Only use xref fallback if `PDFNoValidXRef` is raised and `fallback` is True (#684)

* check obj type

* update changelog

* Update CHANGELOG.md

* add changes

* update change

* update changelog

* Use fallback in except clause

* Update changelog.md

Co-authored-by: Pieter Marsman <pietermarsman@gmail.com>
Co-authored-by: Tony Tong <baojia.tong@kensho.com>

CHANGELOG.md

@@ -15,6 +15,10 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 - Add handling of JPXDecode filter to enable extraction of images for some pdfs ([#645](https://github.com/pdfminer/pdfminer.six/pull/645))
 - Fix extraction of jbig2 files, which was producing invalid files ([#652](https://github.com/pdfminer/pdfminer.six/pull/653))
 - Crash in `pdf2txt.py --boxes-flow=disabled` ([#682](https://github.com/pdfminer/pdfminer.six/pull/682))
+- Only use xref fallback if `PDFNoValidXRef` is raised and `fallback` is True ([#684](https://github.com/pdfminer/pdfminer.six/pull/684))
+
+### Changed
+- Replace warnings.warn with logging.Logger.warning in line with [recommended use](https://docs.python.org/3/howto/logging.html#when-to-use-logging) ([#673](https://github.com/pdfminer/pdfminer.six/pull/673))
 
 ## [20211012]
 
@@ -41,7 +45,6 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 - Support for Python 3.4 and 3.5 ([#522](https://github.com/pdfminer/pdfminer.six/pull/522))
 - Unused dependency on `sortedcontainers` package ([#525](https://github.com/pdfminer/pdfminer.six/pull/525))
 - Support for non-standard output streams that are not binary ([#523](https://github.com/pdfminer/pdfminer.six/pull/523))
-- Replace warnings.warn with logging.Logger.warning in line with [recommended use](https://docs.python.org/3/howto/logging.html#when-to-use-logging) ([#673](https://github.com/pdfminer/pdfminer.six/pull/673))
 - Dependency on typing-extensions introduced by [#661](https://github.com/pdfminer/pdfminer.six/pull/661) ([#677](https://github.com/pdfminer/pdfminer.six/pull/677))
 
 ## [20201018]

pdfminer/pdfdocument.py

@@ -11,7 +11,7 @@ from cryptography.hazmat.primitives.ciphers import Cipher, algorithms, modes
 from . import settings
 from .arcfour import Arcfour
 from .pdfparser import PDFSyntaxError, PDFParser, PDFStreamParser
-from .pdftypes import DecipherCallable, PDFException, PDFTypeError, PDFStream,\
+from .pdftypes import DecipherCallable, PDFException, PDFTypeError, PDFStream, \
     PDFObjectNotFound, decipher_all, int_value, str_value, list_value, \
     uint_value, dict_value, stream_value
 from .psparser import PSEOF, literal_name, LIT, KWD
@@ -706,12 +706,12 @@ class PDFDocument:
             pos = self.find_xref(parser)
             self.read_xref_from(parser, pos, self.xrefs)
         except PDFNoValidXRef:
-            pass  # fallback = True
+            if fallback:
-        if fallback:
+                parser.fallback = True
-            parser.fallback = True
+                newxref = PDFXRefFallback()
-            newxref = PDFXRefFallback()
+                newxref.load(parser)
-            newxref.load(parser)
+                self.xrefs.append(newxref)
-            self.xrefs.append(newxref)
+
         for xref in self.xrefs:
             trailer = xref.get_trailer()
             if not trailer: