* add_input_file() adds a FILE* to the list of input files. We need to
* know all input sources ahead of time to find the last line for the $ match.
*
- * process_files() does actual sedding, reading data lines from each input FILE *
+ * process_files() does actual sedding, reading data lines from each input FILE*
* (which could be stdin) and applying the sed command list (sed_cmd_head) to
* each of the resulting lines.
*
* Todo:
* - Create a wrapper around regex to make libc's regex conform with sed
*
- * Reference http://www.opengroup.org/onlinepubs/007904975/utilities/sed.html
+ * Reference
+ * http://www.opengroup.org/onlinepubs/007904975/utilities/sed.html
+ * http://pubs.opengroup.org/onlinepubs/9699919799/utilities/sed.html
*/
//usage:#define sed_trivial_usage
-//usage: "[-efinr] SED_CMD [FILE]..."
+//usage: "[-inrE] [-f FILE]... [-e CMD]... [FILE]...\n"
+//usage: "or: sed [-inrE] CMD [FILE]..."
//usage:#define sed_full_usage "\n\n"
//usage: " -e CMD Add CMD to sed commands to be executed"
//usage: "\n -f FILE Add FILE contents to sed commands to be executed"
-//usage: "\n -i Edit files in-place (else sends result to stdout)"
+//usage: "\n -i[SFX] Edit files in-place (otherwise sends to stdout)"
+//usage: "\n Optionally back files up, appending SFX"
//usage: "\n -n Suppress automatic printing of pattern space"
-//usage: "\n -r Use extended regex syntax"
+//usage: "\n -r,-E Use extended regex syntax"
//usage: "\n"
//usage: "\nIf no -e or -f, the first non-option argument is the sed command string."
//usage: "\nRemaining arguments are input files (stdin if none)."
static int parse_regex_delim(const char *cmdstr, char **match, char **replace)
{
const char *cmdstr_ptr = cmdstr;
- char delimiter;
+ unsigned char delimiter;
int idx = 0;
/* verify that the 's' or 'y' is followed by something. That something
/* save the replacement string */
cmdstr_ptr += idx + 1;
- idx = index_of_next_unescaped_regexp_delim(-delimiter, cmdstr_ptr);
+ idx = index_of_next_unescaped_regexp_delim(- (int)delimiter, cmdstr_ptr);
*replace = copy_parsing_escapes(cmdstr_ptr, idx);
return ((cmdstr_ptr - cmdstr) + idx);
char *temp;
delimiter = '/';
- if (*my_str == '\\') delimiter = *++pos;
+ if (*my_str == '\\')
+ delimiter = *++pos;
next = index_of_next_unescaped_regexp_delim(delimiter, ++pos);
temp = copy_parsing_escapes(pos, next);
- *regex = xmalloc(sizeof(regex_t));
- xregcomp(*regex, temp, G.regex_type|REG_NEWLINE);
+ *regex = xzalloc(sizeof(regex_t));
+ xregcomp(*regex, temp, G.regex_type);
free(temp);
/* Move position to next character after last delimiter */
pos += (next+1);
/* compile the match string into a regex */
if (*match != '\0') {
/* If match is empty, we use last regex used at runtime */
- sed_cmd->sub_match = xmalloc(sizeof(regex_t));
+ sed_cmd->sub_match = xzalloc(sizeof(regex_t));
+ dbg("xregcomp('%s',%x)", match, cflags);
xregcomp(sed_cmd->sub_match, match, cflags);
+ dbg("regcomp ok");
}
free(match);
}
/* handle edit cmds: (a)ppend, (i)nsert, and (c)hange */
else if (idx <= IDX_c) { /* a,i,c */
- if ((sed_cmd->end_line || sed_cmd->end_match) && sed_cmd->cmd != 'c')
- bb_error_msg_and_die("only a beginning address can be specified for edit commands");
+ if (idx < IDX_c) { /* a,i */
+ if (sed_cmd->end_line || sed_cmd->end_match)
+ bb_error_msg_and_die("command '%c' uses only one address", sed_cmd->cmd);
+ }
for (;;) {
if (*cmdstr == '\n' || *cmdstr == '\\') {
cmdstr++;
}
/* handle file cmds: (r)ead */
else if (idx <= IDX_w) { /* r,w */
- if (sed_cmd->end_line || sed_cmd->end_match)
- bb_error_msg_and_die("command only uses one address");
+ if (idx < IDX_w) { /* r */
+ if (sed_cmd->end_line || sed_cmd->end_match)
+ bb_error_msg_and_die("command '%c' uses only one address", sed_cmd->cmd);
+ }
cmdstr += parse_file_cmd(/*sed_cmd,*/ cmdstr, &sed_cmd->string);
if (sed_cmd->cmd == 'w') {
sed_cmd->sw_file = xfopen_for_write(sed_cmd->string);
sed_cmd->cmd = *cmdstr++;
cmdstr = parse_cmd_args(sed_cmd, cmdstr);
+ /* cmdstr now points past args.
+ * GNU sed requires a separator, if there are more commands,
+ * else it complains "char N: extra characters after command".
+ * Example: "sed 'p;d'". We also allow "sed 'pd'".
+ */
+
/* Add the command to the command array */
*G.sed_cmd_tail = sed_cmd;
G.sed_cmd_tail = &sed_cmd->next;
/* go through the replacement string */
for (i = 0; replace[i]; i++) {
- /* if we find a backreference (\1, \2, etc.) print the backref'ed * text */
+ /* if we find a backreference (\1, \2, etc.) print the backref'ed text */
if (replace[i] == '\\') {
unsigned backref = replace[++i] - '0';
if (backref <= 9) {
static int do_subst_command(sed_cmd_t *sed_cmd, char **line_p)
{
char *line = *line_p;
- int altered = 0;
unsigned match_count = 0;
+ bool altered = 0;
+ bool prev_match_empty = 1;
+ bool tried_at_eol = 0;
regex_t *current_regex;
current_regex = sed_cmd->sub_match;
G.previous_regex_ptr = current_regex;
/* Find the first match */
- if (REG_NOMATCH == regexec(current_regex, line, 10, G.regmatch, 0))
+ dbg("matching '%s'", line);
+ if (REG_NOMATCH == regexec(current_regex, line, 10, G.regmatch, 0)) {
+ dbg("no match");
return 0;
+ }
+ dbg("match");
/* Initialize temporary output buffer. */
G.pipeline.buf = xmalloc(PIPE_GROW);
/* Now loop through, substituting for matches */
do {
+ int start = G.regmatch[0].rm_so;
+ int end = G.regmatch[0].rm_eo;
int i;
- /* Work around bug in glibc regexec, demonstrated by:
- echo " a.b" | busybox sed 's [^ .]* x g'
- The match_count check is so not to break
- echo "hi" | busybox sed 's/^/!/g' */
- if (!G.regmatch[0].rm_so && !G.regmatch[0].rm_eo && match_count) {
- pipe_putc(*line++);
- continue;
- }
-
match_count++;
/* If we aren't interested in this match, output old line to
- end of match and continue */
+ * end of match and continue */
if (sed_cmd->which_match
&& (sed_cmd->which_match != match_count)
) {
- for (i = 0; i < G.regmatch[0].rm_eo; i++)
+ for (i = 0; i < end; i++)
pipe_putc(*line++);
- continue;
+ /* Null match? Print one more char */
+ if (start == end && *line)
+ pipe_putc(*line++);
+ goto next;
}
- /* print everything before the match */
- for (i = 0; i < G.regmatch[0].rm_so; i++)
+ /* Print everything before the match */
+ for (i = 0; i < start; i++)
pipe_putc(line[i]);
- /* then print the substitution string */
- do_subst_w_backrefs(line, sed_cmd->string);
+ /* Then print the substitution string,
+ * unless we just matched empty string after non-empty one.
+ * Example: string "cccd", pattern "c*", repl "R":
+ * result is "RdR", not "RRdR": first match "ccc",
+ * second is "" before "d", third is "" after "d".
+ * Second match is NOT replaced!
+ */
+ if (prev_match_empty || start != 0 || start != end) {
+ //dbg("%d %d %d", prev_match_empty, start, end);
+ dbg("inserting replacement at %d in '%s'", start, line);
+ do_subst_w_backrefs(line, sed_cmd->string);
+ /* Flag that something has changed */
+ altered = 1;
+ } else {
+ dbg("NOT inserting replacement at %d in '%s'", start, line);
+ }
- /* advance past the match */
- line += G.regmatch[0].rm_eo;
- /* flag that something has changed */
- altered++;
+ /* If matched string is empty (f.e. "c*" pattern),
+ * copy verbatim one char after it before attempting more matches
+ */
+ prev_match_empty = (start == end);
+ if (prev_match_empty) {
+ if (!line[end]) {
+ tried_at_eol = 1;
+ } else {
+ pipe_putc(line[end]);
+ end++;
+ }
+ }
+
+ /* Advance past the match */
+ dbg("line += %d", end);
+ line += end;
/* if we're not doing this globally, get out now */
- if (sed_cmd->which_match)
+ if (sed_cmd->which_match != 0)
break;
+ next:
+ /* Exit if we are at EOL and already tried matching at it */
+ if (*line == '\0') {
+ if (tried_at_eol)
+ break;
+ tried_at_eol = 1;
+ }
-//maybe (G.regmatch[0].rm_eo ? REG_NOTBOL : 0) instead of unconditional REG_NOTBOL?
- } while (*line && regexec(current_regex, line, 10, G.regmatch, REG_NOTBOL) != REG_NOMATCH);
+//maybe (end ? REG_NOTBOL : 0) instead of unconditional REG_NOTBOL?
+ } while (regexec(current_regex, line, 10, G.regmatch, REG_NOTBOL) != REG_NOMATCH);
/* Copy rest of string into output pipeline */
while (1) {
static void append(char *s)
{
- llist_add_to_end(&G.append_head, xstrdup(s));
+ llist_add_to_end(&G.append_head, s);
}
-static void flush_append(void)
+/* Output line of text. */
+/* Note:
+ * The tricks with NO_EOL_CHAR and last_puts_char are there to emulate gnu sed.
+ * Without them, we had this:
+ * echo -n thingy >z1
+ * echo -n again >z2
+ * >znull
+ * sed "s/i/z/" z1 z2 znull | hexdump -vC
+ * output:
+ * gnu sed 4.1.5:
+ * 00000000 74 68 7a 6e 67 79 0a 61 67 61 7a 6e |thzngy.agazn|
+ * bbox:
+ * 00000000 74 68 7a 6e 67 79 61 67 61 7a 6e |thzngyagazn|
+ */
+enum {
+ NO_EOL_CHAR = 1,
+ LAST_IS_NUL = 2,
+};
+static void puts_maybe_newline(char *s, FILE *file, char *last_puts_char, char last_gets_char)
+{
+ char lpc = *last_puts_char;
+
+ /* Need to insert a '\n' between two files because first file's
+ * last line wasn't terminated? */
+ if (lpc != '\n' && lpc != '\0') {
+ fputc('\n', file);
+ lpc = '\n';
+ }
+ fputs(s, file);
+
+ /* 'x' - just something which is not '\n', '\0' or NO_EOL_CHAR */
+ if (s[0])
+ lpc = 'x';
+
+ /* had trailing '\0' and it was last char of file? */
+ if (last_gets_char == LAST_IS_NUL) {
+ fputc('\0', file);
+ lpc = 'x'; /* */
+ } else
+ /* had trailing '\n' or '\0'? */
+ if (last_gets_char != NO_EOL_CHAR) {
+ fputc(last_gets_char, file);
+ lpc = last_gets_char;
+ }
+
+ if (ferror(file)) {
+ xfunc_error_retval = 4; /* It's what gnu sed exits with... */
+ bb_error_msg_and_die(bb_msg_write_error);
+ }
+ *last_puts_char = lpc;
+}
+
+static void flush_append(char *last_puts_char, char last_gets_char)
{
char *data;
/* Output appended lines. */
while ((data = (char *)llist_pop(&G.append_head))) {
- fprintf(G.nonstdout, "%s\n", data);
+ puts_maybe_newline(data, G.nonstdout, last_puts_char, last_gets_char);
free(data);
}
}
-static void add_input_file(FILE *file)
-{
- G.input_file_list = xrealloc_vector(G.input_file_list, 2, G.input_file_count);
- G.input_file_list[G.input_file_count++] = file;
-}
-
/* Get next line of input from G.input_file_list, flushing append buffer and
* noting if we ran out of files without a newline on the last line we read.
*/
-enum {
- NO_EOL_CHAR = 1,
- LAST_IS_NUL = 2,
-};
-static char *get_next_line(char *gets_char)
+static char *get_next_line(char *gets_char, char *last_puts_char, char last_gets_char)
{
char *temp = NULL;
int len;
char gc;
- flush_append();
+ flush_append(last_puts_char, last_gets_char);
/* will be returned if last line in the file
* doesn't end with either '\n' or '\0' */
return temp;
}
-/* Output line of text. */
-/* Note:
- * The tricks with NO_EOL_CHAR and last_puts_char are there to emulate gnu sed.
- * Without them, we had this:
- * echo -n thingy >z1
- * echo -n again >z2
- * >znull
- * sed "s/i/z/" z1 z2 znull | hexdump -vC
- * output:
- * gnu sed 4.1.5:
- * 00000000 74 68 7a 6e 67 79 0a 61 67 61 7a 6e |thzngy.agazn|
- * bbox:
- * 00000000 74 68 7a 6e 67 79 61 67 61 7a 6e |thzngyagazn|
- */
-static void puts_maybe_newline(char *s, FILE *file, char *last_puts_char, char last_gets_char)
-{
- char lpc = *last_puts_char;
-
- /* Need to insert a '\n' between two files because first file's
- * last line wasn't terminated? */
- if (lpc != '\n' && lpc != '\0') {
- fputc('\n', file);
- lpc = '\n';
- }
- fputs(s, file);
-
- /* 'x' - just something which is not '\n', '\0' or NO_EOL_CHAR */
- if (s[0])
- lpc = 'x';
-
- /* had trailing '\0' and it was last char of file? */
- if (last_gets_char == LAST_IS_NUL) {
- fputc('\0', file);
- lpc = 'x'; /* */
- } else
- /* had trailing '\n' or '\0'? */
- if (last_gets_char != NO_EOL_CHAR) {
- fputc(last_gets_char, file);
- lpc = last_gets_char;
- }
-
- if (ferror(file)) {
- xfunc_error_retval = 4; /* It's what gnu sed exits with... */
- bb_error_msg_and_die(bb_msg_write_error);
- }
- *last_puts_char = lpc;
-}
-
#define sed_puts(s, n) (puts_maybe_newline(s, G.nonstdout, &last_puts_char, n))
static int beg_match(sed_cmd_t *sed_cmd, const char *pattern_space)
int substituted;
/* Prime the pump */
- next_line = get_next_line(&next_gets_char);
+ next_line = get_next_line(&next_gets_char, &last_puts_char, '\n' /*last_gets_char*/);
/* Go through every line in each file */
again:
/* Read one line in advance so we can act on the last line,
* the '$' address */
- next_line = get_next_line(&next_gets_char);
+ next_line = get_next_line(&next_gets_char, &last_puts_char, last_gets_char);
linenum++;
/* For every line, go through all the commands */
/* or does this line matches our last address regex */
|| (sed_cmd->end_match && old_matched
&& (regexec(sed_cmd->end_match,
- pattern_space, 0, NULL, 0) == 0)
+ pattern_space, 0, NULL, 0) == 0)
)
);
}
}
/* actual sedding */
- //bb_error_msg("pattern_space:'%s' next_line:'%s' cmd:%c",
- //pattern_space, next_line, sed_cmd->cmd);
+ dbg("pattern_space:'%s' next_line:'%s' cmd:%c",
+ pattern_space, next_line, sed_cmd->cmd);
switch (sed_cmd->cmd) {
/* Print line number */
case 's':
if (!do_subst_command(sed_cmd, &pattern_space))
break;
+ dbg("do_subst_command succeeded:'%s'", pattern_space);
substituted |= 1;
/* handle p option */
/* Append line to linked list to be printed later */
case 'a':
- append(sed_cmd->string);
+ append(xstrdup(sed_cmd->string));
break;
/* Insert text before this line */
rfile = fopen_for_read(sed_cmd->string);
if (rfile) {
char *line;
-
while ((line = xmalloc_fgetline(rfile))
!= NULL)
append(line);
- xprint_and_close_file(rfile);
+ fclose(rfile);
}
break;
free(pattern_space);
pattern_space = next_line;
last_gets_char = next_gets_char;
- next_line = get_next_line(&next_gets_char);
+ next_line = get_next_line(&next_gets_char, &last_puts_char, last_gets_char);
substituted = 0;
linenum++;
break;
pattern_space[len] = '\n';
strcpy(pattern_space + len+1, next_line);
last_gets_char = next_gets_char;
- next_line = get_next_line(&next_gets_char);
+ next_line = get_next_line(&next_gets_char, &last_puts_char, last_gets_char);
linenum++;
break;
}
/* Delete and such jump here. */
discard_line:
- flush_append();
+ flush_append(&last_puts_char, last_gets_char);
free(pattern_space);
goto again;
/* It is possible to have a command line argument with embedded
* newlines. This counts as multiple command lines.
* However, newline can be escaped: 's/e/z\<newline>z/'
- * We check for this.
+ * add_cmd() handles this.
*/
static void add_cmd_block(char *cmdstr)
cmdstr = sv = xstrdup(cmdstr);
do {
eol = strchr(cmdstr, '\n');
- next:
- if (eol) {
- /* Count preceding slashes */
- int slashes = 0;
- char *sl = eol;
-
- while (sl != cmdstr && *--sl == '\\')
- slashes++;
- /* Odd number of preceding slashes - newline is escaped */
- if (slashes & 1) {
- overlapping_strcpy(eol - 1, eol);
- eol = strchr(eol, '\n');
- goto next;
- }
+ if (eol)
*eol = '\0';
- }
add_cmd(cmdstr);
cmdstr = eol + 1;
} while (eol);
free(sv);
}
+static void add_input_file(FILE *file)
+{
+ G.input_file_list = xrealloc_vector(G.input_file_list, 2, G.input_file_count);
+ G.input_file_list[G.input_file_count++] = file;
+}
+
int sed_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE;
int sed_main(int argc UNUSED_PARAM, char **argv)
{
unsigned opt;
llist_t *opt_e, *opt_f;
+ char *opt_i;
+
+#if ENABLE_LONG_OPTS
+ static const char sed_longopts[] ALIGN1 =
+ /* name has_arg short */
+ "in-place\0" Optional_argument "i"
+ "regexp-extended\0" No_argument "r"
+ "quiet\0" No_argument "n"
+ "silent\0" No_argument "n"
+ "expression\0" Required_argument "e"
+ "file\0" Required_argument "f";
+#endif
+
int status = EXIT_SUCCESS;
INIT_G();
if (ENABLE_FEATURE_CLEAN_UP) atexit(sed_free_and_close_stuff);
/* Lie to autoconf when it starts asking stupid questions. */
- if (argv[1] && !strcmp(argv[1], "--version")) {
+ if (argv[1] && strcmp(argv[1], "--version") == 0) {
puts("This is not GNU sed version 4.0");
return 0;
}
/* do normal option parsing */
opt_e = opt_f = NULL;
+ opt_i = NULL;
opt_complementary = "e::f::" /* can occur multiple times */
"nn"; /* count -n */
+
+ IF_LONG_OPTS(applet_long_options = sed_longopts);
+
/* -i must be first, to match OPT_in_place definition */
- opt = getopt32(argv, "irne:f:", &opt_e, &opt_f,
+ /* -E is a synonym of -r:
+ * GNU sed 4.2.1 mentions it in neither --help
+ * nor manpage, but does recognize it.
+ */
+ opt = getopt32(argv, "i::rEne:f:", &opt_i, &opt_e, &opt_f,
&G.be_quiet); /* counter for -n */
//argc -= optind;
argv += optind;
if (opt & OPT_in_place) { // -i
atexit(cleanup_outname);
}
- if (opt & 0x2) G.regex_type |= REG_EXTENDED; // -r
- //if (opt & 0x4) G.be_quiet++; // -n
+ if (opt & (2|4))
+ G.regex_type |= REG_EXTENDED; // -r or -E
+ //if (opt & 8)
+ // G.be_quiet++; // -n (implemented with a counter instead)
while (opt_e) { // -e
add_cmd_block(llist_pop(&opt_e));
}
fclose(cmdfile);
}
/* if we didn't get a pattern from -e or -f, use argv[0] */
- if (!(opt & 0x18)) {
+ if (!(opt & 0x30)) {
if (!*argv)
bb_show_usage();
add_cmd_block(*argv++);
fclose(G.nonstdout);
G.nonstdout = stdout;
- /* unlink(argv[i]); */
- xrename(G.outname, argv[i]);
+ if (opt_i) {
+ char *backupname = xasprintf("%s%s", argv[i], opt_i);
+ xrename(argv[i], backupname);
+ free(backupname);
+ }
+ /* else unlink(argv[i]); - rename below does this */
+ xrename(G.outname, argv[i]); //TODO: rollback backup on error?
free(G.outname);
G.outname = NULL;
projects / platform / upstream / busybox.git / blobdiff