[Cocci] Experiences around program transformations (including Java)

[Cocci] Experiences around program transformations (including Java)

[Markus Elfring]

Hello,

I have taken another look also at contents from the document “Semantic Patches
for Java Program Transformation (Experience Report)”.
https://drops.dagstuhl.de/opus/volltexte/2019/10814/
https://dx.doi.org/10.4230/LIPIcs.ECOOP.2019.22

My software development attention grew then for a few implementation details.

1. Listings are presented for scripts of the semantic patch language.
   I recognised opportunities to write them in different ways.
   Would you like to discuss any variations further?

2. The comment “‘ <... ’ indicates that all occurrences should be replaced”
   was specified in the listing 22.
   How does this information fit to the description for the usage of SmPL ellipses
   (or dot variants) in the software documentation?
   https://github.com/coccinelle/coccinelle/blob/ed1eb8e06f800739d3992158d36945c0c4c6f0c7/docs/manual/cocci_syntax.tex#L694

3. I would appreciate further improvements also around the following wording
   on page 24.
   “…
   One painful aspect of working with Coccinelle4J is it does not always report
   errors when parsing SMPL in a user-friendly way. This limitation is inherent
   in yacc-like parsers.
   …”

   * How many resources will be needed to improve the situation in significant
     ways also in this area?

   * Under which circumstances will error diagnostics become more helpful?

4. How will the support evolve for the handling of classes (object-oriented
   programming) besides the restricted functionality of “Middleweight Java”?
   https://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.14.5415

Regards,
Markus
_______________________________________________
Cocci mailing list
Cocci@systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci