
stumbled over a problem which goes a little deeper and is related to a major
refactoring we did just before Christmas. I am pretty puzzled that no one ever
noticed this, but I'll immediately start working on fixing this.
I'm tracking progress here:
https://issues.apache.org/jira/browse/PLC4X-206
If that's done I would like to ask you to use the
0.8.0-SNAPSHOT till we've released the next version of PLC4X.
Chris
Am 01.07.20, 12:00 schrieb "Iñigo Angulo" <[email protected]>:
Hi Chris,
Thanks for the fast answer!
I will keep your advice about not adding the cast to the
values, and letting the driver do its work.
If you want me to try to reproduce any other test just let
me know.
Iñigo
-----------------------------------------
Iñigo Angulo
ZYLK.net :: consultoría.openSource
Ribera de Axpe, 11
Edificio A, modulo 201-203
48950 Erandio (Bizkaia)
+34 944272119
-----------------------------------------
----- Mensaje original -----
De: "Christofer Dutz" <[email protected]>
Para: "dev" <[email protected]>
Enviados: Miércoles, 1 de Julio 2020 11:48:48
Asunto: Re: S7 write test doubt
And let me give you a little more context ... with the INT
type in S7 you happen to know that there an INT is a 16 Bit signed integer.
We wanted to abstract from that. Cause another PLC could
define INT as 32 bit integer ... So you can generally just work with any
numeric type and pass that in. The driver will inspect the value ranges for
that particular protocol and give some sensible error messages (At least it
should ;-) ).
But ... I just noticed that indeed I am seeing the same
problems you are seeing and that my suggestion just causes a different error
... I'll get working on fixing this right away.
I'm really a bit confused about this ... but it does prove
that most people just care about reading ;-)
Will keep you and the others posted on any progress.
Chris
Am 01.07.20, 11:33 schrieb "Christofer Dutz"
<[email protected]>:
Hi Iñigo,
welcome to the party ... I really hope we'll be able to
get you started asap ...
I just had a look at the code and PlcInteger does have
the constructor ...
public PlcInteger(Short value) {
super(value.intValue(), true);
}
So I'll try to reproduce the problem ... but for now
... could you just remove the cast from your items?
builderWriter.addItem("mivariable-1",
"%DB20:DBX6.7:BOOL", true);
builderWriter.addItem("mivariable-4",
"%DB20:DBW06.0:INT", 2);
builderWriter.addItem("mivariable-2",
"%DB20:DBB06:BYTE", 0);
Cause PLC4X internally already ensures everything fits
into the bounds and internally PLC4X handles the "(short) 2" as an "(int) 2"
anyway ;-)
Chris
Am 01.07.20, 11:19 schrieb "Iñigo Angulo"
<[email protected]>:
Hi,
I started doing some basic tests with the S7
protocol using a Siemens S7-300, and the latest 0.7.0 release. The tests
consist of reading and writing several DB variables. I was able to read
different datatypes, for instance:
builder.addItem("mivariable-1",
"%DB20:DBX05.0:BOOL");
builder.addItem("mivariable-4", "%DB20:DBW06:INT");
However, when I try to write the same memory
addresses I only manage to write BOOL values, for example
builderWriter.addItem("mivariable-1",
"%DB20:DBX6.7:BOOL", true); //WORKS OK
builderWriter.addItem("mivariable-4",
"%DB20:DBW06.0:INT", (short)2); // FAILS
builderWriter.addItem("mivariable-2",
"%DB20:DBB06:BYTE", (byte)0); // FAILS
I am unable to write any datatype different than
BOOL.
I am facing the following Error message:
Exception in thread "main"
org.apache.plc4x.java.api.exceptions.PlcRuntimeException: Error initializing
field class PlcInteger
at
org.apache.plc4x.java.s7.readwrite.field.S7PlcFieldHandler.internalEncodeInteger(S7PlcFieldHandler.java:394)
at
org.apache.plc4x.java.s7.readwrite.field.S7PlcFieldHandler.encodeShort(S7PlcFieldHandler.java:86)
at
org.apache.plc4x.java.spi.messages.DefaultPlcWriteRequest$Builder.lambda$build$0(DefaultPlcWriteRequest.java:270)
at
java.base/java.util.TreeMap.forEach(TreeMap.java:1002)
at
org.apache.plc4x.java.spi.messages.DefaultPlcWriteRequest$Builder.build(DefaultPlcWriteRequest.java:265)
at
net.zylk.plc4x.test.protocols.WriteS7.main(WriteS7.java:102)
Caused by: java.lang.NoSuchMethodException:
org.apache.plc4x.java.api.value.PlcInteger.<init>([Ljava.lang.Short;)
at
java.base/java.lang.Class.getConstructor0(Class.java:3349)
at
java.base/java.lang.Class.getDeclaredConstructor(Class.java:2553)
at
org.apache.plc4x.java.s7.readwrite.field.S7PlcFieldHandler.internalEncodeInteger(S7PlcFieldHandler.java:392)
... 5 more
Maybe I am misunderstanding the sintax and how the
values should be written?
---
To try to offer some more information about this
error:
Digging into that class (S7PlcFieldHandler) I saw
that there are some fields that are initialized based on the value of the
datatype
case INT:
minValue = BigInteger.valueOf(Short.MIN_VALUE);
maxValue = BigInteger.valueOf(Short.MAX_VALUE);
fieldType = PlcInteger.class;
valueType = Short[].class;
castedValues = new Short[values.length];
break;
Then, the line that throws the Exception seems to
be trying to inferr a PlcInteger class constructor based on the datatype
// Create the field item.
try {
return
fieldType.getDeclaredConstructor(valueType).newInstance(castedValues);
} catch (InstantiationException |
IllegalAccessException | InvocationTargetException | NoSuchMethodException e) {
throw new PlcRuntimeException("Error initializing
field class " + fieldType.getSimpleName(), e);
}
However there appears to be a "mismatch" between
the valueType (Short[]) and the constructor (Short) in the PlcInteger class,
which leads me to think i am missing something on the write value sintax?
Thank you in advance, any help will be welcome.
Iñigo
-----
the test class code:
public class WriteS7 {
private static final Logger _log =
LoggerFactory.getLogger(WriteS7.class);
public static void main(String[] args) throws
PlcConnectionException {
String connectionString =
"s7://10.105.143.1:102?remote-rack=0&remote-slot=0&controller-type=S7_300";
_log.info(String.format("Establishing connection to
%s", connectionString));;
try (PlcConnection plcConnection = new
PlcDriverManager().getConnection(connectionString)) {
_log.info("Conecting... " + connectionString);
if (!plcConnection.getMetadata().canWrite()) {
_log.error(String.format("Connection %s doesn't
support writing.", connectionString));
return;
}
PlcWriteRequest.Builder builderWriter = null;
builderWriter =
plcConnection.writeRequestBuilder();
// builderWriter.addItem("mivariable-26",
"%DB20:DBX6.7:BOOL", false); //OK
builderWriter.addItem("mivariable-2",
"%DB20:DBB06:BYTE", (byte)0); //FAILS
// builderWriter.addItem("mivariable-4",
"%DB20:DBW06.0:INT", (short)2); //FAILS
PlcWriteRequest writeRequest =
builderWriter.build();
PlcWriteResponse writeResponse =
writeRequest.execute().get();
GeneralFunctions.printPlc4XWriteReponse(writeResponse); //print response
}
}
}
-----------------------------------------
Iñigo Angulo
ZYLK.net :: consultoría.openSource
Ribera de Axpe, 11
Edificio A, modulo 201-203
48950 Erandio (Bizkaia)
+34 944272119
-----------------------------------------
