Allow whitespace and comments in identifiers when parsing

[paul35621]

It is allowed in Protocol Buffers to have whitespace and comments around the dots in identifiers. This is currently (version 5.2.1) not supported in the parser. It would be nice to have this.

As an example, consider the file google/ads/googleads/v18/common/user_lists.proto from the googleapis GitHub repository. Below the part of the file where it goes wrong. The identifier google.ads.googleads.v18.enums.LookalikeExpansionLevelEnum.LookalikeExpansionLevel is split over multiple lines.

message LookalikeUserListInfo {
  // Seed UserList ID from which this list is derived, provided by user.
  repeated int64 seed_user_list_ids = 1;

  // Expansion level, reflecting the size of the lookalike audience
  google.ads.googleads.v18.enums.LookalikeExpansionLevelEnum
      .LookalikeExpansionLevel expansion_level = 2;

  // Countries targeted by the Lookalike. Two-letter country code as defined by
  // ISO-3166
  repeated string country_codes = 3;
}

Parsing this file fails with error Syntax error in user_lists.proto:49:33: expected '=' but was 'e'. (line 49 is the line that starts with .LookalikeExpansionLevel)

[oldergod]

Yep. I tried to address that a while ago but I couldn't do it as fast as I wanted and gave up at the time.