implementation 99% complete

everything and tested except for repeatable shuffling

tested most basic/typical split usecases

cleanup

some more cleanup

fix CI

more ci fix

more ci fixes

more ci fix

more ci fix

more ci fix

added more tests, fixed some bugs

some more clean up and test cases

added shard/shuffle before split warning/error

addressed code review comments and ci

fixed ci

Merge pull request !1192 from h.farahat/string_Tensor

Merge pull request !1129 from zhaozhenlong/op/bounding-box-encoder

Merge pull request !1203 from changzherui/code516

Merge pull request !1176 from jiangjinsheng/issue_transpose

Merge pull request !1195 from jiangjinsheng/issue_geswitch

Merge pull request !1100 from jinyaohui/context_opt_2

iou

NMSWithMask

larsupdate testcase

sgd testcase

Merge pull request !1178 from zhouneng/add_vm_support_roialign

-replace switch'case by indexing

- Add test case
- Add constructors
- Add getItem string

- Fix bugs
- Add more tests

- Tensor iterator
- asNumpy
- TextFileDataset

- Tensor(Numpy)

- Super > 2D
- Add more test cases for GeneratorDataset

- Change StartAddr to GetBuffer and GetMutableNuffer

- Raise an error if batch is used with strings

Clean-up work

Merge pull request !1157 from ms_yan/concat_dataset

Merge pull request !1126 from xiefangqi/voc_support_detection

change to use __add__ operation instead ds.concat

Merge pull request !1171 from fary86/adjust_onnx_related_uts

Merge pull request !1015 from candanzg/me_with_shape

Merge pull request !1026 from peixu_ren/custom_pp_ops

Merge pull request !1084 from SanjayChan/04quant

Merge pull request !1089 from ghzl/add-get-lr-base-on-parameter

Merge pull request !1069 from 李鸿章/context_manager

Merge pull request !1094 from zhouneng/add_vm_support_l2normalize

Signed-off-by: candanzg <zhangshucheng@huawei.com>

Merge pull request !1121 from lihongkang/master

Merge pull request !999 from liyong126/mindrecord_file_list

Merge pull request !1042 from jinyaohui/clean_pylint

Merge pull request !1082 from hewei/support_cont_break

To handle 'break' and 'continue' statement, a loop context is pushed
to a stack before we parse the loop body, and pop it after body parsed.
When a 'break', 'continue' statement is encountered, we retrieve current
loop contex from the stack, and let the current block jump to the end
block or header block;

For 'break' statement, we added an extra 'end_block' follow the 'after_block',
because 'after_block' is called from a ContionalJump in 'header_block', it can
not be set as jump target from other place. to support 'break', we let loop
body jump to the 'end_block' at the 'break' point. and 'after_block'
maybe a good place to handle loop 'else' clause in the future.

Handle 'pass' is simple, just bypass it when doing parse.

Merge pull request !1110 from lichen/fix_gatherv2_and_dataset_bug

Merge pull request !1038 from jinyaohui/clean_context