Re-reverse engineered index meta information layout in JET 4
file format, based particularly on jackcess (Java) JET/Access
database library (https://github.com/jahlborn/jackcess), and
hexdumps of JET 4 database index meta information. Based both
on byte counting of jackcess reading of index definitions and
also expected flag values (0x80, 0x89, etc), the flags of JET 4
index definition are 4 bytes later than mdbtools previously thought.
See IndexData.read() and private static class Jet4Format in
src/main/java/com/healthmarketscience/jackcess/impl/IndexData.java
src/main/java/com/healthmarketscience/jackcess/impl/JetFormat.java
in jackcess source for layout reference.
Now appears to get sensible "CREATE INDEX"/"CREATE UNIQUE INDEX"
behaviour on export to PostgreSQL schema.
Also added extensive index definition byte layout reference to top
of src/libmdb/index.c for ease of reference, plus more debugging
assistance (and comments of phases reading index for readability).
Adds "-B" (--boolean-words) option to mdb-export, which will reconfigure
mdb/data.c to export TRUE/FALSE for boolean values instead of 1/0. The
option is needed to support BOOLEAN fields on PostgreSQL, which will not
implicitly cast bare 1/0 into a BOOLEAN value. Value literals are the
SQL TRUE/FALSE, and _quoted_ words meaning true/false and _quoted_
'1'/'0'. With this flag the SQL TRUE/FALSE values are output, which should
work with several SQL databases.
PostgreSQL Reference:
http://www.postgresql.org/docs/current/static/datatype-boolean.html
This adds support for reading OLE objects to the ODBC driver. The APIs
for reading OLE appear somewhat idiosyncratic, so we read the string
fully and stash it in a static variable.
Tested by reading an old Access database, with checking for memory leaks.
Most 64-bit environments use 64 bit longs and 32 bit ints; as such,
comparing an int to LONG_MIN/MAX makes no sense. Since SQL LONGs
appear limited to 32 bits, I assume INT_MIN/MAX was intended.
'gnome-doc-utils' is a rather large dependency on OS X, and the build
system has checks for gnome-doc-utils surrounding its use, but the
autoconf check errors out in its absence.
The memory for sql->bound_values is now freed by mdb_sql_reset, so these values no longer should be freed in dump_results_pp. Also, this fixes a memory leak that was occurring when results were not being pretty printed.
SQL engine was being reset in both dump_results functions. Rather than having several different functions responsible for resetting the engine depending on the circumstances, the reset now only occurs in run_query, the same function where the query is executed.
