Le vendredi 13 décembre 2013, 16:57:06 Thomas Preud'homme a écrit : > > So I'm not sure about why there is only a single backslash for bytea values > as string litteral but two backslashes are clearly necessary. Indeed, the > two backslashes are first interpreted by the string litteral parser which > leads to a single escape in the result, something of the form '\x...'. Then > this is parsed by the bytea input which transform it into a sequence of > bytes. > > Adrien, are you familiar with gdb? If yes, would you mind helping me debug > this issue? We could set up a meeting on IRC which would be the most > convenient way to do it or do it over email (that would be a bit slower but > allow for more asynchronism between us).
That might not be necessary. I checked libpq's code (the C library to use postgresql from a program) of the function PQescapeByteaInternal and the amount of backslashed depends on whether standard_conforming_strings is set to on or off. Since version -12 of dspam, the default value (on) is used while the value was set to off (by mistake) in version -11. That explains why the bug appears. I just need to understand why two backslashes are not present in both cases and I will be able to fix this bug. So please be patient, this issue should be resolved soon. Best regards, Thomas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
