Added
* Entry point py3odb with commands: dump, geopoints, and query.
* Row objects (immutable dictionary-like, iterable sequences).
* Connection objects now have filename attribute.
* Cursor execution operations now support \<odb> tag in place of filename:
cur.execute("SELECT DISTINCT varno FROM <odb>")
* Constants submodule with ColumnType enum and Varno static class.
* Context-manager Reader class for iterating single cursor executions:
with Reader(filename, "SELECT lat,lon,obsvalue FROM <odb>") as odb_reader:
for row in odb_reader:
print(row)
Changed
* Cursor fetches now return Row objects.
* Cursor metadata now stores column type as ColumnType.
* Split code coverage reports into unit/integration.
Fixed
* Suppressed (by default) STDOUT messages coming from odbql.