Outputting comments to NC file. (post)

[Post dept]

Avs_Fan,

 

Not sure why it's not working for you. I think it's because your sof = 1 (not start of file anymore) flag is in the wrong spot. It's jsut going through and outputing everything on the first pass through the pcomment postblock. It's using rc6 as an automatically incrementing counter so it shouldn't output anything twice.

 

Roger,

 

I'm learning a lot from you here. I didn't know about the 'comm_cnt = 0' command. Nor did I know about the pscomm0 postblock. This will come in handy. Also, I like the way you check through the whole buffer and set a flag if the descriptor was actually used.

 

Brett

[Avs_Fan]

Hi All,

 

Brett, after seeing the very detailed response from Roger I thought I would try that before pursuing the sof=1 placement. Thanks for replying!

 

Roger, I tried your sample posts with some of our files & they all do just about what we want. I now need to copy the changes to one of our posts, tweak placement of comments and test. Maybe a day or 2.

This response was awesome! Thank You!

[Avs_Fan]

Mr. Roger Martin,

 

Did the paste into my post & some tweaks.

THIS IS IT! HOORAH! IT WORKS!

We've been fighting this off & on for over a year.

THANK YOU! THANK YOU! THANK YOU!

 

Should I refer to you in the future as "Saint" Roger?

[crazy^millman]

To all of you that have looked at this Thread a good bunch of things ot add to your Notebook. I have taken 6 sections from this even though I posted no repsonse in this till now. Thank you to Brent and Roger for Sharing it is appericated.

[KevinPro61]

Thanks again to all involved but I seem to be the only one that still hasn't figured it out. I get a variable error when I run the posts that Roger provided.

 

"Variable not defined: comm_cnt

Post line number 742

Program execution halted due to error(s) in .pst"

 

I am not sure what to do now, I seems that Avs_fan does not have the same problems that I have.

 

[crazy^millman]

Kevin this varabile needs to be defined at the top of the post where the other varailbe are defined. Here it is taking from Rogers Martins Psoting in this Thread:

code:

---

comm_cnt = 0  

---

Thne put it in this part of the post:

code:

---

#Misc variables

tcnt : 0 # Count the number of tool changes

toolcount : 1 # Tool counter for logic control

toolcountn : 0 # Next tool counter for logic control

toolcountp : 0 # Previous tool counter for logic control

tooltotal : 0 # Tool tool count

nexttool : 0 # Next actual tool

---

I hope that helps.

[KevinPro61]

Sorry to keep bothering you guy on this but here is the thing. I downloaded Roger's post and ran it with out touching it and I get the same error. Acording to Avs_fan when he ran them they coded fine? I think the problem might be some where else.

[Roger Martin from CNC Software]

quote:

---

"Variable not defined: comm_cnt

Post line number 742

Program execution halted due to error(s) in .pst"

---

COMM_CNT is a pre-defined MP variable.

You would not need to declare it.

 

Contact your Dealer to upgrade your MC.

[crazy^millman]

Ok well thanks for the clairfication. Sorry to steer you astray on that. Good luck and it does sound like you need to upgrade.

[Avs_Fan]

Kevin,

Yes, Roger's posts worked for us "as is". After adding the changes to one of ours I got a similar error to yours except it was the variable "one".

 

I added this in the same place Millman shows.

code:

---

one         : 1     #Define constant

---

That got rid of the error.

For reference, we used the "2-buffer" method.

Also for reference, we have V9.1 SP2 Mill Level 3 Solids with current maintenance.

Hope this helps.

[Roger Martin from CNC Software]

Millman,

 

Adding the COMM_CNT : 0 would make that error disappear, but the output will not be correct.

 

AVS,

I thought that I'd changed all those 'one', and 'two', references.

 

code:

---

 pscomm0  # Comment "pre-read" posblock

if sof = yes,

scomm0 = wbuf(one, wc1) # Save the string to Buffer 1

---

The 'one' in the above line is the reason for this in the 2-Buffer samples.

You can add the definition for 'one' like you did -or- change the (one, wc1) to-> (1, wc1)

 

I'm used to using the 'words' since they are defined in the MPFAN & MPMASTER type posts.

 

code:

---

m_one       : -1    #Define constant

zero : 0 #Define constant

one : 1 #Define constant

two : 2 #Define constant

three : 3 #Define constant

four : 4 #Define constant

five : 5 #Define constant

c9k : 9999 #Define constant

---

Then I realized that people may be plugging this code into PSTs that don't have these definitions and changed 'em all back to digits.

Missed that 'one'

[Avs_Fan]

Hi Roger,

Just a follow up. I found that this line also has to be changed from rc1 = one to rc1 = 1 or you will get the error during posting mentioned in an earlier thread.

code:

---

         rc1 = 1  # reset the read pointer to the start of Buffer 1

---

[Post dept]

Roger,

 

Thanks for the buffer methods. These will come in handy. I prefer the single buffer method. When testing the 2 buffer method I found that if I did not output a file descriptor I did not get the first op comment to appear. I guess it's only writing to the second buffer if it sees the end flag. Works great....thanks again.

 

Brett

[Roger Martin from CNC Software]

Ah... I see what you described.

 

I know that using the 2-Buffer method, if there where FD comments output to the post, you did not need to [/i]output[/i] the FD comment to the NC file.

Or you could output the FD comment anywhere you wished in the NC file.

I know I tested that!

But if no FD comments were in the MC9 file or you did not output them to the PST, then you got no 1st OP comment(s).

 

A lesson to be learned here... Don't do post work late at night while watching TV

 

Here is an easy fix for the 2-Buffer method.

Alter the "poutput_comments" postblock to be this ->

code:

---

poutput_comments  # Output any SOF Comments 'other' than any FD commnets

if EndMarkerIndex > 0, # Comments AFTER the File Descriptor were saved to Buffer 2

[

size2 = rbuf(2, 0) # Get the size of Buffer 2

while (rc2 <= size2),

[

stemp = rbuf(2, rc2)

"(", stemp, ")", e # Output any "non" FD comments

]

]

else, # There wwere NO FD comments, so any 'other' comments are still in Buffer 1

[

rc1 = 1 # Reset the read index to the start of Buffer 1

size1 = rbuf(1, 0) # Get the size of Buffer 1

while (rc1 <= size1),

[

stemp = rbuf(1, rc1)

"(", stemp, ")", e # Output any "non" FD comments

]

]

---

[crazy^millman]

Well thanks for all that I sure learned alot. I now have 7 more pages ot add to my ever growing Mastercam help book. Thanks you guys for askig nthe question and thank you Roger for clearing alot of that up.

[Avs_Fan]

Hi All,

 

Just some (hopefully) useful follow-up on this thread.

 

I encounter a problem when applying Roger's code (2-buffer) to one of our posts that was modified from MPYASNAK.PST. It was already using the rc1 & wc1 strings for some tool change & staging. I just edited Roger's to rc3 & wc3 and also changed rc2 & wc2 to rc4 & wc4 to keep them in sequence.

Worked like a charm!

Thanks again Roger

[cnc-guy]

Update note for version X2:

 

change the following:

 

pscomm0 --> pscomm0$

yes --> yes$

no --> no$

scomm0 --> scomm0$

comm_cnt --> comm_cnt$

e --> e$

 

add:

stemp = ucase(stemp) #Capitalize string output

 

Thanks Roger.

[cnc-guy]

Roger,

 

I've tried outputting the FD inside a set file without success.

 

Is there something that should be looked at.

 

Thanks