Re: [Cocci] Determination of an usage statistic for macro calls “for_each…node…”

[Markus Elfring <Markus.Elfring@web.de>]

> …
> > +(
> > +for_each_node_by_name(n,e1) S
> > +|
> …
> > +|
> > +for_each_node_with_property(n,e1) S
> > +)
> …
> 
> 
> Do you indicate any occurrence frequencies or probabilities for the mentioned macro calls
> by the ordering in this disjunction for the semantic patch language?

I would like to share another source code analysis approach.
I hope that this contribution can trigger further helpful software development ideas.


@initialize:python@
@@
import sys, sqlalchemy
sys.stderr.write("\n".join( ("Using SQLAlchemy version:", sqlalchemy.__version__) ))
sys.stderr.write("\n")
from sqlalchemy import Column, Integer, String, create_engine
from sqlalchemy.ext.declarative import declarative_base
from sqlalchemy.orm import sessionmaker
engine = create_engine("sqlite:///:memory:", echo=False)
base = declarative_base()

class action(base):
   __tablename__ = "macros"
   name = Column(String, primary_key=True)
   source_file = Column(String, primary_key=True)
   macro = Column(String, primary_key=True)
   line = Column(Integer, primary_key=True)
   column = Column(Integer, primary_key=True)
   
   def __repr__(self):
      return """<action(name='%s',
source_file='%s',
macro='%s',
line='%s',
column='%s')>""" % (self.name,
                    self.source_file,
                    self.macro,
                    self.line,
                    self.column)

configured_session = sessionmaker(bind=engine)
session = configured_session()
base.metadata.create_all(engine)

def store_position(source, loop):
    """Add data to an internal table."""
    for place in source:
       entry = action(name = place.current_element,
                      source_file = place.file,
                      macro = loop,
                      line = place.line,
                      column = int(place.column) + 1)
       session.add(entry)

@find@
identifier for_loop, work;
iterator name for_each_node_by_name,
              for_each_node_by_type,
              for_each_node_with_property,
              for_each_matching_node,
              for_each_matching_node_and_match,
              for_each_compatible_node,
              for_each_child_of_node,
              for_each_available_child_of_node;
position pos;
statement s;
type t;
@@
 t work(...)
 {
 <+...
(for_each_node_by_name@for_loop@pos(...) s
|for_each_node_by_type@for_loop@pos(...) s
|for_each_matching_node@for_loop@pos(...) s
|for_each_node_with_property@for_loop@pos(...) s
|for_each_compatible_node@for_loop@pos(...) s
|for_each_matching_node_and_match@for_loop@pos(...) s
|for_each_child_of_node@for_loop@pos(...) s
|for_each_available_child_of_node@for_loop@pos(...) s
)
 ...+>
 }

@script:python collection@
fl << find.for_loop;
place << find.pos;
@@
store_position(place, fl)

@finalize:python@
@@
session.commit()
from sqlalchemy import func
entries = session.query(func.count()).select_from(action).scalar()

if entries > 0:
   from sqlalchemy.sql import literal_column
   delimiter = "|"
   sys.stdout.write(delimiter.join(['"source file"', 'macro', 'incidence']))
   sys.stdout.write("\r\n")
   mark = ['"', '', '"']
   
   for file, \
       macro, \
       incidence in session.query(action.source_file,
                                  action.macro,
                                  func.count(literal_column("*"))
                                 ) \
                           .group_by(action.source_file, action.macro) \
                           .order_by(action.source_file, func.count(literal_column("*")).desc()):
      mark[1] = file
      sys.stdout.write(delimiter.join([''.join(mark), macro, str(incidence)]))
      sys.stdout.write("\r\n")

   sys.stdout.write("=====\r\n")
   sys.stdout.write(delimiter.join(['macro', 'incidence']))
   sys.stdout.write("\r\n")

   for macro, \
       incidence in session.query(action.macro, func.count(literal_column("*"))) \
                           .group_by(action.macro) \
                           .order_by(func.count(literal_column("*")).desc()):
      sys.stdout.write(macro + delimiter + str(incidence))
      sys.stdout.write("\r\n")
else:
   sys.stderr.write("No result for this analysis!\n")


Test result:
elfring@Sonne:~/Projekte/Linux/next-patched> git checkout next-20201016 && XX=$(date) && time spatch --python $(which python3) --dir . ~/Projekte/Coccinelle/janitor/report_for_each_node_macro_calls.cocci; YY=$(date) && echo "$XX | $YY"
…
Using SQLAlchemy version:
1.3.19
518 files match
…
=====
macro|incidence
for_each_child_of_node|357
for_each_available_child_of_node|157
for_each_compatible_node|79
for_each_node_by_name|55
for_each_node_by_type|53
for_each_matching_node|22
for_each_matching_node_and_match|16
for_each_node_with_property|6

real	3m26,039s
user	2m3,453s
sys	0m5,041s
Sa 17. Okt 07:00:42 CEST 2020 | Sa 17. Okt 07:04:08 CEST 2020


Can such facts influence the specification of efficient SmPL disjunctions any more?


Would you like to look for software configuration alternatives for better parallel data processing?

Regards,
Markus


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