…e types internal so they can be included as links from other project

…es.Extensions

…odifications

…der TFMs

….Resources.Extensions.BinaryFormat

This reverts commit 6354a91.

…xception and the stream is seekable

- add missing usings
- remove writer tests
- update stored blobs that had different assembly key
- use hardcoded assembly name as it's used in the resources
- move some of the tests to PayloadReader as they belong there

… be less or equal than the List array length)

…so it must return lengths for all dimensions

- remove env var
- remove redundant attribute annotation
- enable AOT analyzer and annotate the code, use Array.CreateInstanceFromArrayType when possible
- throw immediately once invalid input is read (to reduce the number of places we throw), add missing tests
- use Resources to get exception messages

Co-authored-by: Jan Kotas <jkotas@microsoft.com>

Co-authored-by: Jeremy Barton <jbarton@microsoft.com>

Co-authored-by: Jeremy Barton <jbarton@microsoft.com>

- move the files to match the existing source code path convention
- adjust namespaces (and base64 serialized objects that used the old namespace and library name)
- fix AOT annotation
- fix the build (by not using Windows-specific test data on other platforms)
- rename "Binary Format" and "Binary Formatter" to "NRBF"
- fix formatting
- throw ArgumentException for non-seekable Streams (don't expect that using Position will throw for them)

…tion on Mono formats stack trace in slightly different way)

- remove parts of BinaryFormattedObjectExtensions used only in WinForms
- remove FormatterConverterStub (not used here)
- remove hashing from tests
- add comment(s)
- improve XML doc comments
- rename ArrayType to BinaryArrayType and SerializedStreamHeaderRecord to SerializedStreamHeaderRecord to use same naming as spec everywhere
- rename ContainsBinaryFormatterPayload to StartsWithPayloadHeader and change it accordingly
- use "s_" prefix for static fields

Co-authored-by: Jeremy Barton <jbarton@microsoft.com>

Co-authored-by: Jeremy Barton <jbarton@microsoft.com>

- fix the build
- don't hardcode current default value for ToArray(maxLength) argument, make it Int64 for non-SZArrays as the max length is bigger than int.MaxValue
- add/fix comments
- specify the type in explicit way when using new operator
- local constants need to start with uppercase letter
- make ObjecId abstract (to make it more explicit which records don't provide Id on purpose)
- prefer "private protected" over "protected"
- remove opinionated restriction for ObjectNullMultiple256Record
- don't use seealso
- rename Parse to Decode where input is binary
- add argument validaiton for public methods
- improve exception messages

… at the same time (don't pre-allocate large buffers based on size read from payload when payload does not contain enough data)

…, remove maxLength argument

…w for large multi-dimensional arrays

Co-authored-by: Jeremy Barton <jbarton@microsoft.com>

- prefer NET over NETCOREAPP when writing #if defines
- add ConditionalTheory to make the SkipTestException actually work
- ObjectNullMultiple must not allow for 0 inputs
- preallocate the List<T> for TFMs other than .NET

- ensure that reading arrays of primitive types is as fast as possible when there is enough data in the stream, throw when there is not, fall back to slow path when we don't know
- make GetArray reuse the previously created instance
- and provide tests for all of that

…tes as what is stored in the test source code. The blobs may not be identical (the output is not deterministic), but still valid.