Skip to content

Handle include and exclude the same way bsdtar does #4

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
cerisier merged 4 commits into from
Feb 5, 2026
Merged

Handle include and exclude the same way bsdtar does #4

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
9d033af
Include and exclude paths should apply after strip
cerisier Feb 5, 2026
d218c5b
Revert "Include and exclude paths should apply after strip"
cerisier Feb 5, 2026
4f91cfb
wip
cerisier Feb 5, 2026
b14b47b
wip
cerisier Feb 5, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
187 changes: 119 additions & 68 deletions pkgutil.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -79,9 +79,17 @@ static char *strip_components_path(const char *path, int strip);
static int apply_strip_components(struct archive_entry *e, int strip);
static int path_component_count(const char *path);
static char *normalize_rel_path(const char *path);
static int match_excluded_with_prefix(struct archive *match,
struct archive_entry *e,
const char *prefix);
struct pattern_list {
char **items;
size_t len;
size_t cap;
};
static void pattern_list_add(struct pattern_list *list, const char *pattern);
static void pattern_list_free(struct pattern_list *list);
static int should_extract_path(struct archive *matching, const char *path);
static char *join_prefix_path(const char *prefix, const char *path);
static int has_include_descendant(const struct pattern_list *includes,
const char *path);

static int pkg_getopt(int *argc, char ***argv, const char **arg) {
enum { state_start = 0, state_next_word, state_short, state_long };
Expand Down Expand Up @@ -263,7 +271,7 @@ static int astream_close_cb(struct archive *a, void *client_data) {

static void extract_nested_archive_from_stream(struct astream *in,
const char *outdir, int flags,
struct archive *match,
struct archive *matching,
int strip_components,
const char *prefix) {
struct archive *a = archive_read_new();
Expand Down Expand Up @@ -308,14 +316,14 @@ static void extract_nested_archive_from_stream(struct astream *in,
char *rel = normalize_rel_path(p);
archive_entry_set_pathname(e, rel);

if (match != NULL) {
r = match_excluded_with_prefix(match, e, prefix);
if (r != 0) {
archive_read_data_skip(a);
free(rel);
continue;
}
char *logical_path = join_prefix_path(prefix, rel);
if (!should_extract_path(matching, logical_path)) {
archive_read_data_skip(a);
free(logical_path);
free(rel);
continue;
}
free(logical_path);

if (apply_strip_components(e, strip_components)) {
archive_read_data_skip(a);
Expand Down Expand Up @@ -456,41 +464,79 @@ static int path_component_count(const char *path) {
return (count);
}

static int match_excluded_with_prefix(struct archive *match,
struct archive_entry *e,
const char *prefix) {
if (match == NULL) {
return (0);
static void pattern_list_add(struct pattern_list *list, const char *pattern) {
if (list->len == list->cap) {
size_t new_cap = list->cap == 0 ? 8 : list->cap * 2;
char **new_items = realloc(list->items, new_cap * sizeof(*new_items));
if (new_items == NULL) {
fail_errno("realloc");
}
list->items = new_items;
list->cap = new_cap;
}
if (prefix == NULL || prefix[0] == '\0') {
return archive_match_excluded(match, e);
char *dup = strdup(pattern);
if (dup == NULL) {
fail_errno("strdup");
}
const char *orig = archive_entry_pathname(e);
if (orig == NULL) {
return archive_match_excluded(match, e);
list->items[list->len++] = dup;
}

static void pattern_list_free(struct pattern_list *list) {
for (size_t i = 0; i < list->len; i++) {
free(list->items[i]);
}
char *orig_copy = strdup(orig);
if (orig_copy == NULL) {
fail_errno("strdup");
free(list->items);
list->items = NULL;
list->len = 0;
list->cap = 0;
}

static int should_extract_path(struct archive *matching, const char *path) {
struct archive_entry *entry = archive_entry_new();
if (entry == NULL) {
fail_errno("archive_entry_new");
}
archive_entry_set_pathname(entry, path);
int excluded = archive_match_excluded(matching, entry);
archive_entry_free(entry);
if (excluded < 0) {
fail_archive(matching, "archive_match_excluded");
}
return (excluded == 0);
}

static int has_include_descendant(const struct pattern_list *includes,
const char *path) {
size_t plen = strlen(path);
for (size_t i = 0; i < includes->len; i++) {
const char *pat = includes->items[i];
if (strncmp(pat, path, plen) == 0 && pat[plen] == '/') {
return (1);
}
}
return (0);
}

static char *join_prefix_path(const char *prefix, const char *path) {
if (prefix == NULL || prefix[0] == '\0' ||
(prefix[0] == '.' && prefix[1] == '\0')) {
char *dup = strdup(path);
if (dup == NULL) {
fail_errno("strdup");
}
return (dup);
}
size_t plen = strlen(prefix);
size_t olen = strlen(orig_copy);
size_t total = plen + 1 + olen + 1;
size_t path_len = strlen(path);
size_t total = plen + 1 + path_len + 1;
char *buf = malloc(total);
if (buf == NULL) {
free(orig_copy);
fail_errno("malloc");
}
memcpy(buf, prefix, plen);
buf[plen] = '/';
memcpy(buf + plen + 1, orig_copy, olen + 1);

archive_entry_set_pathname(e, buf);
int r = archive_match_excluded(match, e);
archive_entry_set_pathname(e, orig_copy);
free(buf);
free(orig_copy);
return (r);
memcpy(buf + plen + 1, path, path_len + 1);
return (buf);
}

static int contains_dotdot_segment(const char *path) {
Expand Down Expand Up @@ -582,7 +628,8 @@ int main(int argc, char **argv) {
const char *xar_path = NULL;
const char *outdir = NULL;
struct archive *xar;
struct archive *match = NULL;
struct archive *matching;
struct pattern_list includes = {0};
struct archive *disk;
struct archive_entry *e;
int r;
Expand All @@ -594,6 +641,11 @@ int main(int argc, char **argv) {
int strip_components = 0;
int flags;

matching = archive_match_new();
if (matching == NULL) {
fail_errno("archive_match_new");
}

while ((opt = pkg_getopt(&argc, &argv, &arg)) != -1) {
switch (opt) {
case 'f':
Expand All @@ -611,27 +663,14 @@ int main(int argc, char **argv) {
do_expand_full = 1;
break;
case opt_include:
if (match == NULL) {
match = archive_match_new();
if (match == NULL) {
fail_errno("archive_match_new");
}
}
r = archive_match_include_pattern(match, arg);
if (r != ARCHIVE_OK) {
fail_archive(match, "include pattern");
pattern_list_add(&includes, arg);
if (archive_match_include_pattern(matching, arg) != ARCHIVE_OK) {
fail_archive(matching, "archive_match_include_pattern");
}
break;
case opt_exclude:
if (match == NULL) {
match = archive_match_new();
if (match == NULL) {
fail_errno("archive_match_new");
}
}
r = archive_match_exclude_pattern(match, arg);
if (r != ARCHIVE_OK) {
fail_archive(match, "exclude pattern");
if (archive_match_exclude_pattern(matching, arg) != ARCHIVE_OK) {
fail_archive(matching, "archive_match_exclude_pattern");
}
break;
case opt_strip_components:
Expand Down Expand Up @@ -667,10 +706,6 @@ int main(int argc, char **argv) {
fail_errno("archive_read_new");
}

if (match != NULL) {
archive_match_set_inclusion_recursion(match, 1);
}

disk = archive_write_disk_new();
if (disk == NULL) {
fail_errno("archive_write_disk_new");
Expand Down Expand Up @@ -707,20 +742,29 @@ int main(int argc, char **argv) {
fail_errno("chdir(outdir)");
}

if (archive_match_set_inclusion_recursion(matching, 1) != ARCHIVE_OK) {
fail_archive(matching, "archive_match_set_inclusion_recursion");
}

while ((r = archive_read_next_header(xar, &e)) == ARCHIVE_OK) {
const char *p = archive_entry_pathname(e);
char *rel = normalize_rel_path(p);
archive_entry_set_pathname(e, rel);
if (match != NULL) {
r = archive_match_excluded(match, e);
if (r != 0) {
int is_nested = should_be_treated_as_nested_archive(rel);
if (do_expand_full && is_nested) {
char *logical_path = join_prefix_path(NULL, rel);
int include_nested = should_extract_path(matching, logical_path);
if (!include_nested && includes.len > 0 &&
has_include_descendant(&includes, logical_path)) {
include_nested = 1;
}
free(logical_path);
if (!include_nested) {
archive_read_data_skip(xar);
free(rel);
continue;
}
}
int is_nested = should_be_treated_as_nested_archive(rel);
if (do_expand_full && is_nested) {

char *nested_outdir = strip_components_path(rel, strip_components);
int nested_strip = strip_components;
int rel_components = path_component_count(rel);
Expand Down Expand Up @@ -749,12 +793,20 @@ int main(int argc, char **argv) {
.eof = 0,
};

extract_nested_archive_from_stream(&in, nested_outdir, flags, match,
extract_nested_archive_from_stream(&in, nested_outdir, flags, matching,
nested_strip, rel);
}
free(nested_outdir);
free(rel);
} else {
char *logical_path = join_prefix_path(NULL, rel);
if (!should_extract_path(matching, logical_path)) {
archive_read_data_skip(xar);
free(logical_path);
free(rel);
continue;
}
free(logical_path);
if (apply_strip_components(e, strip_components)) {
archive_read_data_skip(xar);
free(rel);
Expand All @@ -771,8 +823,7 @@ int main(int argc, char **argv) {

archive_write_free(disk);
archive_read_free(xar);
if (match != NULL) {
archive_match_free(match);
}
archive_match_free(matching);
pattern_list_free(&includes);
return (0);
}
54 changes: 54 additions & 0 deletions tests/BUILD.bazel
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,6 +64,32 @@ run_binary(
tags = ["manual"],
)

run_binary(
name = "pkgutil_component_expand_full_strip_6_usr_filter_action",
tool = "//:pkgutil",
srcs = [
"@component_pkg//file",
],
args = [
"--include",
"Payload/Library/Developer/CommandLineTools/SDKs/MacOSX15.5.sdk/usr/*",
"--include",
"Payload/Library/Developer/CommandLineTools/SDKs/MacOSX15.5.sdk/System/*",
"--exclude",
"Payload/Library/Developer/CommandLineTools/SDKs/MacOSX15.5.sdk/System/Library/Frameworks/Ruby.framework/Versions/2.6/Headers/ruby/ruby",
"--exclude",
"Payload/Library/Developer/CommandLineTools/SDKs/MacOSX15.5.sdk/usr/share/*",
"--expand-full",
"--strip-components",
"6",
"$(location @component_pkg//file)",
"$@",
],
out_dirs = ["pkgutil-component-expand-full-strip-6-usr-filter"],
testonly = True,
tags = ["manual"],
)

run_binary(
name = "pkgutil_product_expand_action",
tool = "//:pkgutil",
Expand Down Expand Up @@ -169,6 +195,34 @@ exec_test(
],
)

exec_test(
native_test,
name = "pkgutil_component_expand_full_strip_6_usr_filter_test",
src = ":test",
args = [
"-e",
"$(location :pkgutil_component_expand_full_strip_6_usr_filter_action)/usr",
"$(location :pkgutil_component_expand_full_strip_6_usr_filter_action)/System",
"$(location :pkgutil_component_expand_full_strip_6_usr_filter_action)/usr/lib/libNFC_HAL.tbd",
],
data = [
":pkgutil_component_expand_full_strip_6_usr_filter_action",
],
)

exec_test(
native_test,
name = "pkgutil_component_expand_full_strip_6_usr_filter_missing_test",
src = ":test",
args = [
"-ne",
"$(location :pkgutil_component_expand_full_strip_6_usr_filter_action)/usr/share",
],
data = [
":pkgutil_component_expand_full_strip_6_usr_filter_action",
],
)

exec_test(
native_test,
name = "pkgutil_product_expand_test",
Expand Down
Loading