关于 c : 如何更改 Google 的 protobuf 中的缓冲区限制？

  // Sets the maximum number of bytes that this CodedInputStream will read
  // before refusing to continue.  To prevent integer overflows in the
  // protocol buffers implementation, as well as to prevent servers from
  // allocating enormous amounts of memory to hold parsed messages, the
  // maximum message length should be limited to the shortest length that
  // will not harm usability.  The theoretical shortest message that could
  // cause integer overflows is 512MB.  The default limit is 64MB.  Apps
  // should set shorter limits if possible.  If warning_threshold is not -1,
  // a warning will be printed to stderr after warning_threshold bytes are
  // read.  For backwards compatibility all negative values get squashed to -1,
  // as other negative values might have special internal meanings.
  // An error will always be printed to stderr if the limit is reached.
  //
  // This is unrelated to PushLimit()/PopLimit().
  //
  // Hint:  If you are reading this because your program is printing a
  //   warning about dangerously large protocol messages, you may be
  //   confused about what to do next.  The best option is to change your
  //   design such that excessively large messages are not necessary.
  //   For example, try to design file formats to consist of many small
  //   messages rather than a single large one.  If this is infeasible,
  //   you will need to increase the limit.  Chances are, though, that
  //   your code never constructs a CodedInputStream on which the limit
  //   can be set.  You probably parse messages by calling things like
  //   Message::ParseFromString().  In this case, you will need to change
  //   your code to instead construct some sort of ZeroCopyInputStream
  //   (e.g. an ArrayInputStream), construct a CodedInputStream around
  //   that, then call Message::ParseFromCodedStream() instead.  Then
  //   you can adjust the limit.  Yes, it’s more work, but you’re doing
  //   something unusual.