[protobuf] String literal character set

[Michael Powell]

Hello,

When considering the String Literal character set, there is the
regular expression /[^\0\n\\]/, meaning virtually any other character
than those specified in the expression. That's in addition to the
several or so alphanumeric alternatives preceding that alternative.

As far as I can determine, that's virtually any character, some of
which are escaped in the string itself. Does this sound about right?

I'm just trying to get a handle on how best to construct the character
set when building my parser test case generators.

http://developers.google.com/protocol-buffers/docs/reference/proto2-spec

Thanks!

Michael Powell

[Michael Powell]

Post by Michael Powell
Hello,
When considering the String Literal character set, there is the
regular expression /[^\0\n\\]/, meaning virtually any other character
than those specified in the expression. That's in addition to the
several or so alphanumeric alternatives preceding that alternative.
As far as I can determine, that's virtually any character, some of
which are escaped in the string itself. Does this sound about right?

Escape characters themselves are pretty easy to add as a range of characters.

The troublesome cases IMO are how to discern between a hexEscape and
an octEscape, at least as a character. Makes this somewhat of a unique
"character set" issue. Not sure it's possible as such, but I'm open to
suggestions.

Post by Michael Powell
I'm just trying to get a handle on how best to construct the character
set when building my parser test case generators.
http://developers.google.com/protocol-buffers/docs/reference/proto2-spec
Thanks!
Michael Powell