Usenet 1994 October/usenetsourcesnewsgroupsinfomagicoctober1994disk2.iso/unix/volume21/p2c/part11

Text File | 1990-04-05 | 36KB | 813 lines

Subject:  v21i056:  Pascal to C translator, Part11/32
Newsgroups: comp.sources.unix
Approved: [e-mail]
X-Checksum-Snefru: e506fa19 400f9003 1e369812 30e18e13

Submitted-by: Dave Gillespie <[e-mail]>
Posting-number: Volume 21, Issue 56
Archive-name: p2c/part11

#! /bin/sh
# This is a shell archive.  Remove anything before this line, then unpack
# it by saving it into a file and typing "sh file".  To overwrite existing
# files, type "sh file -c".  You can also feed this as standard input via
# unshar, or by typing "sh <file", e.g..  If this archive is complete, you
# will see the following message at the end:
#        "End of archive 11 (of 32)."
# Contents:  src/NOTES
# Wrapped by [e-mail] on Mon Mar 26 14:29:35 1990
PATH=/bin:/usr/bin:/usr/ucb ; export PATH
if test -f 'src/NOTES' -a "${1}" != "-c" ; then
  echo shar: Will not clobber existing file \"'src/NOTES'\"
else
echo shar: Extracting \"'src/NOTES'\" \(33044 characters\)
sed "s/^X//" >'src/NOTES' <<'END_OF_FILE'
X
XTHE GRAND P2C NOTES FILE:
X
XThis file contains notes to myself recording bugs, flaws, and suggested
Ximprovements to p2c.  They have roughly been separated into "old", "older",
Xand "oldest" groups.  I can't guarantee I'll do any of these.  If you do,
Xplease mail me the diffs so I can incorporate them into the next release.
XThanks!
X                        -- Dave Gillespie
X                           [e-mail]
X
X-----------------------------------------------------------------------------
X
X   The Pascal declaration
X
X      type foo = set of (one, two);
X
X   does not ever write out an "enum { one, two }" declaration in C!
X
X-----------------------------------------------------------------------------
X
X   Handling of comments that trail the ELSE clause of an IF statement
X   could use some work.
X
X-----------------------------------------------------------------------------
X
X   Technically speaking, "for byte := min to max do" is legal even
X   if min > 255, i.e., the limits only need to be in range if the
X   body of the loop executes.  Thus, FOR loops should really use a
X   shadow parameter not just when max is the constant 255, but
X   whenever min or max are not provably in byte range.
X
X-----------------------------------------------------------------------------
X
X   Have a "-M"-like mode in which FREE is suppressed altogether, useful
X   in case p2c crashes because bugs have corrupted malloc's free list.
X
X-----------------------------------------------------------------------------
X
X   For expressions building small-sets whose maximum element is <= 15,
X   use "1 << x" instead of "1L << x".
X
X-----------------------------------------------------------------------------
X
X   Handle VAX Pascal VARYING OF CHARs used as the arguments of WITH
X   statements.
X
X-----------------------------------------------------------------------------
X
X   Have a p2crc feature which identifies a given named constant as
X   being a character, not a string of length 1.  Have an option in
X   which this is the default interpretation.
X
X-----------------------------------------------------------------------------
X
X   StringTruncLimit would be helped by expr.c:strmax() interpreting
X   sprintf control strings.  For %s, use strmax of the corresponding
X   argument.  For %d, use 11, etc.  For %.10s, use min(10, strmax(arg)).
X   For %.*s, just use strmax(arg), I guess.
X
X   Have a mode in which such assignments are automatically truncated.
X
X   Perform truncation testing for non-VAR procedure arguments, too.
X
X-----------------------------------------------------------------------------
X
X   In cref.p, the "strappend(buf,#0)" statement translates into
X   "strcpy(STR1,buf); strcpy(buf,STR1)" with a warning about a null
X   character in a sprintf control string!
X
X-----------------------------------------------------------------------------
X
X   Still having problems where the opening comment of an imported
X   module's interface text is copied into the program.
X
X-----------------------------------------------------------------------------
X
X   VAX Pascal features not yet handled:
X
X      [UNSAFE] attribute is only implemented in a few situations.
X      [UNBOUND] attribute on a procedure says it doesn't need a static link.
X      [TRUNCATE] attribute on a parameter allows optional params w/no default.
X      [LIST] attribute on a parameter is like &rest in Lisp.
X      Support types like, e.g., [LONG] BOOLEAN.
X      Can assign "structurally compatible" but different record types.
X      File intrinsics need serious work, especially OPEN.
X      If a copy param is [READONLY], don't need to copy it.
X      If a procedure is [ASYNCHRONOUS], make all its variables volatile.
X      If a procedure is [NOOPTIMIZE], make all its variables volatile.
X      Provide a real implementation of BIN function and :BIN read format.
X      BIT_OFFSET and CARD intrinsics are not supported.
X
X-----------------------------------------------------------------------------
X
X   Modula-2 features not yet handled:
X
X      Local modules are faked up in a pretty superficial way.
X      WORD is compatible with both pointers and CARDINALs.
X      WORD parameters are compatible with any word-sized object.
X      ARRAY OF WORD parameters are compatible with absolutely anything.
X      Improve treatment of character strings.
X      Find manuals for real implementations of Modula-2 and implement
X         any common language extensions.
X      Fix p2c to read system.m2 instead of system.imp automatically.
X
X-----------------------------------------------------------------------------
X
X   Oregon Software Pascal features not yet handled:
X
X    procedure noioerror(var f:file);
X        Built-in.  Sets a flag on an already-open file so that
X        I/O errors record an error code rather than crashing.
X        Each file has its own error code.
X
X    function ioerror(var f:file) : boolean;
X        True if an error has occurred in the file.
X
X    function iostatus(var f:file) : integer;
X        The error code, when ioerror was true.
X
X    reset and rewrite ignore the third parameter, and allow a fourth
X    param which is an integer variable that receives a status code.
X    Without this param, open errors are fatal.  An optional param
X    may be omitted as in reset(f,'foo',,v);
X
X-----------------------------------------------------------------------------
X
X   In p_search, if a file contains const/var/type/procedure/function
X   declarations without any module declaration, surround the entire
X   file with an implicit "module <generated-name>; {PERMANENT}" ... "end.".
X   This would help the Oregon Software dialect considerably.
X
X-----------------------------------------------------------------------------
X
X   Provide an explicit IncludeFrom syntax for "no include file".
X   E.g., "IncludeFrom dos 0".
X
X-----------------------------------------------------------------------------
X
X   In docast, smallsets are converted to large sets of the requested type.
X   Wouldn't it be better to convert to a set of 0..31 of the base type?
X   This would keep foo([]), where the argument is "set of char", from
X   allocating a full 256-bit array for the temporary.
X
X-----------------------------------------------------------------------------
X
X   When initializing a constant variant record or array of same in which
X   non-first variants are initialized, create a function to do the
X   initialization, plus, for modules w/o initializers, a note to call
X   this function.  Another possibility:  Initialize the array as well as
X   possible, but leave zeros in the variant parts.  Then the function
X   has only to fix up the non-first variant fields.
X
X-----------------------------------------------------------------------------
X
X   Figure out some way to initialize packed array constants, e.g., a short
X   macro PACK4(x,y)=(((x)<<4)+(y)) which is used inside the C initializer.
X   Alternatively, implement initializer functions as above and use those.
X
X-----------------------------------------------------------------------------
X
X   How about declaring Volatile any variables local to a function which
X   are used after the first RECOVER?  GNU also suggests writing the
X   statement:  "&foo;" which will have no side effect except to make
X   foo essentially volatile, without relying on ANSI features.
X
X-----------------------------------------------------------------------------
X
X   Test the macros for GET, PUT, etc.
X
X-----------------------------------------------------------------------------
X
X   Can the #if 0'd code for strinsert in funcs.c be changed to test
X   strcpyleft?
X
X-----------------------------------------------------------------------------
X
X   Even in Ansi mode, p2c seems to be casting Anyptrs into other pointer
X   types explicitly.  This is an automatic conversion in Ansi C.
X
X-----------------------------------------------------------------------------
X
X   A Turbo typed constant or VAX initialized variable with a VarMacro
X   loses its initializer!
X
X-----------------------------------------------------------------------------
X
X   Test the ability of the parser to recover from common problems such
X   as too many/few arguments to a procedure, missing/extra semicolon, etc.
X   One major problem has been with undeclared identifiers being used as
X   type names.
X
X-----------------------------------------------------------------------------
X
X   Line breaker still needs considerable tuning!
X
X-----------------------------------------------------------------------------
X
X   How about indenting trailing comments analogously to the code:
X   Try to indent to column C+(X-Y), where C=original column number,
X   X=output indentation, Y=original input indentation.
X
X   Even fancier would be to study all the comment indentations in the
X   function or struct decl to discover if most comments are at the same
X   absolute indentation; if so, compute the average or minimum amount of
X   space preceding the comments and indent the C comments to an analogous
X   position.
X
X-----------------------------------------------------------------------------
X
X   After "type foo = bar;" variables declared as type foo are translated
X   as type bar.  Ought to assume the user has two names for a reason,
X   and copy the distinction into the C code.
X
X-----------------------------------------------------------------------------
X
X   Warn if address is taken of an arithmetic expression like "v1+1".
X   Allow user to declare certain bicalls as l-values, e.g., so that
X   LSC's topLeft and botRight macros won't generate complaints.
X
X-----------------------------------------------------------------------------
X
X   Consider changing the "language" modes into a set of p2crc files
X   which can be included to support the various modes.
X
X-----------------------------------------------------------------------------
X
X   If we exchange the THEN and ELSE parts of an IF statement, be sure
X   to exchange their comments as well!
X
X-----------------------------------------------------------------------------
X
X   How about checking for a ".p2crc" file in the user's home directory.
X
X-----------------------------------------------------------------------------
X
X   Store comments in the following situations:
X      On the first line of a record decl.
X      On the default clause of a CASE statement
X        (use same trick as for ELSE clauses).
X      On the "end" of a CASE statement.
X      On null statements.
X      Use stealcomments for, e.g., decl_comments and others.
X
X-----------------------------------------------------------------------------
X
X   Think of other formatting options for format_gen to support.
X
X-----------------------------------------------------------------------------
X
X   Consider converting gratuitous BEGIN/END pairs into gratuitous
X   { } pairs.
X
X-----------------------------------------------------------------------------
X
X   The construction "s := copy(s, 1, 3)" converts to a big mess that
X   could be simplified to "s[3] = 0".
X
X-----------------------------------------------------------------------------
X
X   Have a mode (and make it the default!) in which declarations are mixed
X   if and only if the original Pascal decls were mixed.  Simply store
X   a flag in each meaning to mark "mixed-with-preceding-meaning".
X
X-----------------------------------------------------------------------------
X
X   Have a column number at which to put names in variable and typedef
X   declarations.  Have another option to choose whether a '*' preceding
X   a name should be left- or right-justified within the skipped space:
X    int     *foo;     or
X        int *    foo;
X
X-----------------------------------------------------------------------------
X
X   Support the /*
X                *
X                */ form for multi-line comments.
X
X-----------------------------------------------------------------------------
X
X   Have an indentation parameter for the word "else" by itself on a line.
X
X-----------------------------------------------------------------------------
X
X   Have an option to use C++'s "//" comments when possible.
X   (0=never, 1=always, def=only for trailing comments.)
X
X-----------------------------------------------------------------------------
X
X   Allow real comments to come before top-of-file comments like {Language}.
X
X-----------------------------------------------------------------------------
X
X   Teach the line breaker to remove spaces around innermost operators
X   if in a crunch.
X
X-----------------------------------------------------------------------------
X
X   Is it possible that the line breaker is losing counts?  A line that
X   included lots of invisible parens converted to visible ones was
X   allowed to be suspiciously long.
X
X-----------------------------------------------------------------------------
X
X   The notation t^ where t is a text file should convert \n's to
X   spaces if necessary.
X
X-----------------------------------------------------------------------------
X
X   The assignment and type cast "f4 := tf4(i)" where type
X      "tf4 = function (i, j : integer) : str255"
X   generates something really weird.
X
X-----------------------------------------------------------------------------
X
X   The conditional expression  strsub(s,1,4) = 'Spam'
X   could be translated as      strncmp(s, "Spam", 4)
X
X-----------------------------------------------------------------------------
X
X   Consider an option which generates a "file.p2c" or "module.p2c"
X   file, that will in the future be read in by p2c as another p2crc
X   type of file, both when the module is re-translated later and when
X   it is imported.  This file would contain commands like "NoSideEffects"
X   for functions which are found to have this property, etc.
X
X-----------------------------------------------------------------------------
X
X   Extend the "file.log" or "module.log" file to contain a more detailed
X   account of the translation, including all notes and warnings which were
X   even considered.  For example, ALL calls to na_lsl with non-constant
X   shifts would be noted, even if regular notes in this case were not
X   requested.  Also, funny transformations along the lines of
X   "str[0] := chr(len)" and "ch >= #128" should be mentioned in the log.
X   How about a summary of non-default p2crc options and command-line args?
X
X-----------------------------------------------------------------------------
X
X   Create a TypeMacro analogous to FuncMacro, VarMacro, and ConstMacro.
X   Should the definition be expressed in C or Pascal notation?  (Yuck---not
X   a C type parser!)
X
X-----------------------------------------------------------------------------
X
X   In argument type promotions, should "unsigned char" be promoted to
X   "unsigned int"?
X
X-----------------------------------------------------------------------------
X
X   Turbo's FExpand translation is really weird.
X
X-----------------------------------------------------------------------------
X
X   Can we translate Erase(x) to unlink(x)?  (This could just be a FuncMacro.)
X
X-----------------------------------------------------------------------------
X
X   There should be an option that causes a type to be explicitly named,
X   even if it would not otherwise have had a typedef name.  Have a mode
X   that does this for all pointer types.
X
X-----------------------------------------------------------------------------
X
X   Make sure that the construction:  if blah then {comment} else other
X   does not rewrite to if (!blah) other; i.e., a comment in this situation
X   should generate an actual placeholder statement.  Or perhaps, a null
X   statement written explicitly by the Pascal programmer should always
X   produce a placeholder.
X
X-----------------------------------------------------------------------------
X
X   Allow the line breaker to treat a \003 as if it were a \010.  The penalty
X   should be enough less than SameIndentPenalty that same-indent cases will
X   cause the introduction of parentheses.
X
X-----------------------------------------------------------------------------
X
X   A comment of the form "{------}" where the whole comment is 78, 79 or 80
X   columns wide, should be reduced by two to take the larger C comment
X   brackets into account.  Also, "{*****}", etc.
X
X-----------------------------------------------------------------------------
X
X   There should be a mode that translates "halt" as "exit(0)", and another
X   that translates it as "exit(1)".
X
X-----------------------------------------------------------------------------
X
X   There should be a mode in which strread's "j" parameter is completely
X   ignored.  Also, in this mode, don't make a copy of the string being
X   read.
X
X-----------------------------------------------------------------------------
X
X   Is there an option that generates an fflush(stdout) after every write
X   (not writeln) statement?  It should be easy to do---the code is already
X   there to support the prompt statement.
X
X-----------------------------------------------------------------------------
X
X   Check out the Size_T_Long option; size_t appears to be int on most
X   machines, not long.
X
X-----------------------------------------------------------------------------
X
X   The type "size_t" should really be made into a separate type, with a
X   function to cast to type "size_t".  This function would always do
X   the cast unless sizeof(int) == sizeof(long), or unless the expression
X   only involves constants and objects or functions of type "size_t".
X
X-----------------------------------------------------------------------------
X
X   Finish the Turbo Pascal features (in the file turbo.imp).
X
X-----------------------------------------------------------------------------
X
X   Are there any ways to take advantage of "x ?: y" in GCC?
X   Is it worth using GCC constructor expressions for procedure variables?
X   How about generating "volatile" and "const" for suitable functions?
X      (doing this in the .h file would be very difficult...)
X   Use the "asm" notation of 5.17 to implement var x ['y'] declarations.
X
X-----------------------------------------------------------------------------
X
X   Recognize GCC extensions in pc_expr().  (By the way, remember
X   to implement += and friends in pc_expr(), too!)
X
X-----------------------------------------------------------------------------
X
X   Lightspeed C can't handle "typedef char foo[];" which arises from a
X   MAXINT-sized array type declaration.
X
X-----------------------------------------------------------------------------
X
X   "Return" and friends are only introduced once.  In code of the form:
X
X      if (!done) { foo(); }  if (!done) { bar(); }
X
X    p2c should, after patching up bar(), check if the foo() branch is
X    now also ripe for rearranging.
X
X-----------------------------------------------------------------------------
X
X   Have a global "paranoia" flag.  Default=use current defaults for other
X   options.  1=conservative defaults for other options.  0=sloppy defaults
X   for other options.
X
X-----------------------------------------------------------------------------
X
X   Rather than just generating a note, have writes of attribute characters
X   convert into calls to a "set attribute" procedure, such as nc_sethighlight.
X   Is there any way of generalizing this into something useful for
X   non-HP-Pascal-workstation users?
X
X-----------------------------------------------------------------------------
X
X   Warn when character constants which are control codes are produced.
X   (E.g., arrow keys, etc.)  Also, have an option which deletes all
X   highlighting codes from strings being output.
X
X-----------------------------------------------------------------------------
X
X   Think how nice things would be if the arithmetic routines actually
X   maintained the distinction between tp_int and tp_integer themselves,
X   so that makeexpr_longcast didn't have to second-guess them.
X
X-----------------------------------------------------------------------------
X
X   Importing FS *still* copies its "file support" comment into the importing
X   program!
X
X-----------------------------------------------------------------------------
X
X   Should parameterize those last few hard-wired names, such as "P_eoln",
X   "LONG_MAX", ... ?
X
X-----------------------------------------------------------------------------
X
X   Check if we need to cache away any more options' values, as we did for
X   VarStrings.  How about FoldConstants, SetBits, CopyStructs?
X
X
X=============================================================================
X
X
X   Support the "CSignif" option (by not generating C identifiers which
X   would not be unique if only that many characters were significant).
X
X-----------------------------------------------------------------------------
X
X   What if a procedure accesses strmax of a var-string parameter of a
X   parent procedure?  (Right now this generates a note.)
X
X-----------------------------------------------------------------------------
X
X   Handle full constructors for strings.
X   Handle small-array constants.
X
X-----------------------------------------------------------------------------
X
X   Have an option that causes ANYVAR's to be translated to void *'s.  In
X   this mode, all uses of ANYVAR variables will need to be cast to the
X   proper type, in the function body rather than in calls to the function.
X
X-----------------------------------------------------------------------------
X
X   Handle reading enums.  Add full error checking for reading booleans.
X   (And integer subranges?)
X
X-----------------------------------------------------------------------------
X
X   Support the "BigSetConst" option by creating constant arrays just as the
X   Pascal compiler does.
X
X-----------------------------------------------------------------------------
X
X   The 2^(N+1) - 2^M method for generating [M..N] is not safe if N is 31.
X   If the small-sets we are dealing with encompass the value 31 (== setbits-1)
X   then we must use the bitwise construction instead.  (Currently, the
X   translator just issues a note.)
X
X   (If N is 31, 2^32 will most likely evaluate to 0 on most machines, which
X   is the correct value.  So this is only a minor problem.)
X
X-----------------------------------------------------------------------------
X
X   Big-set constants right now are always folded.  Provide a mechanism
X   for defined set constants, say by having a #define with an argument
X   which is the name of the temporary variable to use for the set.
X
X-----------------------------------------------------------------------------
X
X   Should we convert NA_LONGWORD-type variants into C casts?
X
X-----------------------------------------------------------------------------
X
X   Are there implementations of strcpy that do not return their first
X   argument?  If so, should have an option that says so.
X
X-----------------------------------------------------------------------------
X
X   Handle absolute-addressed variables better.  For  var a[12345]:integer,
X   create an initialized int *.  For  var a['foo']:integer, create an int *
X   which is initialized to NULL and accessed by a macro which locates the
X   symbol the first time the variable is used.
X
X-----------------------------------------------------------------------------
X
X   Handle the idiom, "reset(f, name); open(f);"
X
X-----------------------------------------------------------------------------
X
X   Should have an option that lowercases all file names used in "reset",
X   "fixfname", etc.  This should be on by default.
X
X-----------------------------------------------------------------------------
X
X   Add more complete support for conformant arrays.  Specifically, support
X   non-GNU compilers by converting variable-sized array declarations into
X   pointer declarations with calls to alloca or malloc/free (what if the
X   function uses free and contains return statements)?  Also convert
X   variable-array references into explicit index arithmetic.
X
X-----------------------------------------------------------------------------
X
X   Have a mode in which the body of a TRY-RECOVER is moved out into
X   a sub-procedure all its own, communicating with the parent through
X   varstructs as usual, so that the ANSI C warning about what longjmp
X   can do to the local variables is avoided.  Alternatively, have an
X   option in which all necessary locals are declared volatile when
X   setjmps are present.
X
X-----------------------------------------------------------------------------
X
X   If a sub-procedure refers to a parent's variable with the VAX Pascal
X   [STATIC] attribute, that variable is declared "static" inside the
X   varstruct!  Need to convert it into a varref to a static variable in
X   the parent.
X
X-----------------------------------------------------------------------------
X
X   When comparing records and arrays a la UCSD Pascal, should expand
X   into an "&&" expression comparing each field individually.  (What about
X   variants?  Maybe just compare the first variant, or the tagged
X   variant.)  Probably best to write a struct-comparison function the
X   first time a given type of struct is compared; that way, the function
X   can include a complete if tree or switch statement in the case of
X   tagged unions.
X
X-----------------------------------------------------------------------------
X
X   In the checkvarchanged code, take aliasing of VAR parameters into account.
X   For example, in "procedure p(s1 : string; var s2 : string)" p2c now avoids
X   copying s1 if s1 is not changed within p, but probably should also require
X   that s2 not change, or at least that s1 have been read and used before s2
X   is changed.
X
X-----------------------------------------------------------------------------
X
X   Provide an option that tells the code generator to provide helpful
X   comments of its own when generated code may be obscure.
X
X
X=============================================================================
X
X
X   Compact the various data structures.  In particular, typical runs
X   show the majority of memory is used by Meanings and Symbols for
X   global and imported objects.
X
X-----------------------------------------------------------------------------
X
X   The program wastes memory.  Find ways to reduce memory usage, and to
X   avoid leaving dead records on the heap.  (Garbage collection?  Yuck!)
X   (Maybe GC between each function declaration would be okay.)
X
X-----------------------------------------------------------------------------
X
X   Assign better names to temporaries.  Also, could avoid making redundant
X   temporaries by generating a unique temporary every time one is needed,
X   then crunching them down at the end just before the declarations are
X   written out.  Each temporary would maintain a list of all the other
X   temporaries (of the same type) with which it would conflict.  (This
X   would avoid the current method's waste when several temps are created,
X   then most are cancelled.)  Also, note that char STR1[10], STR2[20] can
X   be considered type-compatible and merged into STR[20].
X
X-----------------------------------------------------------------------------
X
X   Don't generate _STR_xxx structure names if they aren't forward-referenced.
X
X-----------------------------------------------------------------------------
X
X   Can optimize, e.g., "strpos(a,b) = 0" to a call to strstr() or strchr(),
X   even though these ANSI functions are not Pascal-like enough to use in
X   the general case.
X
X-----------------------------------------------------------------------------
X
X   Complete the handling of "usecommas=0" mode.
X
X-----------------------------------------------------------------------------
X
X   Optimize "s := strltrim(s)", "s := strrtrim(s)", and both together.
X
X-----------------------------------------------------------------------------
X
X   Convert "(ch < 'a') or (ch > 'z')" to "!islower(ch)", and so on.
X   Also do "!islower(ch) && !isupper(ch)" => "!isalpha(ch)", etc.
X
X-----------------------------------------------------------------------------
X
X   Find other cases in which to call mixassignments().
X
X-----------------------------------------------------------------------------
X
X   The sequence:   sprintf(buf + strlen(buf), "...", ...);
X                   sprintf(buf + strlen(buf), "...", ...);
X   could be changed to a single sprintf.
X   Also, "sprintf(temp, "...%s", ..., buf); strcpy(buf, temp); (above);"
X   could be crunched down.  (This comes from strinsert, then strappend.)
X
X-----------------------------------------------------------------------------
X
X   If there is only one assignment to a structured function's return
X   variable, and that assignment is at the very end and assigns a local
X   variable to the return variable, then merge the variables.
X   (Example: name_list in netcmp's datastruct module.  RET_name_list
X   should be renamed to namestr.)
X
X-----------------------------------------------------------------------------
X
X   Have an option that causes if-then-else's to be replaced by ? :'s in
X   certain cases.  If the branches of the if are either both returns or
X   both assignments to the same variable, which has no side effects, and
X   if the whole conditional will be simple enough to fit on one line when
X   printed, then the ? : transformation is okay.
X
X-----------------------------------------------------------------------------
X
X   Have an option that makes limited use of variable initialization.
X   If the first statement of a function is an assignment of a constant
X   to a local variable, then the assignment is changed to an initialization.
X   (Non-constant initializers are probably too hard to check for safety.)
X   Should variables with initializers still be mixed?
X   (Example: valid_node_class in netcmp's datastructs module.)
X   File variable initialization is an especially good application for this.
X
X-----------------------------------------------------------------------------
X
X   Have an option that finds cases of multiple assignment.  For example:
X      a := x; b := x;   =>  a = b = x;
X      a := x; b := a;   =>  a = b = x;
X   (provided the objects in question have the same type).
X
X-----------------------------------------------------------------------------
X
X   Need an option that causes $if$'s to change to #if's, instead of being
X   evaluated at translation-time.  (This is *really hard* in general;
X   do it for some common cases, such as entire statements being commented
X   out, or fields being commented out of records.)
X
X-----------------------------------------------------------------------------
X
X   Have an option that prevents generation and inclusion of .h files for
X   modules; instead, each module would contain extern declarations for
X   the things it imports.  (Only declare names that are actually used---
X   this means the declarations will have to be emitted on a procedure-by-
X   procedure basis.)
X
X-----------------------------------------------------------------------------
X
X   Extend the ExpandIncludes option to compile include files as separate
X   stand-alone C modules.  The hard part is warning when the new module
X   uses a procedure which was declared static in an earlier C module.
X   Remember to re-emit all the proper #include's at the beginning.
X   Anything else?
X
X-----------------------------------------------------------------------------
X
X   Consider an option where the variables in a varStruct are made into
X   globals instead, or into parameters if there are few of them.
X
X-----------------------------------------------------------------------------
X
X   Perform a flow analysis after everything else is done; use this to
X   eliminate redundant checks, e.g., for nil pointers or non-open or
X   already-open files.  Also eliminate redundant "j" variables for
X   strwrite statements.
X
X-----------------------------------------------------------------------------
X
X   Need a method for simple pattern matching predicates in FuncMacros.
X   For example, allow special translations of functions where a given
X   argument is a known constant, or known to be positive, or two
X   arguments are the same, etc.
X
X-----------------------------------------------------------------------------
X
X   Have some way to provide run-time templates for fixblock/fixexpr-like
X   applications.  The user enters two C expressions A and B, possibly including
X   Prolog-like logical variables.  If fixexpr sees an expression matching A,
X   it rewrites it into the form of B.
X
X-----------------------------------------------------------------------------
X
X   Have an option to cause selected Pascal procedures or functions to be
X   expanded in-line.  Do this either by generating the keyword "inline",
X   or by doing the expansion in the translator.
X
X-----------------------------------------------------------------------------
X
X   Technically speaking, strcmp shouldn't be used to compute < and > for
X   strings on a machine with signed chars.  Should we care?
X
X-----------------------------------------------------------------------------
X
X   Have an option for creating a "display" of LINK pointers local to a
X   function.  Should only create such pointers for static levels which are
X   referred to in the function body.
X
X-----------------------------------------------------------------------------
X
X
END_OF_FILE
if test 33044 -ne `wc -c <'src/NOTES'`; then
    echo shar: \"'src/NOTES'\" unpacked with wrong size!
fi
# end of 'src/NOTES'
fi
echo shar: End of archive 11 \(of 32\).
cp /dev/null ark11isdone
MISSING=""
for I in 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 ; do
    if test ! -f ark${I}isdone ; then
    MISSING="${MISSING} ${I}"
    fi
done
if test "${MISSING}" = "" ; then
    echo You have unpacked all 32 archives.
    echo "Now see PACKNOTES and the README"
    rm -f ark[1-9]isdone ark[1-9][0-9]isdone
else
    echo You still need to unpack the following archives:
    echo "        " ${MISSING}
fi
##  End of shell archive.
exit 0