Support GCS file pattern in grain

[aireenmei]

Description

ArrayRecord Reader will support reading from GCS directly in the coming releases. We update the file_pattern matching logic to include GCS path

Tests

Tested in V5p, steps:

1. Uninstall array-record, and re-install from a pre-released wheel
2. Run maxtext:

python3 -m MaxText.train MaxText/configs/base.yml \
  run_name=test/arrayrecord_gcs2 \
  base_output_directory=gs://aireenmei-multipod/test/ \
  per_device_batch_size=1 \
  dataset_type=grain \
  grain_file_type=arrayrecord \
  grain_train_files="gs://aireenmei-us-central1/maxtext-dataset/array-record/c4/en/3.0.4/c4-train2.tfrecord-0000*-of-01024.array_record,0.4;gs://aireenmei-us-central1/maxtext-dataset/array-record/c4/en/3.0.4/c4-train2.tfrecord-0001*-of-01024.array_record,0.6" \
  steps=5 \
  enable_checkpointing=false

Checklist

Before submitting this PR, please make sure (put X in square brackets):

• I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have run end-to-end tests tests and provided workload links above if applicable.
• I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.