A sample file provided in #213 has an xref table where at least one of the byte offsets is a few bytes too low. When we jump to that offset, we don't find the start of the object as we expect.
However, the following token in the file is the start of the expected object. Other readers I tested with seem to handle this particular corruption gracefully, and now we do too.
A sample file provided in #213 has an xref table where at least one of the byte offsets is a few bytes too low. When we jump to that offset, we don't find the start of the object as we expect.
However, the following token in the file is the start of the expected object. Other readers I tested with seem to handle this particular corruption gracefully, and now we do too.
Fixes #213