Encode filepaths to be safe on Windows

This changes the file paths to use characters that are safe for Windows.

Channges the output of the following:
* Package cache directory
* Generated pkl-doc files
* Kotlin generated code

Unsafe characters are encoded as (<hex>).
For example, the colon character `:` is encoded as `(3a)`.

Additionally, this changes the cache directory prefix (package-1 to
package-2).

Follows the design of https://github.com/apple/pkl-evolution/pull/3
This commit is contained in:
Dan Chao
2024-04-26 07:34:31 -07:00
committed by Daniel Chao
parent 110dc89e86
commit a5c13e325a
36 changed files with 185 additions and 106 deletions

View File

@@ -413,7 +413,7 @@ final class PackageResolvers {
private final Path tmpDir;
private static final String CACHE_DIR_PREFIX = "package-1";
private static final String CACHE_DIR_PREFIX = "package-2";
@GuardedBy("lock")
private final EconomicMap<PackageUri, FileSystem> fileSystems = EconomicMaps.create();
@@ -438,12 +438,14 @@ final class PackageResolvers {
return path;
}
var checksumIdx = path.lastIndexOf("::");
return path.substring(0, checksumIdx);
return IoUtils.encodePath(path.substring(0, checksumIdx));
}
private Path getRelativePath(PackageUri uri) {
return Path.of(
CACHE_DIR_PREFIX, uri.getUri().getAuthority(), getEffectivePackageUriPath(uri));
CACHE_DIR_PREFIX,
IoUtils.encodePath(uri.getUri().getAuthority()),
getEffectivePackageUriPath(uri));
}
private String getLastSegmentName(PackageUri packageUri) {

View File

@@ -597,6 +597,30 @@ public final class IoUtils {
return newUri;
}
/**
* Windows reserves characters {@code <>:"\|?*} in filenames.
*
* <p>For any such characters, enclose their decimal character code with parentheses. Verbatim
* {@code (} is encoded as {@code ((}.
*/
public static String encodePath(String path) {
if (path.isEmpty()) return path;
var sb = new StringBuilder();
for (var i = 0; i < path.length(); i++) {
var character = path.charAt(i);
switch (character) {
case '<', '>', ':', '"', '\\', '|', '?', '*' -> {
sb.append('(');
sb.append(ByteArrayUtils.toHex(new byte[] {(byte) character}));
sb.append(")");
}
case '(' -> sb.append("((");
default -> sb.append(path.charAt(i));
}
}
return sb.toString();
}
private static int getExclamationMarkIndex(String jarUri) {
var index = jarUri.indexOf('!');
if (index == -1) {

View File

@@ -431,4 +431,14 @@ class IoUtilsTest {
IoUtils.readString(URI("http://example.com").toURL())
}
}
@Test
fun `encodePath encodes characters reserved on windows`() {
assertThat(IoUtils.encodePath("foo:bar")).isEqualTo("foo(3a)bar")
assertThat(IoUtils.encodePath("<>:\"\\|?*")).isEqualTo("(3c)(3e)(3a)(22)(5c)(7c)(3f)(2a)")
assertThat(IoUtils.encodePath("foo(3a)bar")).isEqualTo("foo((3a)bar")
assertThat(IoUtils.encodePath("(")).isEqualTo("((")
assertThat(IoUtils.encodePath("3a)")).isEqualTo("3a)")
assertThat(IoUtils.encodePath("foo/bar/baz")).isEqualTo("foo/bar/baz")
}
}