Rename PCRE2_EXTRA_ALLOW_SURROGATE_ESCAPES to PCRE2_EXTRA_ALLOW_SURROGATESso it can be used when characters with those codes are used even when notescaped.  This also resolves the ambuiguity that allowed the following:  PCRE2 version 10.34 2019-11-21    re> /\p{Cs}/match_invalid_utf  data> \x{dfff}\=no_utf_check  No matchSeveral interrelated bug fixes:* An UTF-8 or UTF-32 encoded character with a surrogate codepoint should be  allowed with PCRE2_EXTRA_ALLOW_SURROGATES.  Before this change non UTF-16  will not allow surrogates if encoded in the pattern directly.* In ALT_BSUX mode, any escaped character should represent itself if not  "special", but that was not the case for not ASCII characters.* When match_invalid_utf is enabled a surrogate character in the subject was  being skipped, even if no_utf_check was used together with allow_surrogatesIssues that would be addressed with a new version* the documentation around \o and \x, specially when bigger than 255 needs  reviewing for clarification.* pcre2test support for testing escaped UTF characters is suboptimal, so most  testing was done with custom code that should be integrated instead.* JIT is broken when match_invalid_utf and allow_surrogate.Issues that are still open and had been punted from this version:* ALT_BSUX allows escape characters up to \x{ffff} even without PCRE2_UTF.* PCRE2_ALT_BSUX might assume only UTF-16. It is not clear what 0xf1 might  mean (\u00F1, assume binary matching like it seems to do with surrogates  and match with an extended  0xf1 PCRE2_UCHAR, or error), if the later is  something like PCRE2_EXTRA_ALLOW_SURROGATE_ESCAPES needed?.* PCRE2_ALT_BSUX match shouldn't really do UTF matching when using surrogates.* PCRE2_EXTRA_ALT_BSUX should restrict the use of \, so for example non ASCII  or \U throw an error.  It probably need to break with PCRE2_ALT_SUX.* Behaviour is different between UTF-16 and UTF-32 but tests are still unified  so tests are missing.