From 0068ce2fa0e3dbc618b8ca87ab56e144731da784 Mon Sep 17 00:00:00 2001 From: Rob Landley Date: Tue, 20 Jul 2021 16:02:31 +0200 Subject: cut: add toybox-compatible options -O OUTSEP, -D, -F LIST function old new delta cut_main 884 1201 +317 packed_usage 33823 33885 +62 .rodata 104186 104179 -7 ------------------------------------------------------------------------------ (add/remove: 0/0 grow/shrink: 2/1 up/down: 379/-7) Total: 372 bytes Signed-off-by: Rob Landley Signed-off-by: Denys Vlasenko --- coreutils/cut.c | 207 ++++++++++++++++++++++++++++++-------------------------- 1 file changed, 110 insertions(+), 97 deletions(-) (limited to 'coreutils') diff --git a/coreutils/cut.c b/coreutils/cut.c index cc3c325..7009e74 100644 --- a/coreutils/cut.c +++ b/coreutils/cut.c @@ -14,6 +14,13 @@ //config: help //config: cut is used to print selected parts of lines from //config: each file to stdout. +//config: +//config:config FEATURE_CUT_REGEX +//config: bool "cut -F" +//config: default y +//config: depends on CUT +//config: help +//config: Allow regex based delimiters. //applet:IF_CUT(APPLET_NOEXEC(cut, cut, BB_DIR_USR_BIN, BB_SUID_DROP, cut)) @@ -25,9 +32,14 @@ //usage: "Print selected fields from FILEs to stdout\n" //usage: "\n -b LIST Output only bytes from LIST" //usage: "\n -c LIST Output only characters from LIST" -//usage: "\n -d CHAR Use CHAR instead of tab as field delimiter" +//usage: "\n -d SEP Field delimiter for input (default -f TAB, -F run of whitespace)" +//usage: "\n -O SEP Field delimeter for output (default = -d for -f, one space for -F)" +//usage: "\n -D Don't sort/collate sections or match -fF lines without delimeter" +//usage: "\n -f LIST Print only these fields (-d is single char)" +//usage: IF_FEATURE_CUT_REGEX( +//usage: "\n -F LIST Print only these fields (-d is regex)" +//usage: ) //usage: "\n -s Output only lines containing delimiter" -//usage: "\n -f LIST Print only these fields" //usage: "\n -n Ignored" //(manpage:-n with -b: don't split multibyte characters) //usage: @@ -39,38 +51,49 @@ #include "libbb.h" +#if ENABLE_FEATURE_CUT_REGEX +#include "xregex.h" +#else +#define regex_t int +typedef struct { int rm_eo, rm_so; } regmatch_t; +#define xregcomp(x, ...) *(x) = 0 +#define regexec(...) 0 +#endif + /* This is a NOEXEC applet. Be very careful! */ /* option vars */ -#define OPT_STR "b:c:f:d:sn" +#define OPT_STR "b:c:f:d:O:sD"IF_FEATURE_CUT_REGEX("F:")"n" #define CUT_OPT_BYTE_FLGS (1 << 0) #define CUT_OPT_CHAR_FLGS (1 << 1) #define CUT_OPT_FIELDS_FLGS (1 << 2) #define CUT_OPT_DELIM_FLGS (1 << 3) -#define CUT_OPT_SUPPRESS_FLGS (1 << 4) +#define CUT_OPT_ODELIM_FLGS (1 << 4) +#define CUT_OPT_SUPPRESS_FLGS (1 << 5) +#define CUT_OPT_NOSORT_FLGS (1 << 6) +#define CUT_OPT_REGEX_FLGS ((1 << 7) * ENABLE_FEATURE_CUT_REGEX) struct cut_list { int startpos; int endpos; }; -enum { - BOL = 0, - EOL = INT_MAX, - NON_RANGE = -1 -}; - static int cmpfunc(const void *a, const void *b) { return (((struct cut_list *) a)->startpos - ((struct cut_list *) b)->startpos); } -static void cut_file(FILE *file, char delim, const struct cut_list *cut_lists, unsigned nlists) +static void cut_file(FILE *file, const char *delim, const char *odelim, + const struct cut_list *cut_lists, unsigned nlists) { char *line; unsigned linenum = 0; /* keep these zero-based to be consistent */ + regex_t reg; + int spos, shoe = option_mask32 & CUT_OPT_REGEX_FLGS; + + if (shoe) xregcomp(®, delim, REG_EXTENDED); /* go through every line in the file */ while ((line = xmalloc_fgetline(file)) != NULL) { @@ -80,29 +103,22 @@ static void cut_file(FILE *file, char delim, const struct cut_list *cut_lists, u char *printed = xzalloc(linelen + 1); char *orig_line = line; unsigned cl_pos = 0; - int spos; /* cut based on chars/bytes XXX: only works when sizeof(char) == byte */ if (option_mask32 & (CUT_OPT_CHAR_FLGS | CUT_OPT_BYTE_FLGS)) { /* print the chars specified in each cut list */ for (; cl_pos < nlists; cl_pos++) { - spos = cut_lists[cl_pos].startpos; - while (spos < linelen) { + for (spos = cut_lists[cl_pos].startpos; spos < linelen;) { if (!printed[spos]) { printed[spos] = 'X'; putchar(line[spos]); } - spos++; - if (spos > cut_lists[cl_pos].endpos - /* NON_RANGE is -1, so if below is true, - * the above was true too (spos is >= 0) */ - /* || cut_lists[cl_pos].endpos == NON_RANGE */ - ) { + if (++spos > cut_lists[cl_pos].endpos) { break; } } } - } else if (delim == '\n') { /* cut by lines */ + } else if (*delim == '\n') { /* cut by lines */ spos = cut_lists[cl_pos].startpos; /* get out if we have no more lists to process or if the lines @@ -115,9 +131,7 @@ static void cut_file(FILE *file, char delim, const struct cut_list *cut_lists, u while (spos < (int)linenum) { spos++; /* go to the next list if we're at the end of this one */ - if (spos > cut_lists[cl_pos].endpos - || cut_lists[cl_pos].endpos == NON_RANGE - ) { + if (spos > cut_lists[cl_pos].endpos) { cl_pos++; /* get out if there's no more lists to process */ if (cl_pos >= nlists) @@ -135,55 +149,56 @@ static void cut_file(FILE *file, char delim, const struct cut_list *cut_lists, u puts(line); goto next_line; } else { /* cut by fields */ - int ndelim = -1; /* zero-based / one-based problem */ - int nfields_printed = 0; - char *field = NULL; - char delimiter[2]; - - delimiter[0] = delim; - delimiter[1] = 0; - - /* does this line contain any delimiters? */ - if (strchr(line, delim) == NULL) { - if (!(option_mask32 & CUT_OPT_SUPPRESS_FLGS)) - puts(line); - goto next_line; - } - - /* process each list on this line, for as long as we've got - * a line to process */ - for (; cl_pos < nlists && line; cl_pos++) { - spos = cut_lists[cl_pos].startpos; - do { - /* find the field we're looking for */ - while (line && ndelim < spos) { - field = strsep(&line, delimiter); - ndelim++; - } - - /* we found it, and it hasn't been printed yet */ - if (field && ndelim == spos && !printed[ndelim]) { - /* if this isn't our first time through, we need to - * print the delimiter after the last field that was - * printed */ - if (nfields_printed > 0) - putchar(delim); - fputs_stdout(field); - printed[ndelim] = 'X'; - nfields_printed++; /* shouldn't overflow.. */ + unsigned uu = 0, start = 0, end = 0, out = 0; + int dcount = 0; + + /* Loop through bytes, finding next delimiter */ + for (;;) { + /* End of current range? */ + if (end == linelen || dcount > cut_lists[cl_pos].endpos) { + if (++cl_pos >= nlists) break; + if (option_mask32 & CUT_OPT_NOSORT_FLGS) + start = dcount = uu = 0; + end = 0; + } + /* End of current line? */ + if (uu == linelen) { + /* If we've seen no delimiters, check -s */ + if (!cl_pos && !dcount && !shoe) { + if (option_mask32 & CUT_OPT_SUPPRESS_FLGS) + goto next_line; + } else if (dcount= 0, - * while endpos may be = NON_RANGE (-1) */ + /* NB: startpos is always >= 0 */ cut_lists[nlists].startpos = s; cut_lists[nlists].endpos = e; nlists++; @@ -294,7 +306,8 @@ int cut_main(int argc UNUSED_PARAM, char **argv) /* now that the lists are parsed, we need to sort them to make life * easier on us when it comes time to print the chars / fields / lines */ - qsort(cut_lists, nlists, sizeof(cut_lists[0]), cmpfunc); + if (!(opt & CUT_OPT_NOSORT_FLGS)) + qsort(cut_lists, nlists, sizeof(cut_lists[0]), cmpfunc); } { @@ -309,7 +322,7 @@ int cut_main(int argc UNUSED_PARAM, char **argv) retval = EXIT_FAILURE; continue; } - cut_file(file, delim, cut_lists, nlists); + cut_file(file, delim, odelim, cut_lists, nlists); fclose_if_not_stdin(file); } while (*++argv); -- cgit v1.1