Remove -e flag, remove long options, start rewriting in C, add support for changing shell

Signed-off-by: Mattias Andrée <maandree@kth.se>

7 files changed, 481 insertions, 87 deletions

diff --git a/.gitignore b/.gitignore
index df7be57..4240069 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,16 +1,5 @@
-_/
-/bin/
-/obj/
-__pycache__/
-.*
-!.git*
-\#*\#
+*\#*
 *~
-*.sw[op]
-*.py[co]
-*.bak
+*.o
+*.su
 /gpp
-*.info
-*.gz
-*.install
-
diff --git a/Makefile b/Makefile
index f841829..dae30f0 100644
--- a/Makefile
+++ b/Makefile
@@ -1,31 +1,32 @@
-VERSION = 1.5
+.POSIX:
 
-PREFIX = /usr
-MANPREFIX = $(PREFIX)/share/man
-
-PKGNAME = gpp
-COMMAND = gpp
-
-PY = python
-SHEBANG = /usr/bin/env $(PY)
+CONFIGFILE = config.mk
+include $(CONFIGFILE)
 
 
 all: gpp
+gpp.o: gpp.c arg.h
 
-gpp: gpp.py
-	env SHEBANG='$(SHEBANG)' $(PY) gpp.py < gpp.py > $@
+gpp: gpp.o
+	$(CC) -o $@ gpp.o $(LDFLAGS)
+
+.c.o:
+	$(CC) -c -o $@ $< $(CFLAGS) $(CPPFLAGS)
 
 install: gpp
 	install -dm755 -- "$(DESTDIR)$(PREFIX)/bin"
-	install -m755 -- gpp "$(DESTDIR)$(PREFIX)/bin/$(COMMAND)"
+	install -m755 -- gpp "$(DESTDIR)$(PREFIX)/bin/gpp"
 	install -dm755 -- "$(DESTDIR)$(MANPREFIX)/man"
-	install -m644 -- gpp.1 "$(DESTDIR)$(MANPREFIX)/man/$(COMMAND).1"
+	install -m644 -- gpp.1 "$(DESTDIR)$(MANPREFIX)/man/gpp.1"
 
 uninstall:
-	-rm -- "$(DESTDIR)$(PREFIX)/bin/$(COMMAND)"
-	-rm -- "$(DESTDIR)$(MANPREFIX)/man/$(COMMAND).1"
+	-rm -- "$(DESTDIR)$(PREFIX)/bin/gpp"
+	-rm -- "$(DESTDIR)$(MANPREFIX)/man/gpp.1"
 
 clean:
-	-rm -rf -- gpp bin obj
+	-rm -rf -- gpp *.o *.su
+
+.SUFFIXES:
+.SUFFIXES: .o .c
 
 .PHONY: all install uninstall clean
diff --git a/README b/README
index a703e75..ec06c25 100644
--- a/README
+++ b/README
@@ -2,15 +2,18 @@ NAME
 	gpp - Bash-based preprocessor for anything
 
 SYNOPSIS
-	gpp [OPTION]...
+	gpp [-D name[=value]] [-f file | [-i input-file]
+	    [-o output-file]] [-n count] [-s symbol] [-u [-u]]
+	    [shell [argument] ...]
 
 ETYMOLOGY
 	gpp stands for General Preprocessor.
 
 DESCRIPTION
 	gpp lets a developer embed directives written in GNU
-	Bash into any text document. These directives are used
-	to automate the writting of parts of the document.
+	Bash (this can be changed) into any text document.
+	These directives are used to automate the writting of
+	parts of the document.
 
 	The preprocessing directives start with a symbol
 	(or text string) specified by the developer. By default
@@ -41,40 +44,43 @@ DESCRIPTION
 	echo verbatim.
 
 OPTIONS
-	-s, --symbol SYMBOL
-		Set the prefix symbol for preprocessor directives.
-		Defaults to @.
+	-D name=value
+		Set the environment variable name to hold
+		the value value.
 
-	-e, --encoding ENCODING
-		Specifies the encoding of the file.
-
-	-n, --iterations N
-		Process the file recursively N times. Defaults to 1 time.
+	-D name
+		Set the environment variable name to hold
+		the value 1.
 
-	-u, --unshebang
-		Clear the shebang line, remove it if this flag
-		is used twice. If used twice, an empty line
-		will be inserted after the new first line.
+	-f file
+		Equivalent to -i FILE -o FILE.
 
-	-i, --input FILE
+	-i input-file
 		Select file to process. Defaults to /dev/stdin.
 
-	-o, --output FILE
+	-n count
+		Process the file recursively count times.
+		Defaults to 1 time.
+
+	-o output-file
 		Select output file. Defaults to /dev/stdout.
 
-	-f, --file FILE
-		Equivalent to -i FILE -o FILE.
+	-s symbol
+		Set the prefix symbol for preprocessor directives.
+		Defaults to @.
 
-	-D, --export NAME=VALUE
-		Set the environment variable NAME to hold
-		the value VALUE.
+	-u
+		Clear the shebang line, remove it if this flag
+		is used twice. If used twice, an empty line
+		will be inserted after the new first line.
 
-	-D, --export NAME
-		Set the environment variable NAME to hold
-		the value 1.
+OPERANDS
+	shell
+		The shell to run instead of bash. The shell
+		must be compatible with POSIX shell.
 
-	Short options must be joined. The value of a flag must
-	be in a separate argument from the flag itself.
+	argument ...
+		Command line arguments for the shell.
 
 EXAMPLES
     Conditional hello world
diff --git a/arg.h b/arg.h
new file mode 100644
index 0000000..79d054c
--- /dev/null
+++ b/arg.h
@@ -0,0 +1,44 @@
+/*
+ * Copy me if you can.
+ * by 20h
+ */
+
+#ifndef ARG_H__
+#define ARG_H__
+
+extern char *argv0;
+
+/* use main(int argc, char *argv[]) */
+#define ARGBEGIN	for (argv0 = *argv, argv++, argc--;\
+					argv[0] && argv[0][0] && argv[0][1];\
+					argc--, argv++) {\
+				char argc_;\
+				char **argv_;\
+				int brk_;\
+				if (argv[0][0] == '-') {\
+					if (argv[0][1] == '-' && argv[0][2] == '\0') {\
+						argv++;\
+						argc--;\
+						break;\
+					}\
+					for (brk_ = 0, argv[0]++, argv_ = argv;\
+							argv[0][0] && !brk_;\
+							argv[0]++) {\
+						if (argv_ != argv)\
+							break;\
+						argc_ = argv[0][0];\
+						switch (argc_)
+
+#define ARGEND				}\
+				} else {\
+					break;\
+				}\
+			}
+
+#define EARGF(x)	((argv[0][1] == '\0' && argv[1] == NULL)?\
+				((x), abort(), (char *)0) :\
+				(brk_ = 1, (argv[0][1] != '\0')?\
+					(&argv[0][1]) :\
+					(argc--, argv++, argv[0])))
+
+#endif
diff --git a/config.mk b/config.mk
new file mode 100644
index 0000000..2e90419
--- /dev/null
+++ b/config.mk
@@ -0,0 +1,6 @@
+PREFIX    = /usr
+MANPREFIX = $(PREFIX)/share/man
+
+CPPFLAGS  = -D_DEFAULT_SOURCE -D_BSD_SOURCE -D_XOPEN_SOURCE=700
+CFLAGS    = -std=c99 -Wall -O2
+LDFLAGS   = -s
diff --git a/gpp.1 b/gpp.1
index 5e32c03..1b7bb69 100644
--- a/gpp.1
+++ b/gpp.1
@@ -4,17 +4,31 @@ gpp - Bash-based preprocessor for anything
 
 .SH SYNOPSIS
 .R gpp
-.RI [ OPTION ]...
+[-D
+.IR name [= value ]]
+[-f
+.I file
+| [-i
+.IR input-file ]
+[-o
+.IR output-file ]]
+[-n
+.IR count ]
+[-s
+.IR symbol ]
+[-u [-u]]
+.RI [ shell
+.RI [ argument ]\ ...]
 
 .SH ETYMOLOGY
 gpp stands for General Preprocessor.
 
 .SH DESCRIPTION
 .B gpp
-lets a developer embed directives written in
-.B GNU Bash
-into any text document. These directives are used
-to automate the writting of parts of the document.
+lets a developer embed directives written in GNU Bash
+(this can be changed) into any text document. These
+directives are used to automate the writting of parts
+of the document.
 .PP
 The preprocessing directives start with a symbol (or
 text string) specified by the developer. By default
@@ -23,7 +37,8 @@ this symbol is
 (at).
 .PP
 Any line starting with 
-.B @< (where
+.B @<
+(where
 .B @
 is the selected symbol for preprocessing directives) or
 .BR @> ,
@@ -73,41 +88,120 @@ Everything that is not a preprocessing directive is
 echo verbatim.
 
 .SH OPTIONS
+The
+.B gpp
+utility conforms to the Base Definitions volume of POSIX.1-2017,
+.IR "Section 12.2" ,
+.IR "Utility Syntax Guidelines" .
+.PP
+The following option is supported:
 .TP
-.BR \-s ,\  \-\-symbol \ \fISYMBOL\fP
-Set the prefix symbol for preprocessor directives.
-Defaults to @ (at).
+.BR \-D\  \fIname\fP\fB=\fP\fIvalue\fP
+Set the environment variable \fIname\fP to hold
+the value \fIvalue\fP.
+.TP
+.BR \-D\  \fIname\fP
+Set the environment variable \fIname\fP to hold
+the value 1.
+.TP
+.BI \-f\  file
+Equivalent to \-i
+.I file
+\-o
+.IR file .
+.TP
+.BI \-i\  input-file
+Select file to process. If
+.B -
+is specified, /dev/stdin will be used.
+Default value is /dev/stdin.
 .TP
-.BR \-e ,\  \-\-encoding \ \fIENCODING\fP
-Specifies the encoding of the file.
+.BI \-n\  n
+Process the file recursively
+.I n
+times. Default value is 1.
 .TP
-.BR \-n ,\  \-\-iterations \ \fIN\fP
-Process the file recursively \fIN\fP times. Defaults to 1 time.
+.BI \-o\  output-file
+Select output file. If
+.B -
+is specified, /dev/stdout will be used.
+Default value is /dev/stdout.
 .TP
-.BR \-u ,\  \-\-unshebang
+.BI \-s\  symbol
+Set the prefix symbol for preprocessor directives.
+Default value is
+.BR @ .
+.TP
+.B \-u
 Clear the shebang line, remove it if this flag
 is used twice. If used twice, an empty line
 will be inserted after the new first line.
+
+.SH OPERANDS
+The following operands are supported:
 .TP
-.BR \-i ,\  \-\-input \ \fIFILE\fP
-Select file to process. Defaults to /dev/stdin.
+.I shell
+The shell to run instead of
+.BR bash .
+The
+.I shell
+must be compatible with POSIX shell.
 .TP
-.BR \-o ,\  \-\-output \ \fIFILE\fP
-Select output file. Defaults to /dev/stdout.
+.IR argument \ ...
+Command line arguments for the shell.
+
+.SH STDIN
+The
+.B gpp
+utility does not use the standard input.
+
+.SH INPUT FILES
+The input file may be of any type, except it may not be a directory.
+
+.SH ENVIRONMENT VARIABLES
+The following environment variables affects the execution of
+.BR gpp :
 .TP
-.BR \-f ,\  \-\-file \ \fIFILE\fP
-Equivalent to \-i \fIFILE\fP \-o \fIFILE\fP.
+.B PATH
+Default. See to the Base Definitions volume of POSIX.1-2017, Section 8.3, Other Environment Variables.
+.PP
+.B gpp
+will set the environment variable
+.B _GPP
+to the zeroth argument
+.B gpp
+was execute with (the name of the command or path to the command).
+
+.SH ASYNCHRONOUS EVENTS
+Default.
+
+.SH STDOUT
+The
+.B gpp
+utility does not use the standard output.
+
+.SH STDERR
+The standard error is only used for diagnostic messages.
+
+.SH OUTPUT FILES
+The output file may be of any type, except it may not be a directory.
+
+.SH EXTENDED DESCRIPTION
+None.
+
+.SH EXIT STATUS
 .TP
-.BR \-D ,\  \-\-export \ \fINAME\fP=\fIVALUE\fP
-Set the environment variable \fINAME\fP to hold
-the value \fIVALUE\fP.
+0
+Successful completion.
 .TP
-.BR \-D ,\  \-\-export \ \fINAME\fP
-Set the environment variable \fINAME\fP to hold
-the value 1.
-.PP
-Short options must be joined. The value of a flag must
-be in a separate argument from the flag itself.
+1
+An error occurred.
+
+.SH CONSEQUENCES OF ERRORS
+Default.
+
+.SH APPLICATION USAGE
+None.
 
 .SH EXAMPLES
 .SS Conditional hello world
@@ -236,7 +330,16 @@ C is one of the few languages that includes a preprocessor,
 some times it is not enough; and all languages need
 preprocessors.
 
-.SH "SEE ALSO"
+.SH NOTES
+None.
+
+.SH BUS
+None.
+
+.SH FUTURE DIRECTIONS
+None.
+
+.SH SEE ALSO
 .BR bash (1),
 .BR jpp (1),
 .BR cpp (1),
diff --git a/gpp.c b/gpp.c
new file mode 100644
index 0000000..569ae76
--- /dev/null
+++ b/gpp.c
@@ -0,0 +1,245 @@
+/* See LICENSE file for copyright and license details. */
+#include <ctype.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+
+#include "arg.h"
+
+
+char *argv0;
+
+
+static void
+usage(void)
+{
+	fprintf(stderr, "usage: %s [-D name[=value]] [-f file | [-i input-file] [-o output-file]] "
+	                "[-n count] [-s symbol] [-u [-u]] [shell [argument] ...]\n", argv0);
+	exit(1);
+}
+
+
+static int
+get_fd(const char *path)
+{
+	const char *p;
+	char *s;
+	long int tmp;
+
+	if (!strcmp(path, "/dev/stdin"))
+		return STDIN_FILENO;
+	if (!strcmp(path, "/dev/stdout"))
+		return STDOUT_FILENO;
+	if (!strcmp(path, "/dev/stderr"))
+		return STDERR_FILENO;
+
+	if (!strncmp(path, "/dev/fd/", sizeof("/dev/fd/") - 1))
+		p = &path[sizeof("/dev/fd/") - 1];
+	else if (!strncmp(path, "/proc/self/fd/", sizeof("/proc/self/fd/") - 1))
+		p = &path[sizeof("/proc/self/fd/") - 1];
+	else
+		return -1;
+
+	if (!isdigit(*p))
+		return -1;
+	errno = 0;
+	tmp = strtol(p, &s, 10);
+	if (errno || tmp > INT_MAX || *s)
+		return -1;
+	return (int)tmp;
+}
+
+
+static int
+xopen(const char *path, int *do_close)
+{
+	int fd = get_fd(path);
+	if (fd >= 0) {
+		*do_close = 0;
+		return fd;
+	}
+	fd = open(path, O_RDONLY);
+	if (fd < 0) {
+		fprintf(stderr, "%s: open %s O_RDONLY: %s\n", argv0, path, strerror(errno));
+		exit(1);
+	}
+	*do_close = 1;
+	return fd;
+}
+
+
+static int
+xcreate(const char *path)
+{
+	int fd = get_fd(path);
+	if (fd >= 0)
+		return fd;
+	fd = open(path, O_WRONLY | O_CREAT | O_TRUNC, 0666);
+	if (fd < 0) {
+		fprintf(stderr, "%s: open %s O_WRONLY|O_CREAT|O_TRUNC 0666: %s\n", argv0, path, strerror(errno));
+		exit(1);
+	}
+	return fd;
+}
+
+
+int
+main(int argc, char *argv[])
+{
+	const char *shell_[] = {"bash", NULL}, **shell = shell_;
+	const char *input_file = NULL;
+	const char *output_file = NULL;
+	const char *symbol = NULL;
+	int iterations = -1;
+	int unshebang = 0;
+	long int tmp;
+	char *arg, *p, *q;
+	char *in_data = NULL, *out_data = NULL;
+	size_t in_size = 0, in_len = 0, in_off = 0;
+	size_t out_size = 0, out_len = 0, out_off = 0;
+	int in_fd, out_fd, do_close;
+	ssize_t r;
+
+	ARGBEGIN {
+	case 'D':
+		arg = EARGF(usage());
+		if (!*arg || *arg == '=')
+			usage();
+		p = strchr(arg, '=');
+		if (p)
+			*p++ = '\0';
+		if (setenv(arg, p ? p : "1", 1)) {
+			fprintf(stderr, "%s: setenv %s %s 1: %s\n", argv0, arg, p ? p : "1", strerror(errno));
+			return 1;
+		}
+		break;
+	case 'f':
+		if (input_file || output_file)
+			usage();
+		input_file = output_file = EARGF(usage());
+		if (!*input_file)
+			usage();
+		break;
+	case 'i':
+		if (input_file)
+			usage();
+		input_file = EARGF(usage());
+		if (!*input_file)
+			usage();
+		break;
+	case 'n':
+		if (iterations >= 0)
+			usage();
+		arg = EARGF(usage());
+		if (!isdigit(*arg))
+			usage();
+		errno = 0;
+		tmp = strtol(arg, &arg, 10);
+		if (errno || tmp > INT_MAX || *arg)
+			usage();
+		iterations = (int)tmp;
+		break;
+	case 'o':
+		if (output_file)
+			usage();
+		output_file = EARGF(usage());
+		if (!*output_file)
+			usage();
+		break;
+	case 's':
+		if (symbol)
+			usage();
+		symbol = EARGF(usage());
+		break;
+	case 'u':
+		if (unshebang == 2)
+			usage();
+		unshebang += 1;
+		break;
+	default:
+		usage();
+	} ARGEND;
+
+	if (argc)
+		shell = (void *)argv;
+
+	if (setenv("_GPP", argv0, 1))
+		fprintf(stderr, "%s: setenv _GPP %s 1: %s\n", argv0, argv0, strerror(errno));
+
+	if (iterations < 0)
+		iterations = 1;
+	if (!symbol)
+		symbol = "@";
+	if (!input_file || (input_file[0] == '-' && !input_file[1]))
+		input_file = "/dev/stdin";
+	if (!output_file || (output_file[0] == '-' && !output_file[1]))
+		output_file = "/dev/stdout";
+
+	in_fd = xopen(input_file, &do_close);
+	for (;;) {
+		if (in_len == in_size) {
+			in_data = realloc(in_data, in_size += 8096);
+			if (!in_data) {
+				fprintf(stderr, "%s: realloc: %s\n", argv0, strerror(errno));
+				return 1;
+			}
+		}
+		r = read(in_fd, &in_data[in_len], in_size - in_len);
+		if (r <= 0) {
+			if (!r)
+				break;
+			fprintf(stderr, "%s: read %s: %s\n", argv0, input_file, strerror(errno));
+		}
+		in_len += (size_t)r;
+	}
+	if (do_close)
+		close(in_fd);
+
+	if (unshebang && in_len >= 2 && in_data[0] == '#' && in_data[1] == '!') {
+		p = memchr(in_data, '\n', in_len);
+		if (!p)
+			goto after_unshebang;
+		in_off = (size_t)(++p - in_data);
+		if (in_off == in_len)
+			goto after_unshebang;
+		if (unshebang >= 2) {
+			q = memchr(p--, '\n', in_len - in_off);
+			if (!q)
+				goto after_unshebang;
+			memmove(p, &p[1], (size_t)(q - in_data) - in_off--);
+			*--q = '\n';
+		}
+	}
+after_unshebang:
+
+	while (iterations--) {
+		/* TODO parse: in -> out */
+
+		in_len = 0;
+		in_off = 0;
+
+		/* TODO shell: out -> in */
+
+		out_len = 0;
+		out_off = 0;
+	}
+
+	free(out_data);
+	out_fd = xcreate(output_file);
+	while (in_off < in_len) {
+		r = write(out_fd, &in_data[in_off], in_len - in_off);
+		if (r <= 0) {
+			fprintf(stderr, "%s: write %s: %s\n", argv0, output_file, strerror(errno));
+			return 1;
+		}
+		in_off += (size_t)r;
+	}
+	if (close(out_fd))
+		fprintf(stderr, "%s: write %s: %s\n", argv0, output_file, strerror(errno));
+	free(in_data);
+	return 0;
+}