wmerge.w

模拟器提供了一个简单易用的平台

\def\9#1{} % this hack is explained in CWEB manual Appendix F11

@* Introduction.  This file contains the program \.{wmerge},
which takes two or more files and merges them according
to the conventions of \.{CWEB}. Namely, it takes an ordinary \.{.w}
file and and optional \.{.ch} file and sends the corresponding
\.{.w}-style file to standard output (or to a named file), 
expanding all ``includes''
that might be specified by \.{@@i} in the original \.{.w} file.
(A more precise description appears in the section on ``command line
arguments'' below.)

@c
#include <stdio.h>
#include <stdlib.h> /* declaration of |getenv| */
#include <ctype.h> /* definition of |isalpha|, |isdigit| and so on */
@<Definitions@>@;
@<Predeclarations of functions@>@;
@<Functions@>@;
main (ac,av)
int ac; char **av;
{
  argc=ac; argv=av;
  @<Set the default options@>;
  @<Scan arguments and open output file@>;
  reset_input();
  while (get_line())
    put_line();
  fflush(out_file);
  check_complete();
  fflush(out_file);
  return wrap_up();
}

@ @<Definitions@>=
typedef short boolean;
typedef unsigned char eight_bits;
typedef char ASCII; /* type of characters inside \.{WEB} */

@ We predeclare some standard string-handling functions here instead of
including their system header files, because the names of the header files
are not as standard as the names of the functions. (There's confusion
between \.{<string.h>} and \.{<strings.h>}.)

@<Predecl...@>=
extern int strlen(); /* length of string */
extern char* strcpy(); /* copy one string to another */
extern int strncmp(); /* compare up to $n$ string characters */
extern char* strncpy(); /* copy up to $n$ string characters */

@ @<Predec...@>=

@ The lowest level of input to the \.{WEB} programs
is performed by |input_ln|, which must be told which file to read from.
The return value of |input_ln| is 1 if the read is successful and 0 if
not (generally this means the file has ended).
The characters of the next line of the file
are copied into the |buffer| array,
and the global variable |limit| is set to the first unoccupied position.
Trailing blanks are ignored. The value of |limit| must be strictly less
than |buf_size|, so that |buffer[buf_size-1]| is never filled.

Some of the routines use the fact that it is safe to refer to
|*(limit+2)| without overstepping the bounds of the array.

@d buf_size 1024

@<Definitions...@>=
ASCII buffer[buf_size]; /* where each line of input goes */
ASCII *buffer_end=buffer+buf_size-2; /* end of |buffer| */
ASCII *limit; /* points to the last character in the buffer */
ASCII *loc; /* points to the next character to be read from the buffer */

@ In the unlikely event that your standard I/O library does not
support |feof|, |getc| and |ungetc|, you may have to change things here.
@^system dependencies@>

Incidentally, here's a curious fact about \.{CWEB} for those of you
who are reading this file as an example of \.{CWEB} programming.
The file \.{stdio.h} includes a typedef for
the identifier |FILE|, which is not, strictly speaking, part of \CEE/.
It turns out \.{CWEAVE} knows that |FILE| is a reserved word (after all,
|FILE| is almost as common as |int|); indeed, \.{CWEAVE} knows all
the types of the ISO standard \CEE/ library. But
if you're using other types like {\bf caddr\_t},
@:caddr_t}{\bf caddr_t@>
which is defined in \.{/usr/include/sys/types.h}, you should let
\.{WEAVE} know that this is a type, either by including the \.{.h} file
at \.{WEB} time (saying \.{@@i /usr/include/sys/types.h}), or by
using \.{WEB}'s format command (saying \.{@@f caddr\_t int}).  Either of
these will make {\bf caddr\_t} be treated in the same way as |int|. 

@<Func...@>=
input_ln(fp) /* copies a line into |buffer| or returns 0 */
FILE *fp; /* what file to read from */
{
  register int  c=EOF; /* character read; initialized so some compilers won't complain */
  register char *k;  /* where next character goes */
  if (feof(fp)) return(0);  /* we have hit end-of-file */
  limit = k = buffer;  /* beginning of buffer */
  while (k<=buffer_end && (c=getc(fp)) != EOF && c!='\n')
    if ((*(k++) = c) != ' ') limit = k;
  if (k>buffer_end)
    if ((c=getc(fp))!=EOF && c!='\n') {
      ungetc(c,fp); loc=buffer; err_print("! Input line too long");
@.Input line too long@>
  }
  if (c==EOF && limit==buffer) return(0);  /* there was nothing after
    the last newline */
  return(1);
}

@ Now comes the problem of deciding which file to read from next.
Recall that the actual text that \.{CWEB} should process comes from two
streams: a |web_file|, which can contain possibly nested include
commands \.{@@i}, and a |change_file|, which might also contain
includes.  The |web_file| together with the currently open include
files form a stack |file|, whose names are stored in a parallel stack
|file_name|.  The boolean |changing| tells whether or not we're reading
from the |change_file|.

The line number of each open file is also kept for error reporting.

@f line x /* make |line| an unreserved word */
@d max_include_depth 10 /* maximum number of source files open
  simultaneously, not counting the change file */
@d max_file_name_length 60
@d cur_file file[include_depth] /* current file */
@d cur_file_name file_name[include_depth] /* current file name */
@d cur_line line[include_depth] /* number of current line in current file */
@d web_file file[0] /* main source file */
@d web_file_name file_name[0] /* main source file name */

@<Definitions...@>=
int include_depth; /* current level of nesting */
FILE *file[max_include_depth]; /* stack of non-change files */
FILE *change_file; /* change file */
char file_name[max_include_depth][max_file_name_length];
  /* stack of non-change file names */
char change_file_name[max_file_name_length]; /* name of change file */
char alt_web_file_name[max_file_name_length]; /* alternate name to try */
int line[max_include_depth]; /* number of current line in the stacked files */
int change_line; /* number of current line in change file */
int change_depth; /* where \.{@@y} originated during a change */
boolean input_has_ended; /* if there is no more input */
boolean changing; /* if the current line is from |change_file| */
boolean web_file_open=0; /* if the web file is being read */

@ When |changing=0|, the next line of |change_file| is kept in
|change_buffer|, for purposes of comparison with the next
line of |cur_file|. After the change file has been completely input, we
set |change_limit=change_buffer|,
so that no further matches will be made.

Here's a shorthand expression for inequality between the two lines:

@d lines_dont_match (change_limit-change_buffer != limit-buffer ||
  strncmp(buffer, change_buffer, limit-buffer))

@<Def...@>=
char change_buffer[buf_size]; /* next line of |change_file| */
char *change_limit; /* points to the last character in |change_buffer| */

@ Procedure |prime_the_change_buffer| sets |change_buffer| in
preparation for the next matching operation. Since blank lines in the change
file are not used for matching, we have
|(change_limit==change_buffer && !changing)| if and only if
the change file is exhausted. This procedure is called only when
|changing| is 1; hence error messages will be reported correctly.

@<Func...@>=
void
prime_the_change_buffer()
{
  change_limit=change_buffer; /* this value is used if the change file ends */
  @<Skip over comment lines in the change file; |return| if end of file@>;
  @<Skip to the next nonblank line; |return| if end of file@>;
  @<Move |buffer| and |limit| to |change_buffer| and |change_limit|@>;
}

@ While looking for a line that begins with \.{@@x} in the change file, we
allow lines that begin with \.{@@}, as long as they don't begin with \.{@@y},
\.{@@z} or \.{@@i} (which would probably mean that the change file is fouled up).

@<Skip over comment lines in the change file...@>=
while(1) {
  change_line++;
  if (!input_ln(change_file)) return;
  if (limit<buffer+2) continue;
  if (buffer[0]!='@@') continue;
  if (isupper(buffer[1])) buffer[1]=tolower(buffer[1]);
  if (buffer[1]=='x') break;
  if (buffer[1]=='y' || buffer[1]=='z' || buffer[1]=='i') {
    loc=buffer+2;
    err_print("! Missing @@x in change file");
@.Missing @@x...@>
  }
}

@ Here we are looking at lines following the \.{@@x}.

@<Skip to the next nonblank line...@>=
do {
  change_line++;
  if (!input_ln(change_file)) {
    err_print("! Change file ended after @@x");
@.Change file ended...@>
    return;
  }
} while (limit==buffer);

@ @<Move |buffer| and |limit| to |change_buffer| and |change_limit|@>=
{
  change_limit=change_buffer-buffer+limit;
  strncpy(change_buffer,buffer,limit-buffer+1);
}

@ The following procedure is used to see if the next change entry should
go into effect; it is called only when |changing| is 0.
The idea is to test whether or not the current
contents of |buffer| matches the current contents of |change_buffer|.
If not, there's nothing more to do; but if so, a change is called for:
All of the text down to the \.{@@y} is supposed to match. An error
message is issued if any discrepancy is found. Then the procedure
prepares to read the next line from |change_file|.

This procedure is called only when |buffer<limit|, i.e., when the
current line is nonempty.

@<Func...@>=
void
check_change() /* switches to |change_file| if the buffers match */
{
  int n=0; /* the number of discrepancies found */
  if (lines_dont_match) return;
  while (1) {
    changing=1; change_line++;
    if (!input_ln(change_file)) {
      err_print("! Change file ended before @@y");
@.Change file ended...@>
      change_limit=change_buffer; changing=0;
      return;
    }
    if (limit>buffer+1 && buffer[0]=='@@') {
      if (isupper(buffer[1])) buffer[1]=tolower(buffer[1]);
      @<If the current line starts with \.{@@y},
        report any discrepancies and |return|@>;
    }
    @<Move |buffer| and |limit|...@>;
    changing=0; cur_line++;
    while (!input_ln(cur_file)) { /* pop the stack or quit */
      if (include_depth==0) {
        err_print("! CWEB file ended during a change");
@.CWEB file ended...@>
        input_has_ended=1; return;
      }
      include_depth--; cur_line++;
    }
    if (lines_dont_match) n++;
  }
}

@ @<If the current line starts with \.{@@y}...@>=
if (buffer[1]=='x' || buffer[1]=='z') {
  loc=buffer+2; err_print("! Where is the matching @@y?");
@.Where is the match...@>
  }
else if (buffer[1]=='y') {
  if (n>0) {
    loc=buffer+2;
    fprintf(stderr,"\n! Hmm... %d ",n);
    err_print("of the preceding lines failed to match");
@.Hmm... n of the preceding...@>
  }
  change_depth=include_depth;
  return;
}

@ The |reset_input| procedure gets the program ready to read the
user's \.{WEB} input.

@<Func...@>=
void
reset_input()
{
  limit=buffer; loc=buffer+1; buffer[0]=' ';
  @<Open input files@>;
  include_depth=0; cur_line=0; change_line=0;
  change_depth=include_depth;
  changing=1; prime_the_change_buffer(); changing=!changing;
  limit=buffer; loc=buffer+1; buffer[0]=' '; input_has_ended=0;
}

@ The following code opens the input files.
@^system dependencies@>

@<Open input files@>=
if ((web_file=fopen(web_file_name,"r"))==NULL) {
  strcpy(web_file_name,alt_web_file_name);
  if ((web_file=fopen(web_file_name,"r"))==NULL)
       fatal("! Cannot open input file ", web_file_name);
}
@.Cannot open input file@>
@.Cannot open change file@>
web_file_open=1;
if ((change_file=fopen(change_file_name,"r"))==NULL)
       fatal("! Cannot open change file ", change_file_name);

@ The |get_line| procedure is called when |loc>limit|; it puts the next
line of merged input into the buffer and updates the other variables
appropriately. A space is placed at the right end of the line.
This procedure returns |!input_has_ended| because we often want to
check the value of that variable after calling the procedure.

@<Fun...@>=
get_line() /* inputs the next line */
{
  restart:
  if (changing && include_depth==change_depth)
   @<Read from |change_file| and maybe turn off |changing|@>;
  if (! changing || include_depth>change_depth) {
    @<Read from |cur_file| and maybe turn on |changing|@>;
    if (changing && include_depth==change_depth) goto restart;
  }
  loc=buffer; *limit=' ';
  if (*buffer=='@@' && (*(buffer+1)=='i' || *(buffer+1)=='I')) {
    loc=buffer+2;
    while (loc<=limit && (*loc==' '||*loc=='\t'||*loc=='"')) loc++;
    if (loc>=limit) {
      err_print("! Include file name not given");
@.Include file name ...@>
      goto restart;
    }
    if (include_depth>=max_include_depth-1) {
      err_print("! Too many nested includes");
@.Too many nested includes@>
      goto restart;
    } 
    include_depth++; /* push input stack */
    @<Try to open include file, abort push if unsuccessful, go to |restart|@>;
  }
  return (!input_has_ended);
}

void put_line()
{
  char *ptr=buffer;
  while (ptr<limit) putc(*ptr++,out_file);
  putc('\n',out_file);
}

@ When an \.{@@i} line is found in the |cur_file|, we must temporarily
stop reading it and start reading from the named include file.  The
\.{@@i} line should give a complete file name with or without
double quotes.