[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[GSoC] Parsing using Abstract Syntax Tree

From: Vishal Gupta
Subject: [GSoC] Parsing using Abstract Syntax Tree
Date: Thu, 19 Jul 2018 17:12:07 +0530

In regard to my GSoC project [ Branch : experimental/gsoc/ast ] , I
have some queries :

1)  [ ]
 if  TWO
  test1_LDADD = two.$(OBJEXT)
  test2_LDADD = two.$(OBJEXT)
  test2_DEPENDENCIES = $(test2_LDADD) somethingelse.a
else  !TWO
  test1_LDADD = one.$(OBJEXT)
  test2_LDADD = three.$(OBJEXT)
endif  !TWO

Meaning of the above conditional statement, specifically why is their
a requirement of condition after else/endif and what should be its
expected value ?

2) The parser which I have implemented assumes that their will be a
newline character before EOF character in . Is this
assumption correct ?

3) SUBDIRS variable stores the subdirectories which are to be built.
This feature is implemented in parser by calling the "system" function
call and changing the input and output files to corresponding
subdirectories. [ system( "perl -I.  $val/ >
$val/ast.gv" ); ]

present working directory:-
 |- t
    |- app

t/ contents =>
SUBDIRS = app file parses present in "t" directory. Executing
test,sh on "t/", the system call for subdirs, as mentioned
above, checks for app directory in "present working directory" instead
of "t" directory. This is the expected behavior because perl is
running from " present working directory " but the app directory is
inside the "t" directory. How to overcome this ? Moving the
file into "t" directory, "" file is unavailable and using
"perl ../" for execution, then rest of the modules like, are unavailable.

Vishal Gupta

reply via email to

[Prev in Thread] Current Thread [Next in Thread]