https://zarr.readthedocs.io/en/stable/api/storage.html

• FileSystemStore
• Nested Directory Store (can be a strategy of FileSystemStore) https://zarr.readthedocs.io/en/stable/api/storage.html#zarr.storage.NestedDirectoryStore
• TempStore https://zarr.readthedocs.io/en/stable/api/storage.html#zarr.storage.TempStore
• Zip Store https://zarr.readthedocs.io/en/stable/api/storage.html#zarr.storage.ZipStore

Currently in ZarrArray.java, the findNestedChunks method is implemented as:

 private static boolean findNestedChunks(ZarrPath relativePath, Store store, int[] chunks) throws IOException {
        final String oneOrMoreDigits = "\\d+";
        final StringBuilder sb = new StringBuilder(oneOrMoreDigits);
        for (int i = 0; i < chunks.length - 1; i++) {
            sb.append(DimensionSeparator.SLASH.getSeparatorChar());
            sb.append(oneOrMoreDigits);
        }
        final String regex = sb.toString();
        final Stream<String> keys = store.getRelativeLeafKeys(relativePath.storeKey);
        final boolean nestedChunks = keys.anyMatch(s -> s.matches(regex));
        return nestedChunks;
    }

The java.util.stream.Stream from store.getRelativeLeafKeys() is not wrapped in a try-with-resources block. Since the source of this stream is an interface, it is not known if closing the stream is required or not.

It should be as follows:

    private static boolean findNestedChunks(ZarrPath relativePath, Store store, int[] chunks) throws IOException {
        final String oneOrMoreDigits = "\\d+";
        final StringBuilder sb = new StringBuilder(oneOrMoreDigits);
        for (int i = 0; i < chunks.length - 1; i++) {
            sb.append(DimensionSeparator.SLASH.getSeparatorChar());
            sb.append(oneOrMoreDigits);
        }
        final String regex = sb.toString();
        try (final Stream<String> keys = store.getRelativeLeafKeys(relativePath.storeKey)){
            final boolean nestedChunks = keys.anyMatch(s -> s.matches(regex));
            return nestedChunks;
        }
    }

Default Implementation = ThreadSynchronizer
Additional:
NullSynchronizer, ProzessSynchronizer, CombinedSynchronizer

• Synchronizer Interface
• NullSynchronizer
• ThreadSynchronizer
• ProcessSynchronizer
• CombinedSynchronizer

Hi - thanks for this library! I tried to use the sample POM extract in https://jzarr.readthedocs.io/en/latest/ but it didn't work. I found I needed to use https:// for the repo URL, i.e. https://nexus.senbox.net/nexus/content/groups/public/.

Also I wasn't able to use 0.3.2-SNAPSHOT as there is no POM for this, so I changed it to 0.3.5. This snippet worked for me:

    <dependencies>
        <dependency>
            <groupId>com.bc.zarr</groupId>
            <artifactId>jzarr</artifactId>
            <version>0.3.5</version>
        </dependency>
    </dependencies>

    <repositories>
        <repository>
            <id>bc-nexus-repo</id>
            <name>Brockmann-Consult Public Maven Repository</name>
            <url>https://nexus.senbox.net/nexus/content/groups/public/</url>
        </repository>
    </repositories>

This probably isn't a JZarr problem per se, but I'm struggling to get my Java program to recognise my native c-blosc library.

I'm on an M1 Mac (which might be significant), using JDK17 with NetBeans, using maven as a build system. The c-blosc library was installed with Homebrew.

I've written a simple program and am trying to run it using the command-line switch -Djna.library.path=/opt/homebrew/Cellar/c-blosc/1.21.1/. I only need blosc at runtime, not compile time.

Whatever I try, I end up with:
Exception in thread "main" java.lang.UnsatisfiedLinkError: /private/var/folders/s2/_3zsgdqd6pj0y20x4chv8g640000gn/T/jna-105417/jna11531817638949407677.tmp: dlopen(/private/var/folders/s2/_3zsgdqd6pj0y20x4chv8g640000gn/T/jna-105417/jna11531817638949407677.tmp, 1): no suitable image found.

Are there any more details you can provide about how to set this up in a maven-based system?

I'm wondering whether there may be an issue with the M1 processor (which is ARM based). The output of java -version is:

openjdk version "17.0.1" 2021-10-19 LTS
OpenJDK Runtime Environment Zulu17.30+15-CA (build 17.0.1+12-LTS)
OpenJDK 64-Bit Server VM Zulu17.30+15-CA (build 17.0.1+12-LTS, mixed mode, sharing)

Start a thread (potential feature request) to discuss the options of persisting Zarr arrays/groups to Google Cloud GCS storage bucket. Is there an existing feature or roadmap to support this integration?
@SabineEmbacher

I am trying to read a ome-zarr files available on embassy
https://s3.embassy.ebi.ac.uk/idr/zarr/v0.1/6001247.zarr

I wrote a little client download the zarr locally and I attempted to read the file using jzarr v0.3
I got the following error when reading the file:

com.fasterxml.jackson.databind.exc.InvalidDefinitionException: Cannot construct instance of `com.bc.zarr.ZarrHeader` (no Creators, like default construct, exist): cannot deserialize from Object value (no delegate- or property-based Creator)
 at [Source: (BufferedReader); line: 1, column: 2]
        at com.fasterxml.jackson.databind.exc.InvalidDefinitionException.from(InvalidDefinitionException.java:67)
        at com.fasterxml.jackson.databind.DeserializationContext.reportBadDefinition(DeserializationContext.java:1592)
        at com.fasterxml.jackson.databind.DeserializationContext.handleMissingInstantiator(DeserializationContext.java:1058)
        at com.fasterxml.jackson.databind.deser.BeanDeserializerBase.deserializeFromObjectUsingNonDefault(BeanDeserializerBase.java:1297)
        at com.fasterxml.jackson.databind.deser.BeanDeserializer.deserializeFromObject(BeanDeserializer.java:326)
        at com.fasterxml.jackson.databind.deser.BeanDeserializer.deserialize(BeanDeserializer.java:159)
        at com.fasterxml.jackson.databind.ObjectMapper._readMapAndClose(ObjectMapper.java:4218)
        at com.fasterxml.jackson.databind.ObjectMapper.readValue(ObjectMapper.java:3227)
        at com.bc.zarr.ZarrUtils.fromJson(ZarrUtils.java:104)
        at com.bc.zarr.ZarrArray.open(ZarrArray.java:69)
        at com.bc.zarr.ZarrArray.open(ZarrArray.java:60)
        at com.bc.zarr.ZarrArray.open(ZarrArray.java:56)
        at com.bc.zarr.ZarrArray.open(ZarrArray.java:52)
        at com.example.Zarr.readZarrFile(Zarr.java:193)

When creating a second ZipStore for the same archive a FileSystemAlreadyExistsException is thrown

cc @joshmoore

While running a branch of zarr_implementations with jzarr-based read/write tests (with jzarr 0.3.2), I encountered the following issue.

The data used in the example in that repository has an 8-bit unsigned integer format. The .zattr JSON files created by jzarr currently have the following entry

"dtype" : ">u1",

but .zattr files created by other libraries such as zarr-python, xtensor-zarr and z5py have:

"dtype" : "|u1",

It seems that the spec indicates that | should be used rather than > or < for data types with a single byte width where there is no endianness. A similar issue may also be present for |b1 and |i1 types, but I have not confirmed that.

At @glencoesoftware we're very excited about the work going on with this library and in collaboration with @ome we are looking to add it to a variety of tools we support pursuant to our collective NGFF [1] goals in biomedical imaging. These tools are both open and closed source. A couple of weeks ago in fe2726e license headers were added to all source code specifying GPLv3 or later. The LICENSE at the root of the project remains MIT.

Is it the project's intention to relicense versions after 0.3.1 as GPLv3 or later?

1. https://ngff.openmicroscopy.org/latest/

From zarr-developers/community#15 (comment)

• prove groupId
• all dependencies in maven central

@see zarr.storage._init_array_metadata
@see zarr.util.normalize_chunks

• Auto compute chunk size if chunks==null and chunked==true
• If chunked==false , chunks must be equal to shape (SingleChunk)
• Exceptions like python if chunk dimensions or other things are wrong. @see zarr.util.normalize_chunks

mode:{‘r’, ‘r+’, ‘a’, ‘w’, ‘w-‘}, optional
Persistence mode: ‘r’ means read only (must exist); ‘r+’ means read/write (must exist); ‘a’ means read/write (create if doesn’t exist); ‘w’ means create (overwrite if exists); ‘w-‘ means create (fail if exists).

When a file path is converted to an URL in order to read a zipped ZNAP file in SNAP an IllegalArgumentException is thrown if this path contains a space.

java.net.URISyntaxException: Illegal character in opaque part at index 71: jar:file:/C:/Users/<snip>/RSEO content/S2A_MSIL2A_20230227T160221_N0509_R097_T17QLF_20230227T225659_subset20m.znap.zip at java.net.URI$Parser.fail(URI.java:2848) at java.net.URI$Parser.checkChars(URI.java:3021) at java.net.URI$Parser.parse(URI.java:3058) at java.net.URI.<init>(URI.java:588) at java.net.URI.create(URI.java:850) Caused: java.lang.IllegalArgumentException: Illegal character in opaque part at index 71: jar:file:/C:/Users/<snip>/RSEO content/S2A_MSIL2A_20230227T160221_N0509_R097_T17QLF_20230227T225659_subset20m.znap.zip at java.net.URI.create(URI.java:852) at com.bc.zarr.storage.ZipStore.<init>(ZipStore.java:57) at org.esa.snap.dataio.znap.ZnapProductWriter.writeProductNodesImpl(ZnapProductWriter.java:255) at org.esa.snap.core.dataio.AbstractProductWriter.writeProductNodes(AbstractProductWriter.java:111)

Before the JAR URI is created the URL needs probably be decoded.

Hello, I'm currently using Jzarr and everything works fine
But I discovered that filters (as well as other things) are not implemented yet, so I need to create my own filter manually
Is this going to be available soon ?

Same question for the addition of other compressors, Zlib and Blosc are great but I want to try others just like in the Python library
In the doc it's written "More compressors will be implemented in the future." but since this project has seen no evolution since 1-2 years it might just be abandoned, and I hope it's still alive

Looking at this zlib compressor implementation code, the compression implementation seems to be a simple passthrough implemented here.
Maybe we should use JDK deflate/inflate classes or Apache common compression library.

Implement a chunk name factory to allow nested chunk files.
Such a factory then can be set to and used by any com.bc.zarr.storage.Store implementation to create the name and or path of a chunk file.

Jzarr is implemented in a thread save way.
A process synchronizer shall be implemented.
A process synchronizer then can be used in combination or instead of a thread synchronizer.