struct dep *d;
struct dep **dp;
const char *file_stem = f->stem;
+ const char *fstem;
int initialized = 0;
f->updating = 0;
{
char *p;
struct dep *new, *next;
- char *name = (char *)d->name;
if (! d->name || ! d->need_2nd_expansion)
{
continue;
}
- /* If it's from a static pattern rule, convert the patterns into
- "$*" so they'll expand properly. */
+ /* If it's from a static pattern rule, convert the initial pattern in
+ each word to "$*" so they'll expand properly. */
if (d->staticpattern)
{
- char *o = subst_expand (variable_buffer, name, "%", "$*", 1, 2, 0);
- *o = '\0';
- free (name);
- d->name = name = xstrdup (variable_buffer);
- d->staticpattern = 0;
+ const char *cs = d->name;
+ size_t nperc = 0;
+
+ /* Count the number of % in the string. */
+ while ((cs = strchr (cs, '%')) != NULL)
+ {
+ ++nperc;
+ ++cs;
+ }
+
+ if (nperc)
+ {
+ /* Allocate enough space to replace all % with $*. */
+ size_t slen = strlen (d->name) + nperc + 1;
+ const char *pcs = d->name;
+ char *name = xmalloc (slen);
+ char *s = name;
+
+ /* Substitute the first % in each word. */
+ cs = strchr (pcs, '%');
+
+ while (cs)
+ {
+ memcpy (s, pcs, cs - pcs);
+ s += cs - pcs;
+ *(s++) = '$';
+ *(s++) = '*';
+ pcs = ++cs;
+
+ /* Find the first % after the next whitespace. */
+ cs = strchr (end_of_token (cs), '%');
+ }
+ strcpy (s, pcs);
+
+ free ((char*)d->name);
+ d->name = name;
+ }
}
/* We're going to do second expansion so initialize file variables for
p = variable_expand_for_file (d->name, f);
+ /* Free the un-expanded name. */
+ free ((char*)d->name);
+
if (d->stem != 0)
f->stem = file_stem;
- /* At this point we don't need the name anymore: free it. */
- free (name);
-
/* Parse the prerequisites and enter them into the file database. */
- new = enter_prereqs (split_prereqs (p), d->stem);
+ new = split_prereqs (p);
/* If there were no prereqs here (blank!) then throw this one out. */
if (new == 0)
}
/* Add newly parsed prerequisites. */
+ fstem = d->stem;
next = d->next;
+ free_dep (d);
*dp = new;
- for (dp = &new->next, d = new->next; d != 0; dp = &d->next, d = d->next)
- ;
+ for (dp = &new, d = new; d != 0; dp = &d->next, d = d->next)
+ {
+ d->file = lookup_file (d->name);
+ if (d->file == 0)
+ d->file = enter_file (d->name);
+ d->name = 0;
+ d->stem = fstem;
+ if (!fstem)
+ /* This file is explicitly mentioned as a prereq. */
+ d->file->is_explicit = 1;
+ }
*dp = next;
d = *dp;
}
# perprerequisite's stem (Savannah bug #16053).
#
run_make_test('
+.RECIPEPREFIX := >
+
all.foo.bar: %.foo.bar: %.one
all.foo.bar: %.bar: %.two
all.foo.bar:
- @echo $*
- @echo $^
+> @echo $*
+> @echo $^
.DEFAULT:;@:
',
# (Savannah bug #16053).
#
run_make_test('
+.RECIPEPREFIX := >
.SECONDEXPANSION:
all.foo.bar: %.foo.bar: %.one $$*-one
all.foo.bar: %.bar: %.two $$*-two
all.foo.bar:
- @echo $*
- @echo $^
+> @echo $*
+> @echo $^
.DEFAULT:;@:
',
unlink('hello.z');
+my @dir = ('', 'lib/'); # With and without last slash.
+my @secondexpansion = ('', '.SECONDEXPANSION:');
+
+# The following combinations are generated with and without second expansion.
+# 1.
+# all: bye.x
+# bye.x: %.x: ...
+#
+# 2.
+# all: lib/bye.x
+# lib/bye.x: %.x: ...
+#
+# 3.
+# all: lib/bye.x
+# lib/bye.x: lib/%.x: ...
+#
+# The following combination is not generated, because there is no rule to
+# build bye.x, no stem substitution takes place, not of interest of this test.
+# 4.
+# all: bye.x
+# bye.x: lib/%.x: ...
+#
+
+for my $se (@secondexpansion) {
+for my $d (@dir) { # The directory of the prerequisite of 'all'.
+for my $r (@dir) { # The directory of the prerequisite in the rule definition.
+(!$d && $r) && next; # Combination 4.
+my $dollar = $se ? '$' : '';
+
+# The prerequisite should only have directory if the prerequisite of 'all' has
+# it and if the prequisite pattern in the rule definition does not have it.
+# That is combination 2.
+my $pdir = $d && !$r ? $d : '';
+
+
+# One func, one %.
+my $prereqs = "${pdir}bye.1";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}bye.x
+${d}bye.x: $r%.x: $dollar\$(firstword %.1); \$(info \$@ from \$^)
+", '', "${d}bye.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+
+
+# Multiple funcs, each has one %.
+$prereqs = "${pdir}bye.1 ${pdir}bye.2";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}bye.x
+${d}bye.x: $r%.x: $dollar\$(firstword %.1) $dollar\$(firstword %.2); \$(info \$@ from \$^)
+", '', "${d}bye.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+
+
+# Multiple funcs, each has multiple %.
+$prereqs = "${pdir}bye.1 ${pdir}bye.2 ${pdir}bye.3 ${pdir}bye.4";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}bye.x
+${d}bye.x: $r%.x: $dollar\$(wordlist 1, 99, %.1 %.2) $dollar\$(wordlist 1, 99, %.3 %.4); \$(info \$@ from \$^)
+", '', "${d}bye.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+
+
+# Multiple funcs, each has multiple %, each prerequisite has multiple %.
+$prereqs = "${pdir}bye_%_%.1 ${pdir}bye_%_%.2 ${pdir}bye_%_%.3 ${pdir}bye_%_%.4";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}bye.x
+${d}bye.x: $r%.x: $dollar\$(wordlist 1, 99, %_%_%.1 %_%_%.2) $dollar\$(wordlist 1, 99, %_%_%.3 %_%_%.4); \$(info \$@ from \$^)
+", '', "${d}bye.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+
+
+# Nested functions.
+$prereqs = "${pdir}bye.1 ${pdir}bye.2 ${pdir}bye.3 ${pdir}bye.4";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}bye.x
+${d}bye.x: $r%.x: $dollar\$(wordlist 1, 99, $dollar\$(wordlist 1, 99, %.1 %.2)) $dollar\$(wordlist 1, 99, $dollar\$(wordlist 1,99, %.3 %.4)); \$(info \$@ from \$^)
+", '', "${d}bye.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+
+
+# Multiple funcs, each has multiple words, each word has multiple %, sole %,
+# various corner cases.
+# Make should substitude the first % and only the first % in each word with the
+# stem.
+$prereqs = "${pdir}bye1%2% ${pdir}bye 3${pdir}bye4%5 6${pdir}bye ${pdir}bye7%8 ${pdir}bye9 ${pdir}bye10% 11${pdir}bye12 13";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}bye.x
+${d}bye.x: $r%.x: $dollar\$(wordlist 1, 99, %1%2% % 3%4%5 6%) %7%8 %9 $dollar\$(wordlist 1, 99, %10% 11%12) 13; \$(info \$@ from \$^)
+", '', "${d}bye.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+
+
+if ($port_type eq 'UNIX') {
+# Test that make does not use some hardcoded array of a finite size on stack.
+# Long prerequisite name. This prerequisite name is over 66K long.
+my $prefix = 'abcdefgh' x 128 x 33; # 33K long.
+my $suffix = 'stuvwxyz' x 128 x 33; # 33K long.
+$prereqs = "${prefix}${pdir}bye${suffix}.1 ${prefix}${pdir}bye${suffix}.2";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}bye.x
+${d}bye.x: $r%.x: $dollar\$(wordlist 1, 99, ${prefix}%${suffix}.1 ${prefix}%${suffix}.2); \$(info \$@ from \$^)
+", '', "${d}bye.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+}
+
+
+# Empty stem.
+$prereqs = "${pdir}.1";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}bye.x
+${d}bye.x: $r%bye.x: $dollar\$(firstword %.1); \$(info \$@ from \$^)
+", '', "${d}bye.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+
+
+# A word expands to an empty prerequisite.
+run_make_test("
+$se
+all: ${d}bye.x
+${d}bye.x: $r%.x: $dollar\$(%); \$(info \$@ from \$^)
+", '', "${d}bye.x from \n#MAKE#: Nothing to be done for 'all'.\n");
+
+}
+}
+}
+
+# Escaped %.
+# The following combinations are generated without second expansion.
+# 1.
+# all: the%weird\\_hello_pattern\\.x
+# the\\%weird\\_hello_pattern\\.x: the\\%weird\\_%_pattern\\.x: ...
+#
+# 2.
+# all: lib/the%weird\\_hello_pattern\\.x
+# lib/the\\%weird\\_hello_pattern\\.x: lib/the\\%weird\\_%_pattern\\.x: ...
+#
+# Other combinations or second expansion are not tested, because escaped % is
+# not implemented for those.
+
+for my $d (@dir) {
+my $prereqs = "${d}the%weird\\\\_hello_pattern%\\\\.1 ${d}the%weird\\\\_hello_pattern%\\\\.2";
+run_make_test("
+$se
+.PHONY: $prereqs
+all: ${d}the%weird\\\\_hello_pattern\\\\.x
+${d}the\\%weird\\\\_hello_pattern\\\\.x: ${d}the\\%weird\\\\_%_pattern\\\\.x: $dollar\$(wordlist 1, 99, ${d}the\\%weird\\\\_%_pattern%\\\\.1 ${d}the\\%weird\\\\_%_pattern%\\\\.2); \$(info \$@ from \$^)
+", '', "${d}the%weird\\\\_hello_pattern\\\\.x from $prereqs\n#MAKE#: Nothing to be done for 'all'.\n");
+}
+
1;