GitLab

FormDataParserMultipart contract says that "*OCTET" part of body is read
 and passed as string of bytes. In case header has filename, octet part
 is ignored - store filename instead.
Actual code doesn't implement it. It led to situation when octet data
 were read and were put as string to base::Value. That led to DCHECK as far
 as string value must be utf-8 string.
The same problem occurs with FormDataParserUrlEncoded, where encoded part
 can be as utf-8 string or binary data.

Bug: 813561
Change-Id: I453e5767e44334535253d1bd3fa4e857c2d3a3ff
Reviewed-on: https://chromium-review.googlesource.com/910848Reviewed-by: Devlin <rdevlin.cronin@chromium.org>
Reviewed-by: Vaclav Brozek <vabr@chromium.org>
Commit-Queue: Devlin <rdevlin.cronin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#539907}