Make FPDF_GetText stricter on inputs

The current implementation of this function is problematic. It will
attempt to memcpy to NULL. It will accept obviously wrong inputs like
a negative start index. It will also accept -1 for the count, which in
theory is the amount of space the buffer has allocated to it, so
doesn't make sense, but instead an internal call will calculate the
number of characters to get if the count is -1. This will them lead to
the function attempting to call Left(-1) on a string, which is
invalid.

Ths documentation for this function mentions none of this behaviour,
so I am removing it, since it is inconsistent/bad. The implementation
should now more strictly meet defined API.

BUG=pdfium:828

Change-Id: I18afdb33e12d77c10d856b4bacd615481979c484
Reviewed-on: https://pdfium-review.googlesource.com/12733
Commit-Queue: Ryan Harrison <rharrison@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>

0 files changed, 0 insertions, 0 deletions