rsalz@uunet.uu.net (Rich Salz) (03/28/90)
Submitted-by: Dave Gillespie <daveg@csvax.caltech.edu>
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 rsalz@litchi.bbn.com 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 daveg@csvax.caltech.edu
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
--
Please send comp.sources.unix-related mail to rsalz@uunet.uu.net.
Use a domain-based address or give alternate paths, or you may lose out.