diff options
| -rw-r--r-- | ChangeLog | 48 | ||||
| -rw-r--r-- | array.c | 2 | ||||
| -rw-r--r-- | awkgram.c | 18 | ||||
| -rw-r--r-- | awkgram.y | 18 | ||||
| -rw-r--r-- | builtin.c | 119 | ||||
| -rw-r--r-- | command.c | 10 | ||||
| -rw-r--r-- | command.y | 10 | ||||
| -rw-r--r-- | debug.c | 16 | ||||
| -rw-r--r-- | eval.c | 4 | ||||
| -rw-r--r-- | extension/ChangeLog | 7 | ||||
| -rw-r--r-- | extension/readfile.c | 4 | ||||
| -rw-r--r-- | extension/rwarray.c | 4 | ||||
| -rw-r--r-- | extension/rwarray0.c | 4 | ||||
| -rw-r--r-- | field.c | 4 | ||||
| -rw-r--r-- | gawkapi.h | 2 | ||||
| -rw-r--r-- | helpers/ChangeLog | 8 | ||||
| -rw-r--r-- | helpers/testdfa.c | 21 | ||||
| -rw-r--r-- | interpret.h | 7 | ||||
| -rw-r--r-- | io.c | 11 | ||||
| -rw-r--r-- | node.c | 14 | ||||
| -rw-r--r-- | old-extension/ChangeLog | 9 | ||||
| -rw-r--r-- | old-extension/bindarr.c | 2 | ||||
| -rw-r--r-- | old-extension/fileop.c | 2 | ||||
| -rw-r--r-- | profile.c | 9 | ||||
| -rw-r--r-- | re.c | 4 |
25 files changed, 209 insertions, 148 deletions
@@ -47,6 +47,18 @@ Unrelated: * builtin.c (call_sub): Fix for indirect gensub, 3 args now works. + Unrelated: + + * builtin.c (do_sub): Improve some variable names for readability + and add / expand some comments. + +2015-04-14 Andrew J. Schorr <aschorr@telemetry-investments.com> + Arnold D. Robbins <arnold@skeeve.com> + + * builtin.c (do_sub): Make computations smarter; initial len + to malloc, test for final amount after all matches done and + need to copy in the final part of the original string. + 2015-04-13 Arnold D. Robbins <arnold@skeeve.com> * regcomp.c (analyze): Prevent malloc(0). @@ -68,6 +80,15 @@ * Makefile.am (efence): Make this link again. Thanks to Michal Jaegermann for pointing out the problem. +2015-04-09 Andrew J. Schorr <aschorr@telemetry-investments.com> + + * awkgram.y (yyerror): Rationalize buffer size computations. Remove + old valgrind workarounds. + * debug.c (gprintf): Rationalize buffer size computations. + (serialize_subscript): Ditto. + * io.c (iop_finish): Rationalize buffer size computations. + * profile.c (pp_string): Correct space allocation computation. + 2015-04-08 John E. Malmberg <wb8tyw@qsl.net> * custom.h: VMS shares some code paths with ZOS_USS in @@ -441,6 +462,33 @@ * profile.c (pprint): Be sure to set ip2 in all paths through the code. Thanks to GCC 4.9 for the warning. +2014-12-18 Arnold D. Robbins <arnold@skeeve.com> + + * builtin.c (do_sub): Do not waste a byte at the end of a string. + +2014-12-14 Arnold D. Robbins <arnold@skeeve.com> + + * awkgram.y (yyerror): Do not waste a byte at the end of a string. + * builtin.c (do_match): Ditto. + * command.y (append_statement): Ditto. + * debug.c (gprintf, serialize): Ditto. + * field.c (set_FIELDWIDTHS): Ditto. + * io.c.c (grow_iop_buffer): Ditto. + * profile.c (pp_string, pp_group3): Ditto. + +2014-12-14 Andrew J. Schorr <aschorr@telemetry-investments.com> + + * array.c (concat_exp): Do not waste a byte at the end of a string. + * awkgram.y (common_exp): Ditto. + * builtin.c (do_substr): Ditto. + * eval.c (set_OFS): Ditto. + * field.c (rebuild_record): Ditto. + * gawkapi.h (r_make_string): Ditto. + * interpret.h (r_interpret): Ditto for Op_assign_concat. + * node.c (r_format_val, r_dupnode, make_str_node, str2wstr, wstr2str): + Ditto. + * re.c (make_regexp): Ditto. + 2014-12-20 Arnold D. Robbins <arnold@skeeve.com> Enable non-fatal output on per-file or global basis, @@ -414,7 +414,7 @@ concat_exp(int nargs, bool do_subsep) } len += (nargs - 1) * subseplen; - emalloc(str, char *, len + 2, "concat_exp"); + emalloc(str, char *, len + 1, "concat_exp"); r = args_array[nargs]; memcpy(str, r->stptr, r->stlen); @@ -3399,7 +3399,7 @@ regular_print: n1 = force_string(n1); n2 = force_string(n2); nlen = n1->stlen + n2->stlen; - erealloc(n1->stptr, char *, nlen + 2, "constant fold"); + erealloc(n1->stptr, char *, nlen + 1, "constant fold"); memcpy(n1->stptr + n1->stlen, n2->stptr, n2->stlen); n1->stlen = nlen; n1->stptr[nlen] = '\0'; @@ -4516,7 +4516,6 @@ yyerror(const char *m, ...) char *buf; int count; static char end_of_file_line[] = "(END OF FILE)"; - char save; print_included_from(); @@ -4544,24 +4543,15 @@ yyerror(const char *m, ...) bp = thisline + strlen(thisline); } - /* - * Saving and restoring *bp keeps valgrind happy, - * since the guts of glibc uses strlen, even though - * we're passing an explict precision. Sigh. - * - * 8/2003: We may not need this anymore. - */ - save = *bp; - *bp = '\0'; - msg("%.*s", (int) (bp - thisline), thisline); - *bp = save; va_start(args, m); if (mesg == NULL) mesg = m; - count = (bp - thisline) + strlen(mesg) + 2 + 1; + count = strlen(mesg) + 1; + if (lexptr != NULL) + count += (lexeme - thisline) + 2; emalloc(buf, char *, count, "yyerror"); bp = buf; @@ -1443,7 +1443,7 @@ common_exp n1 = force_string(n1); n2 = force_string(n2); nlen = n1->stlen + n2->stlen; - erealloc(n1->stptr, char *, nlen + 2, "constant fold"); + erealloc(n1->stptr, char *, nlen + 1, "constant fold"); memcpy(n1->stptr + n1->stlen, n2->stptr, n2->stlen); n1->stlen = nlen; n1->stptr[nlen] = '\0'; @@ -2178,7 +2178,6 @@ yyerror(const char *m, ...) char *buf; int count; static char end_of_file_line[] = "(END OF FILE)"; - char save; print_included_from(); @@ -2206,24 +2205,15 @@ yyerror(const char *m, ...) bp = thisline + strlen(thisline); } - /* - * Saving and restoring *bp keeps valgrind happy, - * since the guts of glibc uses strlen, even though - * we're passing an explict precision. Sigh. - * - * 8/2003: We may not need this anymore. - */ - save = *bp; - *bp = '\0'; - msg("%.*s", (int) (bp - thisline), thisline); - *bp = save; va_start(args, m); if (mesg == NULL) mesg = m; - count = (bp - thisline) + strlen(mesg) + 2 + 1; + count = strlen(mesg) + 1; + if (lexptr != NULL) + count += (lexeme - thisline) + 2; emalloc(buf, char *, count, "yyerror"); bp = buf; @@ -1847,7 +1847,7 @@ do_substr(int nargs) * way to do things. */ memset(& mbs, 0, sizeof(mbs)); - emalloc(substr, char *, (length * gawk_mb_cur_max) + 2, "do_substr"); + emalloc(substr, char *, (length * gawk_mb_cur_max) + 1, "do_substr"); wp = t1->wstptr + indx; for (cp = substr; length > 0; length--) { result = wcrtomb(cp, *wp, & mbs); @@ -2597,7 +2597,7 @@ do_match(int nargs) sprintf(buff, "%d", ii); ilen = strlen(buff); - amt = ilen + subseplen + strlen("length") + 2; + amt = ilen + subseplen + strlen("length") + 1; if (oldamt == 0) { emalloc(buf, char *, amt, "do_match"); @@ -2771,42 +2771,42 @@ do_sub(int nargs, unsigned int flags) int ampersands; int matches = 0; Regexp *rp; - NODE *s; /* subst. pattern */ - NODE *t; /* string to make sub. in; $0 if none given */ + NODE *rep_node; /* replacement text */ + NODE *target; /* string to make sub. in; $0 if none given */ NODE *tmp; NODE **lhs = NULL; long how_many = 1; /* one substitution for sub, also gensub default */ - int global; + bool global; long current; bool lastmatchnonzero; char *mb_indices = NULL; if ((flags & GENSUB) != 0) { double d; - NODE *t1; + NODE *glob_flag; tmp = PEEK(3); rp = re_update(tmp); - t = POP_STRING(); /* original string */ + target = POP_STRING(); /* original string */ - t1 = POP_SCALAR(); /* value of global flag */ - if ((t1->flags & (STRCUR|STRING)) != 0) { - if (t1->stlen > 0 && (t1->stptr[0] == 'g' || t1->stptr[0] == 'G')) + glob_flag = POP_SCALAR(); /* value of global flag */ + if ((glob_flag->flags & (STRCUR|STRING)) != 0) { + if (glob_flag->stlen > 0 && (glob_flag->stptr[0] == 'g' || glob_flag->stptr[0] == 'G')) how_many = -1; else { - (void) force_number(t1); - d = get_number_d(t1); - if ((t1->flags & NUMCUR) != 0) + (void) force_number(glob_flag); + d = get_number_d(glob_flag); + if ((glob_flag->flags & NUMCUR) != 0) goto set_how_many; warning(_("gensub: third argument `%.*s' treated as 1"), - (int) t1->stlen, t1->stptr); + (int) glob_flag->stlen, glob_flag->stptr); how_many = 1; } } else { - (void) force_number(t1); - d = get_number_d(t1); + (void) force_number(glob_flag); + d = get_number_d(glob_flag); set_how_many: if (d < 1) how_many = 1; @@ -2817,10 +2817,8 @@ set_how_many: if (d <= 0) warning(_("gensub: third argument %g treated as 1"), d); } - DEREF(t1); - + DEREF(glob_flag); } else { - /* take care of regexp early, in case re_update is fatal */ tmp = PEEK(2); @@ -2832,30 +2830,30 @@ set_how_many: /* original string */ if ((flags & LITERAL) != 0) - t = POP_STRING(); + target = POP_STRING(); else { lhs = POP_ADDRESS(); - t = force_string(*lhs); + target = force_string(*lhs); } } global = (how_many == -1); - s = POP_STRING(); /* replacement text */ + rep_node = POP_STRING(); /* replacement text */ decr_sp(); /* regexp, already updated above */ /* do the search early to avoid work on non-match */ - if (research(rp, t->stptr, 0, t->stlen, RE_NEED_START) == -1 || - RESTART(rp, t->stptr) > t->stlen) + if (research(rp, target->stptr, 0, target->stlen, RE_NEED_START) == -1 || + RESTART(rp, target->stptr) > target->stlen) goto done; - t->flags |= STRING; + target->flags |= STRING; - text = t->stptr; - textlen = t->stlen; + text = target->stptr; + textlen = target->stlen; - repl = s->stptr; - replend = repl + s->stlen; + repl = rep_node->stptr; + replend = repl + rep_node->stlen; repllen = replend - repl; ampersands = 0; @@ -2873,6 +2871,7 @@ set_how_many: index_multibyte_buffer(repl, mb_indices, repllen); } + /* compute length of replacement string, number of ampersands */ for (scan = repl; scan < replend; scan++) { if ((gawk_mb_cur_max == 1 || (repllen > 0 && mb_indices[scan - repl] == 1)) && (*scan == '&')) { @@ -2917,24 +2916,32 @@ set_how_many: lastmatchnonzero = false; - /* guesstimate how much room to allocate; +2 forces > 0 */ - buflen = textlen + (ampersands + 1) * repllen + 2; - emalloc(buf, char *, buflen + 2, "do_sub"); + /* guesstimate how much room to allocate; +1 forces > 0 */ + buflen = textlen + (ampersands + 1) * repllen + 1; + emalloc(buf, char *, buflen + 1, "do_sub"); buf[buflen] = '\0'; - buf[buflen + 1] = '\0'; bp = buf; for (current = 1;; current++) { matches++; - matchstart = t->stptr + RESTART(rp, t->stptr); - matchend = t->stptr + REEND(rp, t->stptr); + matchstart = target->stptr + RESTART(rp, target->stptr); + matchend = target->stptr + REEND(rp, target->stptr); /* * create the result, copying in parts of the original - * string + * string. note that length of replacement string can + * vary since ampersand is actual text of regexp match. + */ + + /* + * add 1 to len to handle "empty" case where + * matchend == matchstart and we force a match on a single + * char. Use 'matchend - text' instead of 'matchstart - text' + * because we may not actually make any substitution depending + * on the 'global' and 'how_many' values. */ - len = matchstart - text + repllen - + ampersands * (matchend - matchstart); + len = matchend - text + repllen + + ampersands * (matchend - matchstart) + 1; sofar = bp - buf; while (buflen < (sofar + len + 1)) { buflen *= 2; @@ -2981,13 +2988,13 @@ set_how_many: if (flags & GENSUB) { /* gensub, behave sanely */ if (isdigit((unsigned char) scan[1])) { int dig = scan[1] - '0'; - if (dig < NUMSUBPATS(rp, t->stptr) && SUBPATSTART(rp, tp->stptr, dig) != -1) { + if (dig < NUMSUBPATS(rp, target->stptr) && SUBPATSTART(rp, tp->stptr, dig) != -1) { char *start, *end; - start = t->stptr - + SUBPATSTART(rp, t->stptr, dig); - end = t->stptr - + SUBPATEND(rp, t->stptr, dig); + start = target->stptr + + SUBPATSTART(rp, target->stptr, dig); + end = target->stptr + + SUBPATEND(rp, target->stptr, dig); for (cp = start; cp < end; cp++) *bp++ = *cp; @@ -3041,19 +3048,29 @@ set_how_many: textlen = text + textlen - matchend; text = matchend; +#if 0 + if (bp - buf > sofar + len) + fprintf(stderr, "debug: len = %zu, but used %ld\n", len, (long)((bp - buf) - (long)sofar)); +#endif + if ((current >= how_many && ! global) || ((long) textlen <= 0 && matchstart == matchend) - || research(rp, t->stptr, text - t->stptr, textlen, RE_NEED_START) == -1) + || research(rp, target->stptr, text - target->stptr, textlen, RE_NEED_START) == -1) break; } sofar = bp - buf; - if (buflen - sofar - textlen - 1) { - buflen = sofar + textlen + 2; + if (buflen < (sofar + textlen + 1)) { + buflen = sofar + textlen + 1; erealloc(buf, char *, buflen, "do_sub"); bp = buf + sofar; } - for (scan = matchend; scan < text + textlen; scan++) + /* + * Note that text == matchend, since that assignment is made before + * exiting the 'for' loop above. Thus we copy in the rest of the + * original string. + */ + for (scan = text; scan < text + textlen; scan++) *bp++ = *scan; *bp = '\0'; textlen = bp - buf; @@ -3062,7 +3079,7 @@ set_how_many: efree(mb_indices); done: - DEREF(s); + DEREF(rep_node); if ((matches == 0 || (flags & LITERAL) != 0) && buf != NULL) { efree(buf); @@ -3072,18 +3089,18 @@ done: if (flags & GENSUB) { if (matches > 0) { /* return the result string */ - DEREF(t); + DEREF(target); assert(buf != NULL); return make_str_node(buf, textlen, ALREADY_MALLOCED); } /* return the original string */ - return t; + return target; } /* For a string literal, must not change the original string. */ if ((flags & LITERAL) != 0) - DEREF(t); + DEREF(target); else if (matches > 0) { unref(*lhs); *lhs = make_str_node(buf, textlen, ALREADY_MALLOCED); @@ -2515,7 +2515,7 @@ append_statement(CMDARG *stmt_list, char *stmt) len += strlen(a->a_string) + 1; /* 1 for ',' */ len += EVALSIZE; - emalloc(s, char *, (len + 2) * sizeof(char), "append_statement"); + emalloc(s, char *, (len + 1) * sizeof(char), "append_statement"); arg = mk_cmdarg(D_string); arg->a_string = s; arg->a_count = len; /* kludge */ @@ -2542,7 +2542,7 @@ append_statement(CMDARG *stmt_list, char *stmt) ssize = stmt_list->a_count; if (len > ssize - slen) { ssize = slen + len + EVALSIZE; - erealloc(s, char *, (ssize + 2) * sizeof(char), "append_statement"); + erealloc(s, char *, (ssize + 1) * sizeof(char), "append_statement"); stmt_list->a_string = s; stmt_list->a_count = ssize; } @@ -2554,7 +2554,7 @@ append_statement(CMDARG *stmt_list, char *stmt) } if (stmt == end_EVAL) - erealloc(stmt_list->a_string, char *, slen + 2, "append_statement"); + erealloc(stmt_list->a_string, char *, slen + 1, "append_statement"); return stmt_list; #undef EVALSIZE @@ -2921,7 +2921,7 @@ again: bool esc_seen = false; toklen = lexend - lexptr; - emalloc(str, char *, toklen + 2, "yylex"); + emalloc(str, char *, toklen + 1, "yylex"); p = str; while ((c = *++lexptr) != '"') { @@ -3100,7 +3100,7 @@ concat_args(CMDARG *arg, int count) arg = arg->next; } - emalloc(str, char *, len + 2, "concat_args"); + emalloc(str, char *, len + 1, "concat_args"); n = tmp[0]; memcpy(str, n->stptr, n->stlen); p = str + n->stlen; @@ -764,7 +764,7 @@ append_statement(CMDARG *stmt_list, char *stmt) len += strlen(a->a_string) + 1; /* 1 for ',' */ len += EVALSIZE; - emalloc(s, char *, (len + 2) * sizeof(char), "append_statement"); + emalloc(s, char *, (len + 1) * sizeof(char), "append_statement"); arg = mk_cmdarg(D_string); arg->a_string = s; arg->a_count = len; /* kludge */ @@ -791,7 +791,7 @@ append_statement(CMDARG *stmt_list, char *stmt) ssize = stmt_list->a_count; if (len > ssize - slen) { ssize = slen + len + EVALSIZE; - erealloc(s, char *, (ssize + 2) * sizeof(char), "append_statement"); + erealloc(s, char *, (ssize + 1) * sizeof(char), "append_statement"); stmt_list->a_string = s; stmt_list->a_count = ssize; } @@ -803,7 +803,7 @@ append_statement(CMDARG *stmt_list, char *stmt) } if (stmt == end_EVAL) - erealloc(stmt_list->a_string, char *, slen + 2, "append_statement"); + erealloc(stmt_list->a_string, char *, slen + 1, "append_statement"); return stmt_list; #undef EVALSIZE @@ -1170,7 +1170,7 @@ again: bool esc_seen = false; toklen = lexend - lexptr; - emalloc(str, char *, toklen + 2, "yylex"); + emalloc(str, char *, toklen + 1, "yylex"); p = str; while ((c = *++lexptr) != '"') { @@ -1349,7 +1349,7 @@ concat_args(CMDARG *arg, int count) arg = arg->next; } - emalloc(str, char *, len + 2, "concat_args"); + emalloc(str, char *, len + 1, "concat_args"); n = tmp[0]; memcpy(str, n->stptr, n->stlen); p = str + n->stlen; @@ -4205,10 +4205,10 @@ gprintf(FILE *fp, const char *format, ...) #define GPRINTF_BUFSIZ 512 if (buf == NULL) { buflen = GPRINTF_BUFSIZ; - emalloc(buf, char *, (buflen + 2) * sizeof(char), "gprintf"); + emalloc(buf, char *, buflen * sizeof(char), "gprintf"); } else if (buflen - bl < GPRINTF_BUFSIZ/2) { buflen += GPRINTF_BUFSIZ; - erealloc(buf, char *, (buflen + 2) * sizeof(char), "gprintf"); + erealloc(buf, char *, buflen * sizeof(char), "gprintf"); } #undef GPRINTF_BUFSIZ @@ -4227,7 +4227,7 @@ gprintf(FILE *fp, const char *format, ...) /* enlarge buffer, and try again */ buflen *= 2; - erealloc(buf, char *, (buflen + 2) * sizeof(char), "gprintf"); + erealloc(buf, char *, buflen * sizeof(char), "gprintf"); } bl = 0; @@ -4267,7 +4267,7 @@ gprintf(FILE *fp, const char *format, ...) static int serialize_subscript(char *buf, int buflen, struct list_item *item) { - int bl = 0, nchar, i; + int bl, nchar, i; NODE *sub; nchar = snprintf(buf, buflen, "%d%c%d%c%s%c%d%c", @@ -4277,7 +4277,7 @@ serialize_subscript(char *buf, int buflen, struct list_item *item) return 0; else if (nchar >= buflen) /* need larger buffer */ return nchar; - bl += nchar; + bl = nchar; for (i = 0; i < item->num_subs; i++) { sub = item->subs[i]; nchar = snprintf(buf + bl, buflen - bl, "%lu%c%s%c", @@ -4356,7 +4356,7 @@ serialize(int type) if (buf == NULL) { /* first time */ buflen = SERIALIZE_BUFSIZ; - emalloc(buf, char *, buflen + 2, "serialize"); + emalloc(buf, char *, buflen + 1, "serialize"); } bl = 0; @@ -4365,7 +4365,7 @@ serialize(int type) if (buflen - bl < SERIALIZE_BUFSIZ/2) { enlarge_buffer: buflen *= 2; - erealloc(buf, char *, buflen + 2, "serialize"); + erealloc(buf, char *, buflen + 1, "serialize"); } #undef SERIALIZE_BUFSIZ @@ -4466,7 +4466,7 @@ enlarge_buffer: } if (nchar > 0) { /* non-empty commands list */ - nchar += (strlen("commands ") + 20 + strlen("end") + 2); /* 20 for cnum (an int) */ + nchar += (strlen("commands ") + 20 + strlen("end") + 1); /* 20 for cnum (an int) */ if (nchar > buflen - bl) { buflen = bl + nchar; erealloc(buf, char *, buflen + 3, "serialize"); @@ -827,9 +827,9 @@ set_OFS() new_ofs_len = OFS_node->var_value->stlen; if (OFS == NULL) - emalloc(OFS, char *, new_ofs_len + 2, "set_OFS"); + emalloc(OFS, char *, new_ofs_len + 1, "set_OFS"); else if (OFSlen < new_ofs_len) - erealloc(OFS, char *, new_ofs_len + 2, "set_OFS"); + erealloc(OFS, char *, new_ofs_len + 1, "set_OFS"); memcpy(OFS, OFS_node->var_value->stptr, OFS_node->var_value->stlen); OFSlen = new_ofs_len; diff --git a/extension/ChangeLog b/extension/ChangeLog index 2769a350..9df8ae00 100644 --- a/extension/ChangeLog +++ b/extension/ChangeLog @@ -86,6 +86,13 @@ (errno_la_SOURCES, errno_la_LDFLAGS, errno_la_LIBADD): Remove. (EXTRA_DIST): Remove errlist.h. +2014-12-14 Andrew J. Schorr <aschorr@telemetry-investments.com> + + * readfile.c (read_file_to_buffer): Do not waste a byte at the end of + a string. + * rwarray.c (read_value): Ditto. + * rwarray0.c (read_value): Ditto. + 2014-11-23 Arnold D. Robbins <arnold@skeeve.com> * inplace.c (do_inplace_begin): Jump through hoops to silence diff --git a/extension/readfile.c b/extension/readfile.c index d4b4aef9..7673589f 100644 --- a/extension/readfile.c +++ b/extension/readfile.c @@ -82,8 +82,8 @@ read_file_to_buffer(int fd, const struct stat *sbuf) goto done; } - emalloc(text, char *, sbuf->st_size + 2, "do_readfile"); - memset(text, '\0', sbuf->st_size + 2); + emalloc(text, char *, sbuf->st_size + 1, "do_readfile"); + memset(text, '\0', sbuf->st_size + 1); if ((ret = read(fd, text, sbuf->st_size)) != sbuf->st_size) { update_ERRNO_int(errno); diff --git a/extension/rwarray.c b/extension/rwarray.c index aa05a0d5..5bead973 100644 --- a/extension/rwarray.c +++ b/extension/rwarray.c @@ -468,8 +468,8 @@ read_value(FILE *fp, awk_value_t *value) len = ntohl(len); value->val_type = AWK_STRING; value->str_value.len = len; - value->str_value.str = gawk_malloc(len + 2); - memset(value->str_value.str, '\0', len + 2); + value->str_value.str = gawk_malloc(len + 1); + memset(value->str_value.str, '\0', len + 1); if (fread(value->str_value.str, 1, len, fp) != (ssize_t) len) { gawk_free(value->str_value.str); diff --git a/extension/rwarray0.c b/extension/rwarray0.c index e2de3cf5..ec3663cc 100644 --- a/extension/rwarray0.c +++ b/extension/rwarray0.c @@ -452,8 +452,8 @@ read_value(int fd, awk_value_t *value) len = ntohl(len); value->val_type = AWK_STRING; value->str_value.len = len; - value->str_value.str = malloc(len + 2); - memset(value->str_value.str, '\0', len + 2); + value->str_value.str = malloc(len + 1); + memset(value->str_value.str, '\0', len + 1); if (read(fd, value->str_value.str, len) != (ssize_t) len) { free(value->str_value.str); @@ -163,7 +163,7 @@ rebuild_record() tlen += (NF - 1) * OFSlen; if ((long) tlen < 0) tlen = 0; - emalloc(ops, char *, tlen + 2, "rebuild_record"); + emalloc(ops, char *, tlen + 1, "rebuild_record"); cops = ops; ops[0] = '\0'; for (i = 1; i <= NF; i++) { @@ -1138,7 +1138,7 @@ set_FIELDWIDTHS() FIELDWIDTHS[0] = 0; for (i = 1; ; i++) { unsigned long int tmp; - if (i + 2 >= fw_alloc) { + if (i + 1 >= fw_alloc) { fw_alloc *= 2; erealloc(FIELDWIDTHS, int *, fw_alloc * sizeof(int), "set_FIELDWIDTHS"); } @@ -827,7 +827,7 @@ r_make_string(const gawk_api_t *api, /* needed for emalloc */ result->str_value.len = length; if (duplicate) { - emalloc(cp, char *, length + 2, "r_make_string"); + emalloc(cp, char *, length + 1, "r_make_string"); memcpy(cp, string, length); cp[length] = '\0'; result->str_value.str = cp; diff --git a/helpers/ChangeLog b/helpers/ChangeLog index a5bbafb1..5a3f2fe3 100644 --- a/helpers/ChangeLog +++ b/helpers/ChangeLog @@ -1,3 +1,11 @@ +2015-04-09 Andrew J. Schorr <aschorr@telemetry-investments.com> + + * testdfa.c (setup_pattern): Rationalize buffer size computations. + +2014-12-18 Arnold D. Robbins <arnold@skeeve.com> + + * testdfa.c (setup_pattern): Do not waste a byte at the end of a string. + 2014-09-04 Arnold D. Robbins <arnold@skeeve.com> * chlistref.awk: New file. Finds @ref{} to non-chapters. diff --git a/helpers/testdfa.c b/helpers/testdfa.c index 25a229a2..2b773467 100644 --- a/helpers/testdfa.c +++ b/helpers/testdfa.c @@ -372,10 +372,10 @@ setup_pattern(const char *pattern, size_t *len) { size_t is_multibyte = 0; int c, c2; - size_t buflen = 0; + size_t buflen; mbstate_t mbs; bool has_anchor = false; - char *buf = NULL; + char *buf; char *dest; const char *src, *end; @@ -391,21 +391,12 @@ setup_pattern(const char *pattern, size_t *len) * escaped characters translated, and generate the regex * from that. */ + buf = (char *) malloc(*len + 1); if (buf == NULL) { - buf = (char *) malloc(*len + 2); - if (buf == NULL) { - fprintf(stderr, "%s: malloc failed\n", __func__); - exit(EXIT_FAILURE); - } - buflen = *len; - } else if (*len > buflen) { - buf = (char *) realloc(buf, *len + 2); - if (buf == NULL) { - fprintf(stderr, "%s: realloc failed\n", __func__); - exit(EXIT_FAILURE); - } - buflen = *len; + fprintf(stderr, "%s: malloc failed\n", __func__); + exit(EXIT_FAILURE); } + buflen = *len; dest = buf; while (src < end) { diff --git a/interpret.h b/interpret.h index b76952f1..03532f43 100644 --- a/interpret.h +++ b/interpret.h @@ -709,7 +709,7 @@ mod: if (t1 != t2 && t1->valref == 1 && (t1->flags & MPFN) == 0) { size_t nlen = t1->stlen + t2->stlen; - erealloc(t1->stptr, char *, nlen + 2, "r_interpret"); + erealloc(t1->stptr, char *, nlen + 1, "r_interpret"); memcpy(t1->stptr + t1->stlen, t2->stptr, t2->stlen); t1->stlen = nlen; t1->stptr[nlen] = '\0'; @@ -719,7 +719,7 @@ mod: size_t wlen = t1->wstlen + t2->wstlen; erealloc(t1->wstptr, wchar_t *, - sizeof(wchar_t) * (wlen + 2), "r_interpret"); + sizeof(wchar_t) * (wlen + 1), "r_interpret"); memcpy(t1->wstptr + t1->wstlen, t2->wstptr, t2->wstlen); t1->wstlen = wlen; t1->wstptr[wlen] = L'\0'; @@ -730,9 +730,10 @@ mod: size_t nlen = t1->stlen + t2->stlen; char *p; - emalloc(p, char *, nlen + 2, "r_interpret"); + emalloc(p, char *, nlen + 1, "r_interpret"); memcpy(p, t1->stptr, t1->stlen); memcpy(p + t1->stlen, t2->stptr, t2->stlen); + /* N.B. No NUL-termination required, since make_str_node will do it. */ unref(*lhs); t1 = *lhs = make_str_node(p, nlen, ALREADY_MALLOCED); } @@ -2709,6 +2709,7 @@ init_awkpath(path_info *pi) end++; len = end - start; if (len > 0) { + /* +2 is correct here; leave room for / */ emalloc(p, char *, len + 2, "init_awkpath"); memcpy(p, start, len); @@ -3163,7 +3164,7 @@ iop_finish(IOBUF *iop) lintwarn(_("data file `%s' is empty"), iop->public.name); iop->errcode = errno = 0; iop->count = iop->scanoff = 0; - emalloc(iop->buf, char *, iop->size += 2, "iop_finish"); + emalloc(iop->buf, char *, iop->size += 1, "iop_finish"); iop->off = iop->buf; iop->dataend = NULL; iop->end = iop->buf + iop->size; @@ -3195,10 +3196,10 @@ grow_iop_buffer(IOBUF *iop) size_t newsize; /* - * Lop off original extra two bytes, double the size, - * add them back. + * Lop off original extra byte, double the size, + * add it back. */ - newsize = ((iop->size - 2) * 2) + 2; + newsize = ((iop->size - 1) * 2) + 1; /* Check for overflow */ if (newsize <= iop->size) @@ -3206,7 +3207,7 @@ grow_iop_buffer(IOBUF *iop) /* Make sure there's room for a disk block */ if (newsize - valid < iop->readsize) - newsize += iop->readsize + 2; + newsize += iop->readsize + 1; /* Check for overflow, again */ if (newsize <= iop->size) @@ -261,7 +261,7 @@ r_format_val(const char *format, int index, NODE *s) } if (s->stptr != NULL) efree(s->stptr); - emalloc(s->stptr, char *, s->stlen + 2, "format_val"); + emalloc(s->stptr, char *, s->stlen + 1, "format_val"); memcpy(s->stptr, sp, s->stlen + 1); no_malloc: s->flags |= STRCUR; @@ -299,12 +299,12 @@ r_dupnode(NODE *n) r->wstlen = 0; if ((n->flags & STRCUR) != 0) { - emalloc(r->stptr, char *, n->stlen + 2, "r_dupnode"); + emalloc(r->stptr, char *, n->stlen + 1, "r_dupnode"); memcpy(r->stptr, n->stptr, n->stlen); r->stptr[n->stlen] = '\0'; if ((n->flags & WSTRCUR) != 0) { r->wstlen = n->wstlen; - emalloc(r->wstptr, wchar_t *, sizeof(wchar_t) * (n->wstlen + 2), "r_dupnode"); + emalloc(r->wstptr, wchar_t *, sizeof(wchar_t) * (n->wstlen + 1), "r_dupnode"); memcpy(r->wstptr, n->wstptr, n->wstlen * sizeof(wchar_t)); r->wstptr[n->wstlen] = L'\0'; r->flags |= WSTRCUR; @@ -377,7 +377,7 @@ make_str_node(const char *s, size_t len, int flags) if ((flags & ALREADY_MALLOCED) != 0) r->stptr = (char *) s; else { - emalloc(r->stptr, char *, len + 2, "make_str_node"); + emalloc(r->stptr, char *, len + 1, "make_str_node"); memcpy(r->stptr, s, len); } r->stptr[len] = '\0'; @@ -681,7 +681,7 @@ str2wstr(NODE *n, size_t **ptr) * realloc the wide string down in size. */ - emalloc(n->wstptr, wchar_t *, sizeof(wchar_t) * (n->stlen + 2), "str2wstr"); + emalloc(n->wstptr, wchar_t *, sizeof(wchar_t) * (n->stlen + 1), "str2wstr"); wsp = n->wstptr; /* @@ -755,7 +755,7 @@ str2wstr(NODE *n, size_t **ptr) n->flags |= WSTRCUR; #define ARBITRARY_AMOUNT_TO_GIVE_BACK 100 if (n->stlen - n->wstlen > ARBITRARY_AMOUNT_TO_GIVE_BACK) - erealloc(n->wstptr, wchar_t *, sizeof(wchar_t) * (n->wstlen + 2), "str2wstr"); + erealloc(n->wstptr, wchar_t *, sizeof(wchar_t) * (n->wstlen + 1), "str2wstr"); return n; } @@ -782,7 +782,7 @@ wstr2str(NODE *n) memset(& mbs, 0, sizeof(mbs)); length = n->wstlen; - emalloc(newval, char *, (length * gawk_mb_cur_max) + 2, "wstr2str"); + emalloc(newval, char *, (length * gawk_mb_cur_max) + 1, "wstr2str"); wp = n->wstptr; for (cp = newval; length > 0; length--) { diff --git a/old-extension/ChangeLog b/old-extension/ChangeLog index 55c8d31d..dad5c794 100644 --- a/old-extension/ChangeLog +++ b/old-extension/ChangeLog @@ -1,3 +1,12 @@ +2015-04-09 Andrew J. Schorr <aschorr@telemetry-investments.com> + + * bindarr.c (do_bind_array): Undo Arnold's change of 2014-12-18. + +2014-12-18 Arnold D. Robbins <arnold@skeeve.com> + + * bindarr.c (do_bind_array): Do not waste a byte at the end of a string. + * fileop.c (do_fread): Ditto. + 2014-01-07 Arnold D. Robbins <arnold@skeeve.com> * dbarray.awk: Use full name for lib to load in extenstion() call. diff --git a/old-extension/bindarr.c b/old-extension/bindarr.c index 60959903..41467425 100644 --- a/old-extension/bindarr.c +++ b/old-extension/bindarr.c @@ -235,7 +235,7 @@ do_bind_array(int nargs) } /* copy the array -- this is passed as the second argument to the functions */ - emalloc(aname, char *, strlen(t->vname) + 2, "do_bind_array"); + emalloc(aname, char *, 1 + strlen(symbol->vname) + 1, "do_bind_array"); aname[0] = '~'; /* any illegal character */ strcpy(& aname[1], symbol->vname); td = make_array(); diff --git a/old-extension/fileop.c b/old-extension/fileop.c index 86f62576..d76a7ded 100644 --- a/old-extension/fileop.c +++ b/old-extension/fileop.c @@ -55,7 +55,7 @@ do_fread(int nargs) force_number(arg); rlen = get_number_ui(arg); - emalloc(rbuf, char *, rlen + 2, "do_fread"); + emalloc(rbuf, char *, rlen + 1, "do_fread"); if ((count = fread(rbuf, 1, rlen, f->fp)) < rlen) { if (! feof(f->fp)) update_ERRNO_int(errno); @@ -1359,7 +1359,7 @@ pp_string(const char *in_str, size_t len, int delim) osiz *= 2; \ } ofre -= (l) - osiz = len + 3 + 2; /* initial size; 3 for delim + terminating null */ + osiz = len + 3 + 1; /* initial size; 3 for delim + terminating null */ emalloc(obuf, char *, osiz, "pp_string"); obufout = obuf; ofre = osiz - 1; @@ -1371,10 +1371,9 @@ pp_string(const char *in_str, size_t len, int delim) *obufout++ = '\\'; *obufout++ = delim; } else if (*str == '\0') { - chksize(4); - *obufout++ = '\\'; *obufout++ = '0'; + chksize(2); /* need 2 more chars for this case */ *obufout++ = '0'; *obufout++ = '0'; } else if ((cp = strchr(escapes, *str)) != NULL) { @@ -1384,7 +1383,7 @@ pp_string(const char *in_str, size_t len, int delim) /* NB: Deliberate use of lower-case versions. */ } else if (isascii(*str) && isprint(*str)) { *obufout++ = *str; - ofre += 1; + ofre += 1; /* used 1 less than expected */ } else { size_t len; @@ -1607,7 +1606,7 @@ pp_group3(const char *s1, const char *s2, const char *s3) len1 = strlen(s1); len2 = strlen(s2); len3 = strlen(s3); - l = len1 + len2 + len3 + 2; + l = len1 + len2 + len3 + 1; emalloc(str, char *, l, "pp_group3"); s = str; if (len1 > 0) { @@ -75,10 +75,10 @@ make_regexp(const char *s, size_t len, bool ignorecase, bool dfa, bool canfatal) * from that. */ if (buf == NULL) { - emalloc(buf, char *, len + 2, "make_regexp"); + emalloc(buf, char *, len + 1, "make_regexp"); buflen = len; } else if (len > buflen) { - erealloc(buf, char *, len + 2, "make_regexp"); + erealloc(buf, char *, len + 1, "make_regexp"); buflen = len; } dest = buf; |