Summary: 59 instances, 50 unique

Text	Count
// TODO assert buffers and lengths size	1
// TODO (yongtang): Support 2.0 which use 4 bytes for length.	1
// TODO: non-resource manager  case	1
// TODO: add support for max = 255 and dtype = uint16; need test file	1
# TODO length may not be at axis=0, if `batch` (axis=0) is present.	1
// TODO: Wire up offset, setting, seeking etc.  note, may only be possible to	1
// TODO: set optional memory to nullptr until input arg is added	1
// Make the Array chunk, TODO null values not currently supported	1
// TODO: DO NOT use a hardcoded path	1
// TODO: Validate filters etc, no nesting, no name conflict in shorthand	1
// TODO: We need to relax the restriction as we want to use `gse://`	1
# TODO handle tuples and flat datasets as well.	1
# TODO: Support multiple inv spectrograms	1
// TODO: pass dtype from Init().	1
// TODO: reference after first?	1
# TODO use base_type() to ensure consistent conversion.	1
// TODO: Check for nested filters and throw error, e.g. disallow	1
// TODO: 0 bytes requested when message body length == 0	1
// TODO: Handle buffer_size = 0 in 2.0 since InputStreamInterface has seek	1
// TODO: Enable HTTP/2.	1
// TODO: replace dims up until rank(start|stop)	3
// TODO: Avro is sync point partitioned and each block is very similiar to	1
// TODO: verify tensor is correct shape, arrow array is within bounds	1
// TODO: add support for max = 65535 and dtype = uint8; need test file	1
// TODO: break if option provided, as no records were returned then	1
# TODO: validate lower == upper except for axis	1
// TODO: LRU cache with adjustable size?	1
# TODO: Use resource to reuse schema initialization	2
# TODO: change to tf.while_loop	2
// TODO null count == 0	1
# TODO: Support audio with channel > 1.	4
// TODO: Much of the logic could be pushed to Golang by passing	1
// TODO: support batch (1-D) input	1
// TODO: Need to enhance the ordering constraint, the below algorithm won't do	1
// Make array data for the list TODO null count == 0	1
// TODO: adjust additional parameter with:	1
// TODO: compression level could be a input tensor node passed in.	1
TODO: Add additional fields besides orientation.	1
# TODO: This file was not generated successfully due to some errors	1
# TODO: support batch (1-D) input	1
// TODO (yongtang): Set channels_ = 4 for now.	2
// TODO: look into why 0 bytes are requested	1
// TODO: look into removing this loop, move tensor instead of copy	1
// TODO: Check that dense shape matches dimensions, handle cases where the	1
// TODO: support batch (1-D) input	2
constexpr int kIOBufferSize = 4096;  // TODO: maybe make this a parameter.	1
// TODO: Log warning here that the writer schema is used for reading	1
// TODO: parquet is RowGroup based so ideally the RowGroup should be	1
// TODO: This is not invoked, memory leak?	1
# TODO: manually check all errors and filter out the unsuccessful ones	1