Upgrading APDL lexer

[germa89]

Fixing some parsing issues (dots mainly).

Upgraded theme.

/batch,list !list = elafune,
!batch 
/title,upf-py1s, test usermat.py with 3D elements
/COM, asdf.ad fasdfasf
/prep7
/upf,usermat.py
/upf, 'asdf'
/com, %asdf%
/com, $
/prep7$
/title,adfs %asdf% asdf
tb,user,1,,2
tbdata,1,1e5, 0.3    ! E, Poisson
adsf = 123/13
adf = adsfsa*adfasdf
asd= (212*12)/123
et,1,185

/eof

/com, expected results: Sx=-1000, epel_x=-1e-2
finish
/exit,nosave

Before:

After:

Close #2315

[jeanas]

This fails the CI as you need to run pytest --update-goldens tests/snippets/apdl to update the golden test output.

[germa89]

@jean-abou-samra I have implemented the new changes. Feel free to re-trigger CI.

Btw, I was trying that the keywords are only at the beginning of each line. I tried to add r"^\s*" to each hardcoded keyword, but it didn't work. Any suggestions?

[jeanas]

@germa89

Btw, I was trying that the keywords are only at the beginning of each line. I tried to add r"^\s*" to each hardcoded keyword, but it didn't work. Any suggestions?

You would need (?m) for ^ to work. However, you don't want the whitespace to be part of the keyword token. I've pushed a commit that takes one possible approach for this. Let me know if I understood this correctly. Otherwise, the rest looks fine to me.

[germa89]

You would need (?m) for ^ to work. However, you don't want the whitespace to be part of the keyword token. I've pushed a commit that takes one possible approach for this. Let me know if I understood this correctly. Otherwise, the rest looks fine to me.

Hi @jean-abou-samra, I'm not sure if I explained myself correctly. I would like that the keywords are only matched when they are at the beginning of the line.
I believe that (?m) are for setting multiline flag. I'm not caring about having ^ in the keywords.

[jeanas]

You said you tried to add r"^\s* as prefix on each keyword. I explained that one reason this does not work is that ^ only matches at the very beginning unless you use (?m).

[germa89]

You said you tried to add r"^\s* as prefix on each keyword. I explained that one reason this does not work is that ^ only matches at the very beginning unless you use (?m).

Ohhh... ok ok... Let's add it then.

[jeanas]

Nope. Because if you do words(..., prefix=r"^\s*"), it mostly works (with some additional changes), but the whitespace is included in the token.

Please, check the commit that I pushed. It's not the only approach, but it should work.

[germa89]

Nope. Because if you do words(..., prefix=r"^\s*"), it mostly works (with some additional changes), but the whitespace is included in the token.

Please, check the commit that I pushed. It's not the only approach, but it should work.

I see....

I checked the lastest commits and I'm happy with the result :). I'm happy to merge it.

Thank a lot!