Pdfreader

---------------------------
- Issue 67: Canvas cache for viewer
- Issue 71: Unsupported security handler version: 1
- Issue 72: Add a method to return document metadata
- Issue 68: Work around wrong stream `Length`

---------------------------
- Issue 63: SimpleCanvas.fromObject images copying issue fixed
- Issue 64: Allow indirect stream objects have missing `endobj`.

---------------------------
- Issue 60: Python 3.9 support
- Issue 59: Use `pycryptodome` instead of `pycrypto`
- Issue 31: Canvas and page iterators added to PDFViewver

---------------------------
- Missing `Resources.Properties` field added.
- Pillow >= 7.1.0 version bump (security issues)
- Issue 51: Current page annotations & hyperlinks.
- Issue 33: Encryption support

---------------------------
- Issue 41: RecursionError: maximum recursion depth exceeded
- Issue 39: Pillow version bump

---------------------------
- Issue 34: Stream Length can be Indirect Reference now.
- Issue 29: Optional Type=Encoding key for Encoding dictionary
- Issue 27: Empty XObject resources cause AttributeError
- Issue 25: pdfreader.exceptions.ParserException: Dict, stream or hexstring expected
- Issue 23: BI/EI: TypeError: can't concat bytes to NoneType
- Issue 18: pdfreader.parsers.cmap.CMapParserException: Regular non-digit character expected
- Issue 17: NotImplementedError: DCTDecode
- Issue 16: RecursionError: maximum recursion depth exceeded