schemas/util.sql: diff_cols(): embed left_query/right_query directly in the main SQL statement, so that the query planner can take the sort order of the queries into account in planning the FULL JOIN (and where possible, use a fast merge join)

schemas/vegbien.sql: regenerated from DB, which adds an autogenerated column alias

schemas/vegbien.sql, validation/aggregating/*/*.sql: renamed public_validation_* to plural public_validations_* since we refer to them as aggregating validation*s*

schemas/util.sql: added eval2col_pair()

schemas/vegbien.sql, validation/aggregating/*/*.sql: renamed public_validation_* to plural public_validations_* since we refer to them as aggregating validation*s*

schemas/vegbien.sql: public_validation_plots: added Brad's notes as comments

schemas/vegbien.sql: public_validation_plots: populated queries from validation/aggregating/plots/bien3_validations_salvias_vegbien.sql

schemas/vegbien.sql: public_validation_traits: added Brad's notes as comments

schemas/vegbien.sql: public_validation_traits: populated queries from validation/aggregating/traits/bien3_validations_traits_bien3.sql

schemas/vegbien.sql: analytical_stem.datasource: removed fkey to source.shortname because this prevents individually-reloaded datasources from being published

schemas/vegbien.sql: publish(), rm(): documented runtime (1 min)

schemas/vegbien.sql: added datasource_unpublish()

schemas/vegbien.sql: public_validation: split into separate subschemas for plots, specimens, traits

schemas/Makefile: vegbien.sql: support auxiliary schemas besides *_validation

fix: schemas/Makefile: %/publish: removed no longer applicable instructions about running `make schemas/rotate`. you should not publish a schema until you are satisfied that it can replace the previous public schema, so that the previous public schema doesn't need to be saved.

fix: schemas/vegbien.sql: public_validation comment: undid incorrect replacement of "validation" with "public_validation"

bugfix: schemas/Makefile: "public": dont

schemas/Makefile: %/publish: don't ignore errors if schema doesn't exist, because it should always exist if you are publishing it

schemas/Makefile: `%/uninstall: vegbien.sql`: use util.schema_bundle_rm() instead of public.rm(), because when the public schema is incompletely imported, public.rm() will not yet exist

schemas/Makefile: `%/install: vegbien.sql`: support auxiliary schemas besides *_validation

schemas/Makefile: %/uninstall, %/publish: use new public.rm(), publish(), which encapsulate the process of renaming multiple related schemas and also allow additional auxiliary schemas beyond just *_validation

schemas/vegbien.sql: added rm(), which deletes this schema

schemas/vegbien.sql: added publish(), which publishes this schema

schemas/vegbien.sql: added public_schema_publish()

schemas/util.sql: added functions for maintaining schema bundles, groups of schemas with a common prefix

schemas/util.sql: added schema_bundle_get_schemas()

schemas/util.sql: added schema_replace()

schemas/util.sql: added schema_rm(), schema_rename()

schemas/vegbien.sql: datasource_publish(): removed nested transaction, which is not needed because Postgres runs each top-level command in a transaction (including in autocommit mode)

bugfix: schemas/Makefile,vegbien.sql: renamed validation to public_validation to match the suffixed name applied by `make schemas/public/install`

schemas/Makefile: %/install, %/publish for public schema: added usage

removed schemas/VegX/run.py, which is a duplicate of _archive/bienDjango/trunk/bien/v/models.py (this may have ended up here because it was a wiki page attachment, but the initial revision did not indicate the source)

schemas/VegCore/vegbien_identifier_examples.xlsx: changed to symlink to Brad_Boyle/vegbien_identifier_examples.xlsx, to avoid duplicate copies

schemas/Makefile: removed no longer used rename/%, rotate. use %/install, %/publish instead to manage public schema versions.

schemas/Makefile: removed no longer needed public/uninstall, which is handled more generally by `%/uninstall: vegbien.sql` instead

schemas/Makefile: `%/uninstall: vegbien.sql`: added support for *_validation schema

schemas/Makefile: %/uninstall: split into 2 rules like for %/install, so that different actions can easily be added for the public schema

schemas/Makefile: %/publish: support auxiliary *_validation schemas

bugfix: schemas/Makefile: `%/install: vegbien.sql`: need -e before each s/// expression

schemas/Makefile: `%/install: vegbien.sql`: create a custom _validation schema for *each public schema, rather than one for the most recently-created public schema. this allows validations to continue to be run against a previous version of the DB while a new version is being imported.

schemas/vegbien.ERD.mwb: regenerated exports

schemas/vegbien.ERD.mwb: synced to schemas/vegbien.my.sql, which adds the validation schema

schemas/util.sql: added diff_views()

schemas/util.sql: added diff_cols(), for use with the aggregating validations

bugfix: schemas/Makefile: public/uninstall: need to uninstall the validation schema in addition to the public schema (deleting the public schema would only cascade to the contents of the validation schema)

schemas/vegbien.sql: added validation schema, which contains queries used in the aggregating validations. schemas/Makefile: include schemas that depend on `public` (eg. `validation`) so they are restored along with it.

fix: schemas/Makefile: %/uninstall: don't prompt to delete auxiliary schemas, as these do not contain data

schemas/vegbien.sql: added top_plot view

schemas/vegbien.sql: location_set_top_plot(): only do the lookup of the parent's top_plot when there is a parent_id

schemas/vegbien.sql: location: added autopopulated top_plot

schemas/vegbien.sql: locationevent: added locationevent_pull_forward_from_parent() trigger which populates project_id from the parent locationevent

schemas/vegbien.sql: locationevent__pull_forward_from_parent(): renamed to locationevent_pull_forward_from_parent_for_stratum() since it actually has a filter that causes it to apply only to stratum events

bugfix: schemas/vegbien.sql: commconcept: added missing unique constraint on commdescription, which is also an identifying field in addition to the ID (which is in commname.commname)

bugfix: schemas/vegbien.sql: locationevent__communities(): inner-join to commdetermination instead of left-joining, so that this does not produce community entries (which occur because there is one commclass per locationevent, but only some commclasses will then have a commdetermination)

schemas/vegbien.sql: projectcontributor: added missing unique constraint

moved everything into /trunk/ to create the standard svn layout, for use with tools that require this (eg. git-svn). IMPORTANT: do NOT do an `svn up`. instead, re-use your working copy's existing files with `svn switch` (http://svnbook.red-bean.com/en/1.6/svn.ref.svn.c.switch.html).

schemas/vegbien.sql: plot.**, analytical_stem_view, analytical_plot: added project_contributors

schemas/vegbien.sql: added project_contributors()

schemas/vegbien.sql: analytical_stem_view: scientificName_verbatim: don't use taxonverbatim.taxonname+author as the scientificName_verbatim if only the author is provided. (this lead to weird scientificName_verbatims that contain just the author.)

bugfix: schemas/vegbien.sql: location: use the place_id from the parent location when no place_id is specified. this fixes a bug in analytical_stem_view where the parent location's place_id was used because it was sometimes missing from the sublocation, but the parent place_id itself was sometimes missing instead if sublocations each had their own place information. this way, it is always available directly in the sublocation, populated from the parent location if needed.

bugfix: schemas/vegbien.sql: location: added place_id which is autopopulated from the current locationplace. join on this in plot.**, to avoid a 1:many join when a location has multiple locationplaces.

bugfix: schemas/vegbien.sql: locationevent_unique_within_parent_by_location unique index: need COALESCE around location_id since it's nullable

fix: schemas/vegbien.sql: analytical_specimen: added specimens-related columns that are in analytical_plot

bugfix: schemas/vegbien.sql: updated for renamed county_centroids column names

schemas/vegbien.sql: include the family_higher_plant_group lookup table values so that these don't need to be regenerated from the NCBI nodes whenever the DB is reloaded

schemas/vegbien.sql: taxonlabel_update_ancestors(): don't do an index scan if the value being scanned for is NULL, to support testing this function without the indexes in place, without extra full-table scans for NULL values affecting things. this can be used to determine if the function is actually using the indexes, by turning them off and seeing if the runtime changes.

schemas/util.sql: explain2table(): documented usage:
PERFORM util.explain2table($$
query
$$);

schemas/util.sql: explain2table(): by default, use the util.explain table

schemas/util.sql: added explain table

schemas/util.sql: added explain2notice()

schemas/util.sql: added explain2str()

schemas/util.sql: added explain2table()

schemas/util.sql: added explain()

schemas/vegbien.sql: taxonlabel_update_ancestors(): don't create a performance-intensive nested transaction (EXCEPTION block) for each INSERT, because there should no longer be duplicate ancestors, so it's OK to abort the whole transaction if this assertion fails

bugfix: schemas/vegbien.sql: taxonlabel_update_ancestors_on_{insert,update}(): only use either the matched taxon's ancestors or the parent's ancestors, to avoid issues related to duplication between these two ancestors lists. this also fixes a bug where the 2nd taxonlabel_update_ancestors() call assumes that the existing ancestors are for the old parent, when in fact they have actually just been set to those for the new matched taxon (which horribly confuses taxonlabel_update_ancestors()).

schemas/vegbien.sql: _taxonlabel_set_parent_id(): just use a plain UPDATE statement, to avoid the significant parsing and stringification overhead of EXECUTE and quote_nullable(). it is not clear that EXECUTE is actually necessary to avoid caching the query plan, because the cache should be invalidated automatically when the table's ANALYZE statistics are regenerated.

schemas/vegbien.sql: removed unused function _taxonlabel_set_matched_label_id(), which refers to obsolete fields

schemas/vegbien.sql: synced to DB (the view renderer apparently changed the text of a view)

bugfix: schemas/Makefile: %/uninstall: always confirm before removing an existing schema, not just for public and r*, because an auxiliary schema might also be used as $version and reinstalled by bin/import_all

schemas/vegbien.sql: analytical_stem_view: scrubbed_author: removed empty COALESCE around value (left over from when multiple values needed to be combined for many TNRS fields)

schemas/vegbien.sql: provider_count_view: source totals: use the much faster query developed for Brad (wiki.vegpath.org/VegBIEN_FAQ#from-Brad-on-2013-12-4), which avoids the need to do a GROUP BY on all of analytical_stem. eventually, we will want to apply the same optimization to the first publisher subtotals.

schemas/vegbien.ERD.mwb: regenerated exports

schemas/postgresql.conf: work_mem: documented that this seemingly small # is multiplied by max_connections, i.e. 256 MB * 100 = 26 GB, which approaches total memory (32 GB)

bugfix: schemas/Makefile: $(confirmRmPublicSchema): only prompt to delete the schema if it actually exists. this avoids prompting to remove a non-existent schema at the beginning of bin/import_all, which requires user attention. since bin/import_all is often run with a delayed start (e.g. to wait for a staging table reinstall to complete), the user may not be at the terminal when this message is displayed, and without this fix, the import would be prevented from running until they return.

schemas/vegbien.sql: collected_dates: documented runtime (2.5 min)

schemas/vegbien.sql: collected_date_min: replaced with collected_dates view that lists all dates we have, so that we can determine which of these may be valid. it turns out that we have data collected from very far back (to the year 1), which are not merely 2-digit years because PostgreSQL will only parse early years when there are 4 digits.

bugfix: schemas/vegbien.sql: collected_date_min: exclude invalid dates < 1000-01-01

bugfix: schemas/vegbien.sql: collected_date_min: exclude -infinity

schemas/vegbien.sql: added collected_date_min view

schemas/vegbien.sql: analytical_stem_view: added taxonomic_status. notice that PostgreSQL 9.3 puts each view column on a separate line, making it much easier to review the svn diff!

fix: schemas/vegbien.sql: taxon_trait_view: include only TNRS-valid names

schemas/vegbien.sql: sync_analytical_stem_to_view(): use new util.force_recreate() instead of manually dropping and re-creating every view that uses this. this avoids the need to add several lines to this function every time we add a new scientific view (of which we expect to have many), because force_recreate()'s error parsing handles this automatically. this makes it possible for a non-expert user to add scientific views without compromising the ability to add columns to analytical_stem_view, because they don't need to understand Postgres's dependency error messages when updating analytical_stem with this function.

schemas/util.sql: added force_recreate(), for use by sync_analytical_stem_to_view(). this uses the new `GET STACKED DIAGNOSTICS` in PostgreSQL 9.3 to access the DETAIL section of the dependent_objects_still_exist error.

schemas/*.conf: upgraded to PostgreSQL 9.3, which is needed for proper exception parsing in the auto-re-create-views functionality

bugfix: schemas/vegbien.sql: plantobservation_aggregateoccurrence_count_1(): only default aggregateoccurrence.count to 1 for specimens data, because plots data may have any number of individuals in a taxon_presence record that has no explicit individual_count

schemas/*.sql: updated for PostgreSQL 9.3. this reorders some functions, adds empty comment headers for omitted SEQUENCE SET commands, and (best of all) finally splits view columns onto multiple lines, so that changes in the columns are actually legible (and produce their own svn diff!)

Makefile, schemas/.Mac.conf: upgraded to PostgreSQL 9.3, which is needed for proper exception parsing in the auto-re-create-views functionality. this also removes the Mac 10.8 Mountain Lion quirks, such as renaming the postgres user to _postgres (which messed everything up, but is now back to normal).

schemas/util.sql: added save_drop_views()

schemas/util.sql: added is_empty(anyarray)

schemas/util.sql: added regexp_matches_group()