not parse negated

[hiperiondev]

Hello,
I have tried to compile the project on Fedora 35 without success.
I have isolated the VM and did a test of the LD parser. It works fine but it doesn't seem to recognize negated states.
Does it also happen in the original project?

The test is the following:

 i0/1--+--+-!i0/5---+---i0/6--+--+--+---(Q0/0 
       |  |         |         |  |  |         
 i0/2--+  |         +---i0/7--+  |  |         
       |  |         |         |  |  |         
 i0/3--+  |         +---i1/0--+  |  |         
          +-----i1/2------------+   |         
 i0/4-------------------------------+         

And the result:

0.LD  i0/4
1.OR (i1/2
2.AND (i0/3
3.OR (i0/2
4.OR (i0/1
5.) 
6.) 
7.OR (i0/6
8.AND (i1/0
9.OR (i0/7
10.OR (i0/5
11.AND (i0/3
12.OR (i0/2
13.OR (i0/1
14.) 
15.) 
16.) 
17.) 
18.) 
19.) 
20.ST  Q0/0

this is the project:
https://github.com/hiperiondev/plc-ladder

[hiperiondev]

but parse_ld_line works ok:

memset(&line, 0, sizeof(line));
line.buf = " ---!i0/5--(Q0/3 ";
result = parse_ld_line(&line);

printf("test: %d\n",result == PLC_OK);
printf("test: %d\n",line.cursor == strlen(line.buf) - 4);
printf("test: %d\n",line.status == STATUS_RESOLVED);
printf("test: %d\n",line.stmt->tag == TAG_ASSIGNMENT);
printf("test: %d\n",line.stmt->v.ass.type == LD_COIL);

printf("test: %d\n",line.stmt->v.ass.right->v.exp.op == IL_AND);
printf("test: %d\n",line.stmt->v.ass.right->v.exp.mod == IL_NEG);
printf("test: %d\n",line.stmt->v.ass.right->v.exp.a->tag == TAG_IDENTIFIER);
printf("test: %d\n",line.stmt->v.ass.right->v.exp.a->v.id.operand == OP_INPUT);
printf("test: %d\n",line.stmt->v.ass.right->v.exp.a->v.id.byte == 0);
printf("test: %d\n",line.stmt->v.ass.right->v.exp.a->v.id.bit == 5);
printf("test: %d\n",line.stmt->v.ass.right->v.exp.b == NULL);

result all 1;

[kalamara]

Thank you for spotting this, I noticed this case is not covered in the unit tests. I will add it and come back to you. ps. Isolating the vm is actually a very good idea, i was currently in the works of a fork of the PLC library that is stand alone. Best regards Στις 2022-02-14 19:58, Hiperión έγραψε:

…

but parse_ld_line works ok: memset(&line, 0, sizeof(line)); line.buf = " ---!i0/5--(Q0/3 "; result = parse_ld_line(&line); printf("test: %d\n",result == PLC_OK); printf("test: %d\n",line.cursor == strlen(line.buf) - 4); printf("test: %d\n",line.status == STATUS_RESOLVED); printf("test: %d\n",line.stmt->tag == TAG_ASSIGNMENT); printf("test: %d\n",line.stmt->v.ass.type == LD_COIL); printf("test: %d\n",line.stmt->v.ass.right->v.exp.op == IL_AND); printf("test: %d\n",line.stmt->v.ass.right->v.exp.mod == IL_NEG); printf("test: %d\n",line.stmt->v.ass.right->v.exp.a->tag == TAG_IDENTIFIER); printf("test: %d\n",line.stmt->v.ass.right->v.exp.a->v.id.operand == OP_INPUT); printf("test: %d\n",line.stmt->v.ass.right->v.exp.a->v.id.byte == 0); printf("test: %d\n",line.stmt->v.ass.right->v.exp.a->v.id.bit == 5); printf("test: %d\n",line.stmt->v.ass.right->v.exp.b == NULL); result all 1; -- Reply to this email directly, view it on GitHub [1], or unsubscribe [2]. Triage notifications on the go with GitHub Mobile for iOS [3] or Android [4]. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***> Links: ------ [1] #2 (comment) [2] https://github.com/notifications/unsubscribe-auth/ADZCTAUE445XH2X42EGUYCLU3E7EXANCNFSM5OLVJSGA [3] https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 [4] https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub

[hiperiondev]

Thanks for the reply,
I work with ESP32 and my idea is to port the VM to this processor.
Apparently the negated state is "lost" when you compile the entire program because when you compile just the line it does so correctly.

[kalamara]

Interesting, i think you will also need a hardware abstraction layer to control the I/O on the ESP32, and possibly an OS abstraction layer, or is your board running linux? There are plans of implementing such abstraction layers in the standalone PLC libary (but nothing concrete yet) Στις 2022-02-15 15:26, Hiperión έγραψε:

…

Thanks for the reply, I work with ESP32 and my idea is to port the VM to this processor. Apparently the negated state is "lost" when you compile the entire program because when you compile just the line it does so correctly. -- Reply to this email directly, view it on GitHub [1], or unsubscribe [2]. Triage notifications on the go with GitHub Mobile for iOS [3] or Android [4]. You are receiving this because you commented.Message ID: ***@***.***> Links: ------ [1] #2 (comment) [2] https://github.com/notifications/unsubscribe-auth/ADZCTAQ7YK3NFDUBUBA2BXDU3JH7XANCNFSM5OLVJSGA [3] https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 [4] https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub

[hiperiondev]

Yes, is needed a HAL, but isn't a complex task. I use IDF, then freertos

[hiperiondev]

Have you seen the mentioned error? I really don't see where the denied status disappears ... very strange!
This week I want to finish a task of another of my projects (hpsdr-p1-rpitx) but next week I want to work in this port
The compiler is a must for this project!

[kalamara]

I believe this is a printing issue rather than a parsing issue, from the look of it its probably a problem in dump_instruction().
I will confirm it with a unit test and get back to you.
ps. it will take a while as this is a very low priority project for me.
ps2. we can also continue the discussion directly via email to kalamara AT users.sourceforge.net
thanks

[hiperiondev]

Txs,
I have delete all references to VM and isolated compiler.
Compiler, VM and editor are three separate projects.
Probably tomorrow I will write an email to you.

[hiperiondev]

I have send an email to you.