writing a few lines of comments that work in most languages

chrism01 · 04-04-2023, 12:04 AM

Quote:

PAN64

But you can almost never translate a real source code from one language to another, there is no such tool. And because of the complexity (and incompatibility) of the existing languages it will never exist.

Actually, back in the days of the Y2K project, there were some one-to-one tools.

I ended up trying a FORTRAN to C tool to try and save time.
The code prob worked (I can't remember), but it was very odd looking 'C' of course ...
Iirc in the end I did re-code it into C by hand, but it was handy to have a 'C' version as an extra reference ...

pan64 · 04-04-2023, 01:31 AM

Quote:

Originally Posted by chrism01

Actually, back in the days of the Y2K project, there were some one-to-one tools.

I ended up trying a FORTRAN to C tool to try and save time.
The code prob worked (I can't remember), but it was very odd looking 'C' of course ...
Iirc in the end I did re-code it into C by hand, but it was handy to a 'C' version as an extra reference ...

Oh yes, there are examples like this, but there is no general "translator". Interestingly the gnu compiler set (and probably other compilers) will produce byte code [during the compilation] which is quite similar (or identical) for every language handled by the tool, but the reverse direction does not work very well (reconstructing the original source files).
By the way using those "reasonably arbitrary text as comments" would mean just a new language - which can be easily converted to some(?) other languages.

Skaperen · 04-06-2023, 08:13 PM

my purpose is to attach a code or certificate to each file. i probably will need to create something different for each class of language (classed by comment syntax) that reads a common form of the text/cert (probably 8 or so lines of 64 chars). python and bash would be in one class with a bunch of others. C would be in another with some. Fortran might have is own. same for COBOL.

NevemTeve · 04-07-2023, 01:07 AM

Maybe you mean `signature`?

dugan · 04-07-2023, 01:15 PM

You'd want to store stuff like that in the filesystem, not in the file. Here's one way you can do it:

https://learn.microsoft.com/en-us/wi...reparse-points

Of course, it's important what you want that "code or certificate" to say say/do. Are you withholding that information, or do you not know the answer yourself?

Before you answer, read the following very carefully:

https://xyproblem.info/

Skaperen · 04-08-2023, 02:37 PM

the current need is to give a file to other people for them to add to each file. it can be added anywhere in their file, and the process using it will work. they may have lots of files of many languages to add it to. i do not want to burden them with the need to figure out which of many variants for many languages they need to use with that file, which maybe is too hard for them to automate.

i do not believe a universal tool exists for that purpose. for example, when using the "file" utility, many of my Python source files are typed as "CLIPPER instruction trace" (no idea what CLIPPER is). a nice tool would be able to read a file and give the correct language it is (with its common name extension) or list of possible languages it could be. maybe some new AI could do that (what would we train it with?).

Skaperen · 04-08-2023, 02:40 PM

Quote:

Originally Posted by NevemTeve

Maybe you mean `signature`?

for what?

Skaperen · 04-08-2023, 02:55 PM

here is a string to attach to every source file involved:

Code:

voodoo code = 4e3c1f5b5e617112b85872c4243a60f927e528023a0577f76f1a190c3b12668ef3a073f7ad7ff50393dd98d6bbe81e425aaa2753ee23d92a9d7eec08285ba3f041b13fcea40a16bf29e2fd73b7b2f35c09db44a592f77300b0e2895be19482d1abc747d64e9f695e6580f3d7cc8eda1471974d29077ddc498af4b83d6adc56bc5cc0c6fb091dd532ef8cbea668297ae8

if you can embed this string inside of another such that this string can still be seen, but will look like something that does not change the working of the source file when compiled (such as a comment or discarded literal) then that might be the solution. as you can see, this will be very difficult, if not impossible.

C:

Code:

/* voodoo code = 4e3c1f5b5e617112b85872c4243a60f927e528023a0577f76f1a190c3b12668ef3a073f7ad7ff50393dd98d6bbe81e425aaa2753ee23d92a9d7eec08285ba3f041b13fcea40a16bf29e2fd73b7b2f35c09db44a592f77300b0e2895be19482d1abc747d64e9f695e6580f3d7cc8eda1471974d29077ddc498af4b83d6adc56bc5cc0c6fb091dd532ef8cbea668297ae8 */

C++ or C:

Code:

// voodoo code = 4e3c1f5b5e617112b85872c4243a60f927e528023a0577f76f1a190c3b12668ef3a073f7ad7ff50393dd98d6bbe81e425aaa2753ee23d92a9d7eec08285ba3f041b13fcea40a16bf29e2fd73b7b2f35c09db44a592f77300b0e2895be19482d1abc747d64e9f695e6580f3d7cc8eda1471974d29077ddc498af4b83d6adc56bc5cc0c6fb091dd532ef8cbea668297ae8

Python or Bash:

Code:

# voodoo code = 4e3c1f5b5e617112b85872c4243a60f927e528023a0577f76f1a190c3b12668ef3a073f7ad7ff50393dd98d6bbe81e425aaa2753ee23d92a9d7eec08285ba3f041b13fcea40a16bf29e2fd73b7b2f35c09db44a592f77300b0e2895be19482d1abc747d64e9f695e6580f3d7cc8eda1471974d29077ddc498af4b83d6adc56bc5cc0c6fb091dd532ef8cbea668297ae8

Fortran:

Code:

C voodoo code = 4e3c1f5b5e617112b85872c4243a60f927e528023a0577f76f1a190c3b12668ef3a073f7ad7ff50393dd98d6bbe81e425aaa2753ee23d92a9d7eec08285ba3f041b13fcea40a16bf29e2fd73b7b2f35c09db44a592f77300b0e2895be19482d1abc747d64e9f695e6580f3d7cc8eda1471974d29077ddc498af4b83d6adc56bc5cc0c6fb091dd532ef8cbea668297ae8

dugan · 04-08-2023, 02:56 PM

How would that voodoo code be generated, what information would it contain, and what would read it?

I see that in your post, the voodoo code is the same every time. If it's just an identifier, and the the idea is to locate files that contain it, then I have a better idea. Create a directory named after the voodoo code, and symlink all of the source files into it.

pan64 · 04-09-2023, 02:23 AM

Quote:

Originally Posted by Skaperen

here is a string to attach to every source file involved:

Code:

voodoo code = 4e3c1f5b5e617112b85872c4243a60f927e528023a0577f76f1a190c3b12668ef3a073f7ad7ff50393dd98d6bbe81e425aaa2753ee23d92a9d7eec08285ba3f041b13fcea40a16bf29e2fd73b7b2f35c09db44a592f77300b0e2895be19482d1abc747d64e9f695e6580f3d7cc8eda1471974d29077ddc498af4b83d6adc56bc5cc0c6fb091dd532ef8cbea668297ae8

if you can embed this string inside of another such that this string can still be seen, but will look like something that does not change the working of the source file when compiled (such as a comment or discarded literal) then that might be the solution. as you can see, this will be very difficult, if not impossible.

I still have no idea what do you want to solve and how will it solve anything. Obviously we can embed [almost] any string into any language.

dugan · 04-09-2023, 11:19 AM

Quote:

Originally Posted by Skaperen

the current need is to give a file to other people for them to add to each file. it can be added anywhere in their file, and the process using it will work. they may have lots of files of many languages to add it to. i do not want to burden them with the need to figure out which of many variants for many languages they need to use with that file, which maybe is too hard for them to automate.

So you're intending to have source code files edited by people who don't even know how to write comments in those languages?

Silliness of this aside, it's standard practice to have a Continuous Integration server that builds the source code and ensures that it's at least is still syntactically correct. Yes this is also (largely) true for interpreted languages.

dugan · 04-10-2023, 12:53 PM

Why can't you just do a separate "echo >>" operation for each file extension?

Why would you have people do it with a text editor?

dugan · 04-10-2023, 01:16 PM

Deleted. Oops. Wrong thread. Carry on.

MadeInGermany · 04-17-2023, 11:03 AM

You can assume that all interpreted languages support #comment, in order to ignore a #! shebang.

Code:

#!/bin/sh
voodoo="4e3c1f5b5e617112b85872c4243a60f927e528023a0577f76f1a190c3b12668ef3a073f7ad7ff50393dd98d6bbe81e425aaa2753ee23d92a9d7eec08285ba3f041b13fcea40a16bf29e2fd73b7b2f35c09db44a592f77300b0e2895be19482d1abc747d64e9f695e6580f3d7cc8eda1471974d29077ddc498af4b83d6adc56bc5cc0c6fb091dd532ef8cbea668297ae8"
bv=" $voodoo"
for fn
do
  fgrep -q "$bv" "$fn" ||
  # only if not yet found
  case $fn in
  ( *.c ) echo "/*$bv */"
  ;;
  ( *.cpp ) echo "//$bv"
  ;;
  ( *.f ) echo "c   $bv"
  ;;
  ( * ) echo "#$bv"
  esac >> "$fn"
done

pan64 · 04-17-2023, 11:15 AM

Quote:

Originally Posted by MadeInGermany

You can assume that all interpreted languages support #comment, in order to ignore a #! shebang.

Code:

#!/bin/sh
voodoo="4e3c1f5b5e617112b85872c4243a60f927e528023a0577f76f1a190c3b12668ef3a073f7ad7ff50393dd98d6bbe81e425aaa2753ee23d92a9d7eec08285ba3f041b13fcea40a16bf29e2fd73b7b2f35c09db44a592f77300b0e2895be19482d1abc747d64e9f695e6580f3d7cc8eda1471974d29077ddc498af4b83d6adc56bc5cc0c6fb091dd532ef8cbea668297ae8"
bv=" $voodoo"
for fn
do
  fgrep -q "$bv" "$fn" ||
  # only if not yet found
  case $fn in
  ( *.c ) echo "/*$bv */"
  ;;
  ( *.cpp ) echo "//$bv"
  ;;
  ( *.f ) echo "c   $bv"
  ;;
  ( * ) echo "#$bv"
  esac >> "$fn"
done

and what about perl, php, bash, java, js, cobol, fortran, assembler, python, basic, pascal (just to name a few)
https://www.simplilearn.com/best-pro...-today-article
(nightmare)