dparkins / language-fortran Goto Github PK
View Code? Open in Web Editor NEWSyntax highlighting for FORTRAN for atom
License: MIT License
Syntax highlighting for FORTRAN for atom
License: MIT License
I start typing the word "class" and it gets to "clas" and then atom freezes and the Atom Helper process consumes 100% cpu. I've narrowed it down to this package after disabling everything else and then trying them one by one. I've only been using Atom for a few days and it was working fine before and I was able to type the "Class" without issue and it just started.
It seems to affect all my .f90 files regardless of their directory. I did a fresh install of Atom, making sure to delete the app, preference files, hidden files, restarted my computer, etc and I'm still encountering this issue with no other non-default packages installed.
I'm running OSX 10.10.3, Atom 0.194.0 and language-fortran 1.5.1
Hello,
It's been a while I'm having trouble with auto-indent in a very long module. I didn't take the time to track back the bug in language-fortran's auto-indent. Here it is! :)
Let's type this in my favorite Atom editor:
program toto
implicit none
real :: x
real :: y ! no problem
real :: z ! this line contains the word "function" in the comment.
real :: q ! <= this is sufficient to mess the auto-indent
real :: messy
end program toto
Now, let's apply Edit > Lines > Auto-Indent ...
program toto
implicit none
real :: x
real :: y ! no problem
real :: z ! this line contains the word "function" in the comment.
real :: q ! <= this is sufficient to mess the auto-indent
real :: messy
end program toto
Now let's replace function
in the declaration of z
by anything else
and apply again Auto-Indent ...
program toto
implicit none
real :: x
real :: y ! no problem
real :: z ! this line contains the word "anything else" in the comment.
real :: q ! <= this is sufficient to mess the auto-indent
real :: messy
end program toto
When the word function is found in a line, even if it is in a comment, then auto-indent gets wrong.
Hi,
Good example being better than long sentences:
Source file has extension .f90
Also affects other intrinsics starting with c and likely to appear in first column of the file, like character
Might be related to recent issue #52
Hi, this is more a question than an issue.. I'm new to atom. I have fortran code in files with .F90 and .f90 extension, which have some C preprocessor directives. These look like
Some Fortran code...
for example. I see that code between # symbols is left of color white in atom with the language-fortran package.
So, my question is: can we highlight with a particular color (say red) all lines that start with the symbol # (i.e. preprocessor directives) and leave the fortran code inside preprocessor conditionals with its usual colors?
Thank you!
Marcos
Hi,
I'm surprised that the auto-indentation does not take into account the ampersand "&" at the end of a given line to add an additional indentation at the beginning of the next one, so as to be able to distinguish between new lines and lines being continuation of previous ones. For instance, after having key-binded tab to the auto-indent command, here is what I get when I only use tab :
Wouldn't something like this be more confortable :
?
Could you publish this package so that it is downloadable from apm?
comments like
cxx
xx would make it not to detect as comment.
Consider the following code snippet:
function hello()
IF ( (condition1) .AND. &
(condition2) ) THEN
CALL DUMMY (1, 32544, &
'DUMMY1', &
'DUMMY2')
END IF
end function hello
function hello()
IF ( (condition1) .AND. (condition2) ) THEN
CALL DUMMY (1, 32544, &
'DUMMY1', &
'DUMMY2')
END IF
end function hello
as can be seen from the attached figure, the if-then block does not work correctly when the condition is split over multiple lines.
Considering the the 'begin'
match for the if-then-construct`` in https://github.com/dparkins/language-fortran/blob/master/grammars/fortran%20-%20punchcard.cson where
(?=[^\n]*\bthen)``` is part of the match, would this net not exclude new-lines?
I am still new at the definitions of grammars in this way, but by replacing the remainder of the match with a .*?
the syntax highlighting is corrected.
Just noticed that comments aren't working in derived type definition constructs. The fix for this is easy and I've already done/tested it on my local machine. I'll upload a fix tonight.
For some reason attributes in type definition statements are being arbitrarily skipped in some cases as shown below
This bug was not present in earlier versions of Atom and I believe it's a bug in first-mate possibly related to this issue. I'll see what I can do about fixing this but we may have to wait for a newer release of Atom to fix it.
I'm trying to open an arguably large source code (~5000 lines of code) and the syntax highlighting makes Atom slower and slower until freezing completely the more I scroll down in the file.
I've tried disabling the syntax highlighting to make sure it was indeed the issue, and it works perfectly fine without syntax highlighting.
If you want to try with the same file, you can download this archive http://s-provencher.com/pub/contin/contin.for.gz that contains the .for file that I was working on.
Store package settings files in the settings/
directory instead of scoped-properties/
Package.loadSettings (/usr/share/atom/resources/app.asar/src/package.js:584:9)
<unknown> (/usr/share/atom/resources/app.asar/src/package.js:186:43)
Package.measure (/usr/share/atom/resources/app.asar/src/package.js:165:15)
Package.load (/usr/share/atom/resources/app.asar/src/package.js:179:12)
PackageManager.loadPackage (/usr/share/atom/resources/app.asar/src/package-manager.js:372:14)
PackageManager.loadPackages (/usr/share/atom/resources/app.asar/src/package-manager.js:336:14)
There seems to be an issue with loop indentation. In emacs:
do i = 1, Ntot
...
end do
In atom using this package:
do i = 1, Ntot
...
end do
This seems to be at least for if and do loops. Not sure if it effects anything else.
Howdy @dparkins,
was just messing around with some c-bindings in atom and noticed that this package doesn't pick up the bind
statement as needing some tlc. See below:
module example
use iso_c_binding, only: c_double
implicit none
contains
subroutine square(a, out) bind(c) <-- See here
real(c_double), intent(in) :: a
real(c_double), intent(out) :: out
out = a*a
return
end subroutine square
end module example
I'm not sure how this language is structured - if it is only a part of the modern grammar or if you need to include in all of them. I think we it should be included only in the modern grammar because the c-binding was added in F2003.
Cheers!
C
What's the way to indicate how toggle comment should work (Command-/) in fixed-form source files? Currently, Command-/ inserts C-style comments. I tried adding this to my config.cson
:
".fixed.fortran.source":
editor:
commentStart: "^C "
commentEnd: ""
but clearly, that's not the way to insert a C
in column 1... Any ideas?
See the attached image, where the function block is prematurely terminated due the presence of a local variable an_end
.
Looking at the match rules in https://github.com/dparkins/language-fortran/blob/master/grammars/fortran%20-%20punchcard.cson, it would seem as if the end rule for the function-definition
(and other blocks, program
, subroutine
, etc.) should include ^\\s*
before the capture of (end)
. For example:
'end': '(?i)^\\s*(end)(?:\\s*(\\1)(\\s+\\2)?\\b)?\\s*([^;!\\n]+)?'
The Atom.io deprecation cop shows the following deprecation warning:
Store package settings files in the settings/ directory instead of scoped-properties/
Called 2 times:
Package.loadSettings - C:\Users\lezar\AppData\Local\atom\app-0.189.0\resources\app\src\package.js:565:9
<unknown> - C:\Users\lezar\AppData\Local\atom\app-0.189.0\resources\app\src\package.js:187:43
I'm opening a file with .f
extension, the grammer is "Fortran - Fixed Form". There are several #include
statements which are not highlighted correctly. Also, when adding e.g. #ifdef
statements they are not highlighted correctly either.
atom 1.3.3, language-fortran 2.0.8
Hi there :)
The error stop
statement of Fortran 2008 is not detected.
http://fortranwiki.org/fortran/show/Fortran+2008
Thanks again for your time and useful work!
"Go to Declaration" does not work
I'm having some problems with the 'do concurrent' statement
do concurrent (i=1:n)
! code here
end do
the 'concurrent' part has no highlight.
I'm also having problem with the following lines:
interface setGaussLegendre
module procedure setGaussLegendre01, setGaussLegendreAB
end interface setGaussLegendre
I always get a highlighted 'end' inside the 'setGaussLegendre01'.
Thanks!
I just noticed there is a change log file in the main directory. Sadly it stops at v0.3.0. I feel it's time to update it so I'm putting this here mostly as a note to myself to do so. Of course if anyone else wants to comb through my old commit messages they're more than welcome to! ๐
The current rules for do-constructs work perfectly well for standard modern Fortran syntax but there are two use cases missing from the older punchcard Fortran (Fortran 77) standards.
In most cases do-constructs are writen like
do i = 1, 10, 1
...
end do
However, one could also replace the enddo
statement with a continue
statement
do i = 1, 10, 1
...
continue
and still be correct. This change on it's own would be rather trivial to make in the rules however there is one last form of the do-construct that complicates things significantly.
The previous two examples are both forms of un-labeled do-constructs. A do-construct can also be written as a labeled do-construct, such as:
do 99 i = 1, 10, 1
...
99 k = k + i
When written in this format the line terminating the loop can be almost any executable statement (including continue
and end do
). Furthermore multiple labeled do-constructs can terminate on the same line. For instance
do 99 i = 1, 10, 1
do 99 j = 2, 20, 2
...
99 k = k + i
where both the inner and outer loop terminate on line 99
.
I've spent a little time toying around with fixed for this but I haven't come up with anything particularly satisfying that didn't involved abandoning the more structured rules approach (i.e., getting rid of the multi-line do-construct rules and replacing them with the old do/enddo keywords.
In the Fortran code that I develop, we use our own custom preprocessor, and I would like to extend the language-fortran syntax to add support for this, as well as our custom file extensions.
I have tried the simple "include" solution given here: http://stackoverflow.com/questions/22363070/how-do-i-make-a-default-syntax-by-filetype-in-atom-text-editor, but in this case, no highlighting works -- even if I don't yet add the support for the preprocessor directives.
What would be the correct way to go about this?
Using an exclamation (!) character to comment code does not highlight the code as comment
Hi there, once again : a big thank for all the efforts :), especially to you @tomedunn !
Modules names should be coloured, shouldn't they?
I have found that if I have a line like this in my code (the third one in the code snippet)
implicit none
integer, parameter :: MAXITER=20,MAXJ=5
dimension g(MAXJ+1)
where the variable g has no assigned type, the atom editor will just hang, and will require a restart. It is very bad coding form, but I would like it if the formatting package didn't crash the editor. Any suggestions?
In following code
do i = 1, 5
end do
do not getting highlighted
This was discussed in issue #57 which is now closed so I wanted to give it it's own issue. To summarize, subroutines called via the call statement are highlighted whether they are intrinsic or not with the exception of type bound subroutines as illustrated below
Frankly I'm not sure if it's possible to make a rule that will correctly highlight these without also highlighting the variable part as well. But for now I'll just leave this issue open as a reminder.
Atom correctly indents different statements within a module or subroutine as expected. The contain
statement is also indented at the level of the procedure it is contained, which I didn't expect but I actually really like:
The issue is, when folding code that contains a contain
statement, the folding stops at the contain
statement instead of down to where it should fold at end module
. This seems to me like atom is repeating some behavior I would expect from python, where everything is progressively nested; however in Fortran that isn't the case.
See gif:
The following code is is for an enum
, which is not detected/highlighted by the grammar:
!-----------------------------------------------------------------------------------------
! Enum type definitions
!-----------------------------------------------------------------------------------------
! Return codes
enum, bind(c)
enumerator :: &
METIS_OK = 1, & ! Returned normally
METIS_ERROR_INPUT = -2, & ! Returned due to erroneous inputs and/or options
METIS_ERROR_MEMORY = -3, & ! Returned due to insufficient memory
METIS_ERROR = -4 ! Some other errors
end enum
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.