2017-01-26 00:13:56 +00:00
// Copyright 2016 Mozilla
//
// Licensed under the Apache License, Version 2.0 (the "License"); you may not use
// this file except in compliance with the License. You may obtain a copy of the
// License at http://www.apache.org/licenses/LICENSE-2.0
// Unless required by applicable law or agreed to in writing, software distributed
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
// specific language governing permissions and limitations under the License.
#![ allow(dead_code) ]
2018-07-16 20:58:34 +00:00
#![ allow(unused_macros) ]
2017-01-26 00:13:56 +00:00
/// Low-level functions for testing.
2018-07-16 20:58:34 +00:00
// Macro to parse a `Borrow<str>` to an `edn::Value` and assert the given `edn::Value` `matches`
// against it.
//
// This is a macro only to give nice line numbers when tests fail.
#[ macro_export ]
macro_rules ! assert_matches {
( $input : expr , $expected : expr ) = > { {
// Failure to parse the expected pattern is a coding error, so we unwrap.
let pattern_value = edn ::parse ::value ( $expected . borrow ( ) )
. expect ( format! ( " to be able to parse expected {} " , $expected ) . as_str ( ) )
. without_spans ( ) ;
let input_value = $input . to_edn ( ) ;
2020-01-14 15:46:21 +00:00
assert! (
input_value . matches ( & pattern_value ) ,
" Expected value: \n {} \n to match pattern: \n {} \n " ,
input_value . to_pretty ( 120 ) . unwrap ( ) ,
pattern_value . to_pretty ( 120 ) . unwrap ( )
) ;
} } ;
2018-07-16 20:58:34 +00:00
}
// Transact $input against the given $conn, expecting success or a `Result<TxReport, String>`.
//
// This unwraps safely and makes asserting errors pleasant.
#[ macro_export ]
macro_rules ! assert_transact {
( $conn : expr , $input : expr , $expected : expr ) = > { {
trace! ( " assert_transact: {} " , $input ) ;
let result = $conn . transact ( $input ) . map_err ( | e | e . to_string ( ) ) ;
assert_eq! ( result , $expected . map_err ( | e | e . to_string ( ) ) ) ;
} } ;
( $conn : expr , $input : expr ) = > { {
trace! ( " assert_transact: {} " , $input ) ;
let result = $conn . transact ( $input ) ;
2020-01-14 15:46:21 +00:00
assert! (
result . is_ok ( ) ,
" Expected Ok(_), got `{}` " ,
result . unwrap_err ( )
) ;
2018-07-16 20:58:34 +00:00
result . unwrap ( )
} } ;
}
Extract partial storage abstraction; use error-chain throughout. Fixes #328. r=rnewman (#341)
* Pre: Drop unneeded tx0 from search results.
* Pre: Don't require a schema in some of the DB code.
The idea is to separate the transaction applying code, which is
schema-aware, from the concrete storage code, which is just concerned
with getting bits onto disk.
* Pre: Only reference Schema, not DB, in debug module.
This is part of a larger separation of the volatile PartitionMap,
which is modified every transaction, from the stable Schema, which is
infrequently modified.
* Pre: Fix indentation.
* Extract part of DB to new SchemaTypeChecking trait.
* Extract part of DB to new PartitionMapping trait.
* Pre: Don't expect :db.part/tx partition to advance when tx fails.
This fails right now, because we allocate tx IDs even when we shouldn't.
* Sketch a db interface without DB.
* Add ValueParseError; use error-chain in tx-parser.
This can be simplified when
https://github.com/Marwes/combine/issues/86 makes it to a published
release, but this unblocks us for now. This converts the `combine`
error type `ParseError<&'a [edn::Value]>` to a type with owned
`Vec<edn::Value>` collections, re-using `edn::Value::Vector` for
making them `Display`.
* Pre: Accept Borrow<Schema> instead of just &Schema in debug module.
This makes it easy to use Rc<Schema> or Arc<Schema> without inserting
&* sigils throughout the code.
* Use error-chain in query-parser.
There are a few things to point out here:
- the fine grained error types have been flattened into one crate-wide
error type; it's pretty easy to regain the granularity as needed.
- edn::ParseError is automatically lifted to
mentat_query_parser::errors::Error;
- we use mentat_parser_utils::ValueParser to maintain parsing error
information from `combine`.
* Patch up top-level.
* Review comment: Only `borrow()` once.
2017-02-24 23:32:41 +00:00
use std ::borrow ::Borrow ;
2018-07-16 20:58:34 +00:00
use std ::collections ::BTreeMap ;
2020-01-14 15:46:21 +00:00
use std ::io ::Write ;
2017-02-08 22:04:32 +00:00
use itertools ::Itertools ;
2017-01-26 00:13:56 +00:00
use rusqlite ;
2020-01-14 15:46:21 +00:00
use rusqlite ::types ::ToSql ;
use rusqlite ::TransactionBehavior ;
2017-02-08 22:04:32 +00:00
use tabwriter ::TabWriter ;
2017-01-26 00:13:56 +00:00
2020-08-06 03:03:58 +00:00
use crate ::bootstrap ;
use crate ::db ::* ;
use crate ::db ::{ read_attribute_map , read_ident_map } ;
use crate ::entids ;
2020-01-14 15:46:21 +00:00
use db_traits ::errors ::Result ;
2017-02-08 22:04:32 +00:00
use edn ;
2018-08-08 17:35:06 +00:00
2020-01-14 15:46:21 +00:00
use core_traits ::{ Entid , TypedValue , ValueType } ;
2020-08-06 03:03:58 +00:00
use crate ::internal_types ::TermWithTempIds ;
use crate ::schema ::SchemaBuilding ;
use crate ::tx ::{ transact , transact_terms } ;
use crate ::types ::* ;
use crate ::watcher ::NullWatcher ;
2020-01-14 15:46:21 +00:00
use edn ::entities ::{ EntidOrIdent , TempId } ;
use edn ::InternSet ;
use mentat_core ::{ HasSchema , SQLValueType , TxReport } ;
2017-01-26 00:13:56 +00:00
2017-02-08 22:04:32 +00:00
/// Represents a *datom* (assertion) in the store.
2020-01-14 15:46:21 +00:00
#[ derive(Clone, Debug, Eq, Hash, Ord, PartialOrd, PartialEq) ]
2018-07-16 20:58:34 +00:00
pub struct Datom {
2017-01-26 00:13:56 +00:00
// TODO: generalize this.
2018-07-16 20:58:34 +00:00
pub e : EntidOrIdent ,
pub a : EntidOrIdent ,
pub v : edn ::Value ,
pub tx : i64 ,
pub added : Option < bool > ,
2017-02-08 22:04:32 +00:00
}
/// Represents a set of datoms (assertions) in the store.
2017-03-20 18:29:17 +00:00
///
/// To make comparision easier, we deterministically order. The ordering is the ascending tuple
/// ordering determined by `(e, a, (value_type_tag, v), tx)`, where `value_type_tag` is an internal
/// value that is not exposed but is deterministic.
2018-07-16 20:58:34 +00:00
pub struct Datoms ( pub Vec < Datom > ) ;
2017-02-08 22:04:32 +00:00
/// Represents an ordered sequence of transactions in the store.
2017-03-20 18:29:17 +00:00
///
/// To make comparision easier, we deterministically order. The ordering is the ascending tuple
/// ordering determined by `(e, a, (value_type_tag, v), tx, added)`, where `value_type_tag` is an
/// internal value that is not exposed but is deterministic, and `added` is ordered such that
/// retracted assertions appear before added assertions.
2018-07-16 20:58:34 +00:00
pub struct Transactions ( pub Vec < Datoms > ) ;
2017-02-08 22:04:32 +00:00
2017-03-21 20:12:10 +00:00
/// Represents the fulltext values in the store.
2018-07-16 20:58:34 +00:00
pub struct FulltextValues ( pub Vec < ( i64 , String ) > ) ;
2017-03-21 20:12:10 +00:00
2017-02-08 22:04:32 +00:00
impl Datom {
2018-07-16 20:58:34 +00:00
pub fn to_edn ( & self ) -> edn ::Value {
2018-06-26 23:34:18 +00:00
let f = | entid : & EntidOrIdent | -> edn ::Value {
2017-02-08 22:04:32 +00:00
match * entid {
2020-01-23 18:16:19 +00:00
EntidOrIdent ::Entid ( ref y ) = > edn ::Value ::Integer ( * y ) ,
2018-06-26 23:34:18 +00:00
EntidOrIdent ::Ident ( ref y ) = > edn ::Value ::Keyword ( y . clone ( ) ) ,
2017-02-08 22:04:32 +00:00
}
} ;
2017-03-20 18:29:17 +00:00
let mut v = vec! [ f ( & self . e ) , f ( & self . a ) , self . v . clone ( ) ] ;
2017-02-08 22:04:32 +00:00
if let Some ( added ) = self . added {
2017-03-20 18:29:17 +00:00
v . push ( edn ::Value ::Integer ( self . tx ) ) ;
2017-02-08 22:04:32 +00:00
v . push ( edn ::Value ::Boolean ( added ) ) ;
}
edn ::Value ::Vector ( v )
}
2017-01-26 00:13:56 +00:00
}
2017-02-08 22:04:32 +00:00
impl Datoms {
2018-07-16 20:58:34 +00:00
pub fn to_edn ( & self ) -> edn ::Value {
2020-01-23 18:16:19 +00:00
edn ::Value ::Vector ( ( & self . 0 ) . iter ( ) . map ( | x | x . to_edn ( ) ) . collect ( ) )
2017-02-08 22:04:32 +00:00
}
}
impl Transactions {
2018-07-16 20:58:34 +00:00
pub fn to_edn ( & self ) -> edn ::Value {
2020-01-23 18:16:19 +00:00
edn ::Value ::Vector ( ( & self . 0 ) . iter ( ) . map ( | x | x . to_edn ( ) ) . collect ( ) )
2017-02-08 22:04:32 +00:00
}
}
2017-03-21 20:12:10 +00:00
impl FulltextValues {
2018-07-16 20:58:34 +00:00
pub fn to_edn ( & self ) -> edn ::Value {
2020-01-14 15:46:21 +00:00
edn ::Value ::Vector (
( & self . 0 )
2020-01-23 18:16:19 +00:00
. iter ( )
2020-01-14 15:46:21 +00:00
. map ( | & ( x , ref y ) | {
edn ::Value ::Vector ( vec! [ edn ::Value ::Integer ( x ) , edn ::Value ::Text ( y . clone ( ) ) ] )
} )
. collect ( ) ,
)
2017-03-21 20:12:10 +00:00
}
}
Schema alteration. Fixes #294 and #295. (#370) r=rnewman
* Pre: Don't retract :db/ident in test.
Datomic (and eventually Mentat) don't allow to retract :db/ident in
this way, so this runs afoul of future work to support mutating
metadata.
* Pre: s/VALUETYPE/VALUE_TYPE/.
This is consistent with the capitalization (which is "valueType") and
the other identifier.
* Pre: Remove some single quotes from error output.
* Part 1: Make materialized views be uniform [e a v value_type_tag].
This looks ahead to a time when we could support arbitrary
user-defined materialized views. For now, the "idents" materialized
view is those datoms of the form [e :db/ident :namespaced/keyword] and
the "schema" materialized view is those datoms of the form [e a v]
where a is in a particular set of attributes that will become clear in
the following commits.
This change is not backwards compatible, so I'm removing the open
current (really, v2) test. It'll be re-instated when we get to
https://github.com/mozilla/mentat/issues/194.
* Pre: Map TypedValue::Ref to TypedValue::Keyword in debug output.
* Part 3: Separate `schema_to_mutate` from the `schema` used to interpret.
This is just to keep track of the expected changes during
bootstrapping. I want bootstrap metadata mutations to flow through
the same code path as metadata mutations during regular transactions;
by differentiating the schema used for interpretation from the schema
that will be updated I expect to be able to apply bootstrap metadata
mutations to an empty schema and have things like materialized views
created (using the regular code paths).
This commit has been re-ordered for conceptual clarity, but it won't
compile because it references the metadata module. It's possible to
make it compile -- the functionality is there in the schema module --
but it's not worth the rebasing effort until after review (and
possibly not even then, since we'll squash down to a single commit to
land).
* Part 2: Maintain entids separately from idents.
In order to support historical idents, we need to distinguish the
"current" map from entid -> ident from the "complete historical" map
ident -> entid. This is what Datomic does; in Datomic, an ident is
never retracted (although it can be replaced). This approach is an
important part of allowing multiple consumers to share a schema
fragment as it migrates forward.
This fixes a limitation of the Clojure implementation, which did not
handle historical idents across knowledge base close and re-open.
The "entids" materialized view is naturally a slice of the "datoms"
table. The "idents" materialized view is a slice of the
"transactions" table. I hope that representing in this way, and
casting the problem in this light, might generalize to future
materialized views.
* Pre: Add DiffSet.
* Part 4: Collect mutations to a `Schema`.
I haven't taken your review comment about consuming AttributeBuilder
during each fluent function. If you read my response and still want
this, I'm happy to do it in review.
* Part 5: Handle :db/ident and :db.{install,alter}/attribute.
This "loops" the committed datoms out of the SQL store and back
through the metadata (schema, but in future also partition map)
processor. The metadata processor updates the schema and produces a
report of what changed; that report is then used to update the SQL
store. That update includes:
- the materialized views ("entids", "idents", and "schema");
- if needed, a subset of the datoms themselves (as flags change).
I've left a TODO for handling attribute retraction in the cases that
it makes sense. I expect that to be straight-forward.
* Review comment: Rename DiffSet to AddRetractAlterSet.
Also adds a little more commentary and a simple test.
* Review comment: Use ToIdent trait.
* Review comment: partially revert "Part 2: Maintain entids separately from idents."
This reverts commit 23a91df9c35e14398f2ddbd1ba25315821e67401.
Following our discussion, this removes the "entids" materialized
view. The next commit will remove historical idents from the "idents"
materialized view.
* Post: Use custom Either rather than std::result::Result.
This is not necessary, but it was suggested that we might be paying an
overhead creating Err instances while using error_chain. That seems
not to be the case, but this change shows that we don't actually use
any of the Result helper methods, so there's no reason to overload
Result. This change might avoid some future confusion, so I'm going
to land it anyway.
Signed-off-by: Nick Alexander <nalexander@mozilla.com>
* Review comment: Don't preserve historical idents.
* Review comment: More prepared statements when updating materialized views.
* Post: Test altering :db/cardinality and :db/unique.
These tests fail due to a Datomic limitation, namely that the marker
flag :db.alter/attribute can only be asserted once for an attribute!
That is, [:db.part/db :db.alter/attribute :attribute] will only be
transacted at most once. Since older versions of Datomic required the
:db.alter/attribute flag, I can only imagine they either never wrote
:db.alter/attribute to the store, or they handled it specially. I'll
need to remove the marker flag system from Mentat in order to address
this fundamental limitation.
* Post: Remove some more single quotes from error output.
* Post: Add assert_transact! macro to unwrap safely.
I was finding it very difficult to track unwrapping errors while
making changes, due to an underlying Mac OS X symbolication issue that
makes running tests with RUST_BACKTRACE=1 so slow that they all time
out.
* Post: Don't expect or recognize :db.{install,alter}/attribute.
I had this all working... except we will never see a repeated
`[:db.part/db :db.alter/attribute :attribute]` assertion in the store!
That means my approach would let you alter an attribute at most one
time. It's not worth hacking around this; it's better to just stop
expecting (and recognizing) the marker flags. (We have all the data
to distinguish the various cases that we need without the marker
flags.)
This brings Mentat in line with the thrust of newer Datomic versions,
but isn't compatible with Datomic, because (if I understand correctly)
Datomic automatically adds :db.{install,alter}/attribute assertions to
transactions.
I haven't purged the corresponding :db/ident and schema fragments just
yet:
- we might want them back
- we might want them in order to upgrade v1 and v2 databases to the
new on-disk layout we're fleshing out (v3?).
* Post: Don't make :db/unique :db.unique/* imply :db/index true.
This patch avoids a potential bug with the "schema" materialized view.
If :db/unique :db.unique/value implies :db/index true, then what
happens when you _retract_ :db.unique/value? I think Datomic defines
this in some way, but I really want the "schema" materialized view to
be a slice of "datoms" and not have these sort of ambiguities and
persistent effects. Therefore, to ensure that we don't retract a
schema characteristic and accidentally change more than we intended
to, this patch stops having any schema characteristic imply any other
schema characteristic(s). To achieve that, I added an
Option<Unique::{Value,Identity}> type to Attribute; this helps with
this patch, and also looks ahead to when we allow to retract
:db/unique attributes.
* Post: Allow to retract :db/ident.
* Post: Include more details about invalid schema changes.
The tests use strings, so they hide the chained errors which do in
fact provide more detail.
* Review comment: Fix outdated comment.
* Review comment: s/_SET/_SQL_LIST/.
* Review comment: Use a sub-select for checking cardinality.
This might be faster in practice.
* Review comment: Put `attribute::Unique` into its own namespace.
2017-03-20 20:18:59 +00:00
/// Turn TypedValue::Ref into TypedValue::Keyword when it is possible.
trait ToIdent {
2020-01-14 15:46:21 +00:00
fn map_ident ( self , schema : & Schema ) -> Self ;
Schema alteration. Fixes #294 and #295. (#370) r=rnewman
* Pre: Don't retract :db/ident in test.
Datomic (and eventually Mentat) don't allow to retract :db/ident in
this way, so this runs afoul of future work to support mutating
metadata.
* Pre: s/VALUETYPE/VALUE_TYPE/.
This is consistent with the capitalization (which is "valueType") and
the other identifier.
* Pre: Remove some single quotes from error output.
* Part 1: Make materialized views be uniform [e a v value_type_tag].
This looks ahead to a time when we could support arbitrary
user-defined materialized views. For now, the "idents" materialized
view is those datoms of the form [e :db/ident :namespaced/keyword] and
the "schema" materialized view is those datoms of the form [e a v]
where a is in a particular set of attributes that will become clear in
the following commits.
This change is not backwards compatible, so I'm removing the open
current (really, v2) test. It'll be re-instated when we get to
https://github.com/mozilla/mentat/issues/194.
* Pre: Map TypedValue::Ref to TypedValue::Keyword in debug output.
* Part 3: Separate `schema_to_mutate` from the `schema` used to interpret.
This is just to keep track of the expected changes during
bootstrapping. I want bootstrap metadata mutations to flow through
the same code path as metadata mutations during regular transactions;
by differentiating the schema used for interpretation from the schema
that will be updated I expect to be able to apply bootstrap metadata
mutations to an empty schema and have things like materialized views
created (using the regular code paths).
This commit has been re-ordered for conceptual clarity, but it won't
compile because it references the metadata module. It's possible to
make it compile -- the functionality is there in the schema module --
but it's not worth the rebasing effort until after review (and
possibly not even then, since we'll squash down to a single commit to
land).
* Part 2: Maintain entids separately from idents.
In order to support historical idents, we need to distinguish the
"current" map from entid -> ident from the "complete historical" map
ident -> entid. This is what Datomic does; in Datomic, an ident is
never retracted (although it can be replaced). This approach is an
important part of allowing multiple consumers to share a schema
fragment as it migrates forward.
This fixes a limitation of the Clojure implementation, which did not
handle historical idents across knowledge base close and re-open.
The "entids" materialized view is naturally a slice of the "datoms"
table. The "idents" materialized view is a slice of the
"transactions" table. I hope that representing in this way, and
casting the problem in this light, might generalize to future
materialized views.
* Pre: Add DiffSet.
* Part 4: Collect mutations to a `Schema`.
I haven't taken your review comment about consuming AttributeBuilder
during each fluent function. If you read my response and still want
this, I'm happy to do it in review.
* Part 5: Handle :db/ident and :db.{install,alter}/attribute.
This "loops" the committed datoms out of the SQL store and back
through the metadata (schema, but in future also partition map)
processor. The metadata processor updates the schema and produces a
report of what changed; that report is then used to update the SQL
store. That update includes:
- the materialized views ("entids", "idents", and "schema");
- if needed, a subset of the datoms themselves (as flags change).
I've left a TODO for handling attribute retraction in the cases that
it makes sense. I expect that to be straight-forward.
* Review comment: Rename DiffSet to AddRetractAlterSet.
Also adds a little more commentary and a simple test.
* Review comment: Use ToIdent trait.
* Review comment: partially revert "Part 2: Maintain entids separately from idents."
This reverts commit 23a91df9c35e14398f2ddbd1ba25315821e67401.
Following our discussion, this removes the "entids" materialized
view. The next commit will remove historical idents from the "idents"
materialized view.
* Post: Use custom Either rather than std::result::Result.
This is not necessary, but it was suggested that we might be paying an
overhead creating Err instances while using error_chain. That seems
not to be the case, but this change shows that we don't actually use
any of the Result helper methods, so there's no reason to overload
Result. This change might avoid some future confusion, so I'm going
to land it anyway.
Signed-off-by: Nick Alexander <nalexander@mozilla.com>
* Review comment: Don't preserve historical idents.
* Review comment: More prepared statements when updating materialized views.
* Post: Test altering :db/cardinality and :db/unique.
These tests fail due to a Datomic limitation, namely that the marker
flag :db.alter/attribute can only be asserted once for an attribute!
That is, [:db.part/db :db.alter/attribute :attribute] will only be
transacted at most once. Since older versions of Datomic required the
:db.alter/attribute flag, I can only imagine they either never wrote
:db.alter/attribute to the store, or they handled it specially. I'll
need to remove the marker flag system from Mentat in order to address
this fundamental limitation.
* Post: Remove some more single quotes from error output.
* Post: Add assert_transact! macro to unwrap safely.
I was finding it very difficult to track unwrapping errors while
making changes, due to an underlying Mac OS X symbolication issue that
makes running tests with RUST_BACKTRACE=1 so slow that they all time
out.
* Post: Don't expect or recognize :db.{install,alter}/attribute.
I had this all working... except we will never see a repeated
`[:db.part/db :db.alter/attribute :attribute]` assertion in the store!
That means my approach would let you alter an attribute at most one
time. It's not worth hacking around this; it's better to just stop
expecting (and recognizing) the marker flags. (We have all the data
to distinguish the various cases that we need without the marker
flags.)
This brings Mentat in line with the thrust of newer Datomic versions,
but isn't compatible with Datomic, because (if I understand correctly)
Datomic automatically adds :db.{install,alter}/attribute assertions to
transactions.
I haven't purged the corresponding :db/ident and schema fragments just
yet:
- we might want them back
- we might want them in order to upgrade v1 and v2 databases to the
new on-disk layout we're fleshing out (v3?).
* Post: Don't make :db/unique :db.unique/* imply :db/index true.
This patch avoids a potential bug with the "schema" materialized view.
If :db/unique :db.unique/value implies :db/index true, then what
happens when you _retract_ :db.unique/value? I think Datomic defines
this in some way, but I really want the "schema" materialized view to
be a slice of "datoms" and not have these sort of ambiguities and
persistent effects. Therefore, to ensure that we don't retract a
schema characteristic and accidentally change more than we intended
to, this patch stops having any schema characteristic imply any other
schema characteristic(s). To achieve that, I added an
Option<Unique::{Value,Identity}> type to Attribute; this helps with
this patch, and also looks ahead to when we allow to retract
:db/unique attributes.
* Post: Allow to retract :db/ident.
* Post: Include more details about invalid schema changes.
The tests use strings, so they hide the chained errors which do in
fact provide more detail.
* Review comment: Fix outdated comment.
* Review comment: s/_SET/_SQL_LIST/.
* Review comment: Use a sub-select for checking cardinality.
This might be faster in practice.
* Review comment: Put `attribute::Unique` into its own namespace.
2017-03-20 20:18:59 +00:00
}
impl ToIdent for TypedValue {
fn map_ident ( self , schema : & Schema ) -> Self {
if let TypedValue ::Ref ( e ) = self {
2020-01-14 15:46:21 +00:00
schema
. get_ident ( e )
. cloned ( )
. map ( | i | i . into ( ) )
. unwrap_or ( TypedValue ::Ref ( e ) )
Schema alteration. Fixes #294 and #295. (#370) r=rnewman
* Pre: Don't retract :db/ident in test.
Datomic (and eventually Mentat) don't allow to retract :db/ident in
this way, so this runs afoul of future work to support mutating
metadata.
* Pre: s/VALUETYPE/VALUE_TYPE/.
This is consistent with the capitalization (which is "valueType") and
the other identifier.
* Pre: Remove some single quotes from error output.
* Part 1: Make materialized views be uniform [e a v value_type_tag].
This looks ahead to a time when we could support arbitrary
user-defined materialized views. For now, the "idents" materialized
view is those datoms of the form [e :db/ident :namespaced/keyword] and
the "schema" materialized view is those datoms of the form [e a v]
where a is in a particular set of attributes that will become clear in
the following commits.
This change is not backwards compatible, so I'm removing the open
current (really, v2) test. It'll be re-instated when we get to
https://github.com/mozilla/mentat/issues/194.
* Pre: Map TypedValue::Ref to TypedValue::Keyword in debug output.
* Part 3: Separate `schema_to_mutate` from the `schema` used to interpret.
This is just to keep track of the expected changes during
bootstrapping. I want bootstrap metadata mutations to flow through
the same code path as metadata mutations during regular transactions;
by differentiating the schema used for interpretation from the schema
that will be updated I expect to be able to apply bootstrap metadata
mutations to an empty schema and have things like materialized views
created (using the regular code paths).
This commit has been re-ordered for conceptual clarity, but it won't
compile because it references the metadata module. It's possible to
make it compile -- the functionality is there in the schema module --
but it's not worth the rebasing effort until after review (and
possibly not even then, since we'll squash down to a single commit to
land).
* Part 2: Maintain entids separately from idents.
In order to support historical idents, we need to distinguish the
"current" map from entid -> ident from the "complete historical" map
ident -> entid. This is what Datomic does; in Datomic, an ident is
never retracted (although it can be replaced). This approach is an
important part of allowing multiple consumers to share a schema
fragment as it migrates forward.
This fixes a limitation of the Clojure implementation, which did not
handle historical idents across knowledge base close and re-open.
The "entids" materialized view is naturally a slice of the "datoms"
table. The "idents" materialized view is a slice of the
"transactions" table. I hope that representing in this way, and
casting the problem in this light, might generalize to future
materialized views.
* Pre: Add DiffSet.
* Part 4: Collect mutations to a `Schema`.
I haven't taken your review comment about consuming AttributeBuilder
during each fluent function. If you read my response and still want
this, I'm happy to do it in review.
* Part 5: Handle :db/ident and :db.{install,alter}/attribute.
This "loops" the committed datoms out of the SQL store and back
through the metadata (schema, but in future also partition map)
processor. The metadata processor updates the schema and produces a
report of what changed; that report is then used to update the SQL
store. That update includes:
- the materialized views ("entids", "idents", and "schema");
- if needed, a subset of the datoms themselves (as flags change).
I've left a TODO for handling attribute retraction in the cases that
it makes sense. I expect that to be straight-forward.
* Review comment: Rename DiffSet to AddRetractAlterSet.
Also adds a little more commentary and a simple test.
* Review comment: Use ToIdent trait.
* Review comment: partially revert "Part 2: Maintain entids separately from idents."
This reverts commit 23a91df9c35e14398f2ddbd1ba25315821e67401.
Following our discussion, this removes the "entids" materialized
view. The next commit will remove historical idents from the "idents"
materialized view.
* Post: Use custom Either rather than std::result::Result.
This is not necessary, but it was suggested that we might be paying an
overhead creating Err instances while using error_chain. That seems
not to be the case, but this change shows that we don't actually use
any of the Result helper methods, so there's no reason to overload
Result. This change might avoid some future confusion, so I'm going
to land it anyway.
Signed-off-by: Nick Alexander <nalexander@mozilla.com>
* Review comment: Don't preserve historical idents.
* Review comment: More prepared statements when updating materialized views.
* Post: Test altering :db/cardinality and :db/unique.
These tests fail due to a Datomic limitation, namely that the marker
flag :db.alter/attribute can only be asserted once for an attribute!
That is, [:db.part/db :db.alter/attribute :attribute] will only be
transacted at most once. Since older versions of Datomic required the
:db.alter/attribute flag, I can only imagine they either never wrote
:db.alter/attribute to the store, or they handled it specially. I'll
need to remove the marker flag system from Mentat in order to address
this fundamental limitation.
* Post: Remove some more single quotes from error output.
* Post: Add assert_transact! macro to unwrap safely.
I was finding it very difficult to track unwrapping errors while
making changes, due to an underlying Mac OS X symbolication issue that
makes running tests with RUST_BACKTRACE=1 so slow that they all time
out.
* Post: Don't expect or recognize :db.{install,alter}/attribute.
I had this all working... except we will never see a repeated
`[:db.part/db :db.alter/attribute :attribute]` assertion in the store!
That means my approach would let you alter an attribute at most one
time. It's not worth hacking around this; it's better to just stop
expecting (and recognizing) the marker flags. (We have all the data
to distinguish the various cases that we need without the marker
flags.)
This brings Mentat in line with the thrust of newer Datomic versions,
but isn't compatible with Datomic, because (if I understand correctly)
Datomic automatically adds :db.{install,alter}/attribute assertions to
transactions.
I haven't purged the corresponding :db/ident and schema fragments just
yet:
- we might want them back
- we might want them in order to upgrade v1 and v2 databases to the
new on-disk layout we're fleshing out (v3?).
* Post: Don't make :db/unique :db.unique/* imply :db/index true.
This patch avoids a potential bug with the "schema" materialized view.
If :db/unique :db.unique/value implies :db/index true, then what
happens when you _retract_ :db.unique/value? I think Datomic defines
this in some way, but I really want the "schema" materialized view to
be a slice of "datoms" and not have these sort of ambiguities and
persistent effects. Therefore, to ensure that we don't retract a
schema characteristic and accidentally change more than we intended
to, this patch stops having any schema characteristic imply any other
schema characteristic(s). To achieve that, I added an
Option<Unique::{Value,Identity}> type to Attribute; this helps with
this patch, and also looks ahead to when we allow to retract
:db/unique attributes.
* Post: Allow to retract :db/ident.
* Post: Include more details about invalid schema changes.
The tests use strings, so they hide the chained errors which do in
fact provide more detail.
* Review comment: Fix outdated comment.
* Review comment: s/_SET/_SQL_LIST/.
* Review comment: Use a sub-select for checking cardinality.
This might be faster in practice.
* Review comment: Put `attribute::Unique` into its own namespace.
2017-03-20 20:18:59 +00:00
} else {
self
}
}
}
2017-02-08 22:04:32 +00:00
/// Convert a numeric entid to an ident `Entid` if possible, otherwise a numeric `Entid`.
2018-07-16 20:58:34 +00:00
pub fn to_entid ( schema : & Schema , entid : i64 ) -> EntidOrIdent {
2020-01-14 15:46:21 +00:00
schema
. get_ident ( entid )
. map_or ( EntidOrIdent ::Entid ( entid ) , | ident | {
EntidOrIdent ::Ident ( ident . clone ( ) )
} )
2017-01-26 00:13:56 +00:00
}
2018-07-16 20:58:34 +00:00
// /// Convert a symbolic ident to an ident `Entid` if possible, otherwise a numeric `Entid`.
// pub fn to_ident(schema: &Schema, entid: i64) -> Entid {
// schema.get_ident(entid).map_or(Entid::Entid(entid), |ident| Entid::Ident(ident.clone()))
// }
2017-02-08 22:04:32 +00:00
/// Return the set of datoms in the store, ordered by (e, a, v, tx), but not including any datoms of
/// the form [... :db/txInstant ...].
2018-07-16 20:58:34 +00:00
pub fn datoms < S : Borrow < Schema > > ( conn : & rusqlite ::Connection , schema : & S ) -> Result < Datoms > {
Extract partial storage abstraction; use error-chain throughout. Fixes #328. r=rnewman (#341)
* Pre: Drop unneeded tx0 from search results.
* Pre: Don't require a schema in some of the DB code.
The idea is to separate the transaction applying code, which is
schema-aware, from the concrete storage code, which is just concerned
with getting bits onto disk.
* Pre: Only reference Schema, not DB, in debug module.
This is part of a larger separation of the volatile PartitionMap,
which is modified every transaction, from the stable Schema, which is
infrequently modified.
* Pre: Fix indentation.
* Extract part of DB to new SchemaTypeChecking trait.
* Extract part of DB to new PartitionMapping trait.
* Pre: Don't expect :db.part/tx partition to advance when tx fails.
This fails right now, because we allocate tx IDs even when we shouldn't.
* Sketch a db interface without DB.
* Add ValueParseError; use error-chain in tx-parser.
This can be simplified when
https://github.com/Marwes/combine/issues/86 makes it to a published
release, but this unblocks us for now. This converts the `combine`
error type `ParseError<&'a [edn::Value]>` to a type with owned
`Vec<edn::Value>` collections, re-using `edn::Value::Vector` for
making them `Display`.
* Pre: Accept Borrow<Schema> instead of just &Schema in debug module.
This makes it easy to use Rc<Schema> or Arc<Schema> without inserting
&* sigils throughout the code.
* Use error-chain in query-parser.
There are a few things to point out here:
- the fine grained error types have been flattened into one crate-wide
error type; it's pretty easy to regain the granularity as needed.
- edn::ParseError is automatically lifted to
mentat_query_parser::errors::Error;
- we use mentat_parser_utils::ValueParser to maintain parsing error
information from `combine`.
* Patch up top-level.
* Review comment: Only `borrow()` once.
2017-02-24 23:32:41 +00:00
datoms_after ( conn , schema , bootstrap ::TX0 - 1 )
2017-02-08 22:04:32 +00:00
}
/// Return the set of datoms in the store with transaction ID strictly greater than the given `tx`,
/// ordered by (e, a, v, tx).
///
/// The datom set returned does not include any datoms of the form [... :db/txInstant ...].
2020-01-14 15:46:21 +00:00
pub fn datoms_after < S : Borrow < Schema > > (
conn : & rusqlite ::Connection ,
schema : & S ,
tx : i64 ,
) -> Result < Datoms > {
Schema alteration. Fixes #294 and #295. (#370) r=rnewman
* Pre: Don't retract :db/ident in test.
Datomic (and eventually Mentat) don't allow to retract :db/ident in
this way, so this runs afoul of future work to support mutating
metadata.
* Pre: s/VALUETYPE/VALUE_TYPE/.
This is consistent with the capitalization (which is "valueType") and
the other identifier.
* Pre: Remove some single quotes from error output.
* Part 1: Make materialized views be uniform [e a v value_type_tag].
This looks ahead to a time when we could support arbitrary
user-defined materialized views. For now, the "idents" materialized
view is those datoms of the form [e :db/ident :namespaced/keyword] and
the "schema" materialized view is those datoms of the form [e a v]
where a is in a particular set of attributes that will become clear in
the following commits.
This change is not backwards compatible, so I'm removing the open
current (really, v2) test. It'll be re-instated when we get to
https://github.com/mozilla/mentat/issues/194.
* Pre: Map TypedValue::Ref to TypedValue::Keyword in debug output.
* Part 3: Separate `schema_to_mutate` from the `schema` used to interpret.
This is just to keep track of the expected changes during
bootstrapping. I want bootstrap metadata mutations to flow through
the same code path as metadata mutations during regular transactions;
by differentiating the schema used for interpretation from the schema
that will be updated I expect to be able to apply bootstrap metadata
mutations to an empty schema and have things like materialized views
created (using the regular code paths).
This commit has been re-ordered for conceptual clarity, but it won't
compile because it references the metadata module. It's possible to
make it compile -- the functionality is there in the schema module --
but it's not worth the rebasing effort until after review (and
possibly not even then, since we'll squash down to a single commit to
land).
* Part 2: Maintain entids separately from idents.
In order to support historical idents, we need to distinguish the
"current" map from entid -> ident from the "complete historical" map
ident -> entid. This is what Datomic does; in Datomic, an ident is
never retracted (although it can be replaced). This approach is an
important part of allowing multiple consumers to share a schema
fragment as it migrates forward.
This fixes a limitation of the Clojure implementation, which did not
handle historical idents across knowledge base close and re-open.
The "entids" materialized view is naturally a slice of the "datoms"
table. The "idents" materialized view is a slice of the
"transactions" table. I hope that representing in this way, and
casting the problem in this light, might generalize to future
materialized views.
* Pre: Add DiffSet.
* Part 4: Collect mutations to a `Schema`.
I haven't taken your review comment about consuming AttributeBuilder
during each fluent function. If you read my response and still want
this, I'm happy to do it in review.
* Part 5: Handle :db/ident and :db.{install,alter}/attribute.
This "loops" the committed datoms out of the SQL store and back
through the metadata (schema, but in future also partition map)
processor. The metadata processor updates the schema and produces a
report of what changed; that report is then used to update the SQL
store. That update includes:
- the materialized views ("entids", "idents", and "schema");
- if needed, a subset of the datoms themselves (as flags change).
I've left a TODO for handling attribute retraction in the cases that
it makes sense. I expect that to be straight-forward.
* Review comment: Rename DiffSet to AddRetractAlterSet.
Also adds a little more commentary and a simple test.
* Review comment: Use ToIdent trait.
* Review comment: partially revert "Part 2: Maintain entids separately from idents."
This reverts commit 23a91df9c35e14398f2ddbd1ba25315821e67401.
Following our discussion, this removes the "entids" materialized
view. The next commit will remove historical idents from the "idents"
materialized view.
* Post: Use custom Either rather than std::result::Result.
This is not necessary, but it was suggested that we might be paying an
overhead creating Err instances while using error_chain. That seems
not to be the case, but this change shows that we don't actually use
any of the Result helper methods, so there's no reason to overload
Result. This change might avoid some future confusion, so I'm going
to land it anyway.
Signed-off-by: Nick Alexander <nalexander@mozilla.com>
* Review comment: Don't preserve historical idents.
* Review comment: More prepared statements when updating materialized views.
* Post: Test altering :db/cardinality and :db/unique.
These tests fail due to a Datomic limitation, namely that the marker
flag :db.alter/attribute can only be asserted once for an attribute!
That is, [:db.part/db :db.alter/attribute :attribute] will only be
transacted at most once. Since older versions of Datomic required the
:db.alter/attribute flag, I can only imagine they either never wrote
:db.alter/attribute to the store, or they handled it specially. I'll
need to remove the marker flag system from Mentat in order to address
this fundamental limitation.
* Post: Remove some more single quotes from error output.
* Post: Add assert_transact! macro to unwrap safely.
I was finding it very difficult to track unwrapping errors while
making changes, due to an underlying Mac OS X symbolication issue that
makes running tests with RUST_BACKTRACE=1 so slow that they all time
out.
* Post: Don't expect or recognize :db.{install,alter}/attribute.
I had this all working... except we will never see a repeated
`[:db.part/db :db.alter/attribute :attribute]` assertion in the store!
That means my approach would let you alter an attribute at most one
time. It's not worth hacking around this; it's better to just stop
expecting (and recognizing) the marker flags. (We have all the data
to distinguish the various cases that we need without the marker
flags.)
This brings Mentat in line with the thrust of newer Datomic versions,
but isn't compatible with Datomic, because (if I understand correctly)
Datomic automatically adds :db.{install,alter}/attribute assertions to
transactions.
I haven't purged the corresponding :db/ident and schema fragments just
yet:
- we might want them back
- we might want them in order to upgrade v1 and v2 databases to the
new on-disk layout we're fleshing out (v3?).
* Post: Don't make :db/unique :db.unique/* imply :db/index true.
This patch avoids a potential bug with the "schema" materialized view.
If :db/unique :db.unique/value implies :db/index true, then what
happens when you _retract_ :db.unique/value? I think Datomic defines
this in some way, but I really want the "schema" materialized view to
be a slice of "datoms" and not have these sort of ambiguities and
persistent effects. Therefore, to ensure that we don't retract a
schema characteristic and accidentally change more than we intended
to, this patch stops having any schema characteristic imply any other
schema characteristic(s). To achieve that, I added an
Option<Unique::{Value,Identity}> type to Attribute; this helps with
this patch, and also looks ahead to when we allow to retract
:db/unique attributes.
* Post: Allow to retract :db/ident.
* Post: Include more details about invalid schema changes.
The tests use strings, so they hide the chained errors which do in
fact provide more detail.
* Review comment: Fix outdated comment.
* Review comment: s/_SET/_SQL_LIST/.
* Review comment: Use a sub-select for checking cardinality.
This might be faster in practice.
* Review comment: Put `attribute::Unique` into its own namespace.
2017-03-20 20:18:59 +00:00
let borrowed_schema = schema . borrow ( ) ;
2017-03-20 18:29:17 +00:00
let mut stmt : rusqlite ::Statement = conn . prepare ( " SELECT e, a, v, value_type_tag, tx FROM datoms WHERE tx > ? ORDER BY e ASC, a ASC, value_type_tag ASC, v ASC, tx ASC " ) ? ;
2017-02-08 22:04:32 +00:00
2020-01-14 15:46:21 +00:00
let r : Result < Vec < _ > > = stmt
. query_and_then ( & [ & tx ] , | row | {
let e : i64 = row . get ( 0 ) ? ;
let a : i64 = row . get ( 1 ) ? ;
2017-02-08 22:04:32 +00:00
2020-01-14 15:46:21 +00:00
if a = = entids ::DB_TX_INSTANT {
return Ok ( None ) ;
}
2017-02-08 22:04:32 +00:00
2020-01-14 15:46:21 +00:00
let v : rusqlite ::types ::Value = row . get ( 2 ) ? ;
let value_type_tag : i32 = row . get ( 3 ) ? ;
2017-02-08 22:04:32 +00:00
2020-01-14 15:46:21 +00:00
let attribute = borrowed_schema . require_attribute_for_entid ( a ) ? ;
let value_type_tag = if ! attribute . fulltext {
value_type_tag
} else {
ValueType ::Long . value_type_tag ( )
} ;
2017-03-21 20:12:10 +00:00
2020-01-14 15:46:21 +00:00
let typed_value =
TypedValue ::from_sql_value_pair ( v , value_type_tag ) ? . map_ident ( borrowed_schema ) ;
let ( value , _ ) = typed_value . to_edn_value_pair ( ) ;
2017-02-08 22:04:32 +00:00
2020-01-14 15:46:21 +00:00
let tx : i64 = row . get ( 4 ) ? ;
2017-02-08 22:04:32 +00:00
2020-01-14 15:46:21 +00:00
Ok ( Some ( Datom {
e : EntidOrIdent ::Entid ( e ) ,
a : to_entid ( borrowed_schema , a ) ,
v : value ,
2020-01-23 18:16:19 +00:00
tx ,
2020-01-14 15:46:21 +00:00
added : None ,
} ) )
} ) ?
. collect ( ) ;
2017-02-08 22:04:32 +00:00
Ok ( Datoms ( r ? . into_iter ( ) . filter_map ( | x | x ) . collect ( ) ) )
}
2017-01-26 00:13:56 +00:00
2017-02-08 22:04:32 +00:00
/// Return the sequence of transactions in the store with transaction ID strictly greater than the
/// given `tx`, ordered by (tx, e, a, v).
///
2018-04-26 22:48:27 +00:00
/// Each transaction returned includes the [(transaction-tx) :db/txInstant ...] datom.
2020-01-14 15:46:21 +00:00
pub fn transactions_after < S : Borrow < Schema > > (
conn : & rusqlite ::Connection ,
schema : & S ,
tx : i64 ,
) -> Result < Transactions > {
Schema alteration. Fixes #294 and #295. (#370) r=rnewman
* Pre: Don't retract :db/ident in test.
Datomic (and eventually Mentat) don't allow to retract :db/ident in
this way, so this runs afoul of future work to support mutating
metadata.
* Pre: s/VALUETYPE/VALUE_TYPE/.
This is consistent with the capitalization (which is "valueType") and
the other identifier.
* Pre: Remove some single quotes from error output.
* Part 1: Make materialized views be uniform [e a v value_type_tag].
This looks ahead to a time when we could support arbitrary
user-defined materialized views. For now, the "idents" materialized
view is those datoms of the form [e :db/ident :namespaced/keyword] and
the "schema" materialized view is those datoms of the form [e a v]
where a is in a particular set of attributes that will become clear in
the following commits.
This change is not backwards compatible, so I'm removing the open
current (really, v2) test. It'll be re-instated when we get to
https://github.com/mozilla/mentat/issues/194.
* Pre: Map TypedValue::Ref to TypedValue::Keyword in debug output.
* Part 3: Separate `schema_to_mutate` from the `schema` used to interpret.
This is just to keep track of the expected changes during
bootstrapping. I want bootstrap metadata mutations to flow through
the same code path as metadata mutations during regular transactions;
by differentiating the schema used for interpretation from the schema
that will be updated I expect to be able to apply bootstrap metadata
mutations to an empty schema and have things like materialized views
created (using the regular code paths).
This commit has been re-ordered for conceptual clarity, but it won't
compile because it references the metadata module. It's possible to
make it compile -- the functionality is there in the schema module --
but it's not worth the rebasing effort until after review (and
possibly not even then, since we'll squash down to a single commit to
land).
* Part 2: Maintain entids separately from idents.
In order to support historical idents, we need to distinguish the
"current" map from entid -> ident from the "complete historical" map
ident -> entid. This is what Datomic does; in Datomic, an ident is
never retracted (although it can be replaced). This approach is an
important part of allowing multiple consumers to share a schema
fragment as it migrates forward.
This fixes a limitation of the Clojure implementation, which did not
handle historical idents across knowledge base close and re-open.
The "entids" materialized view is naturally a slice of the "datoms"
table. The "idents" materialized view is a slice of the
"transactions" table. I hope that representing in this way, and
casting the problem in this light, might generalize to future
materialized views.
* Pre: Add DiffSet.
* Part 4: Collect mutations to a `Schema`.
I haven't taken your review comment about consuming AttributeBuilder
during each fluent function. If you read my response and still want
this, I'm happy to do it in review.
* Part 5: Handle :db/ident and :db.{install,alter}/attribute.
This "loops" the committed datoms out of the SQL store and back
through the metadata (schema, but in future also partition map)
processor. The metadata processor updates the schema and produces a
report of what changed; that report is then used to update the SQL
store. That update includes:
- the materialized views ("entids", "idents", and "schema");
- if needed, a subset of the datoms themselves (as flags change).
I've left a TODO for handling attribute retraction in the cases that
it makes sense. I expect that to be straight-forward.
* Review comment: Rename DiffSet to AddRetractAlterSet.
Also adds a little more commentary and a simple test.
* Review comment: Use ToIdent trait.
* Review comment: partially revert "Part 2: Maintain entids separately from idents."
This reverts commit 23a91df9c35e14398f2ddbd1ba25315821e67401.
Following our discussion, this removes the "entids" materialized
view. The next commit will remove historical idents from the "idents"
materialized view.
* Post: Use custom Either rather than std::result::Result.
This is not necessary, but it was suggested that we might be paying an
overhead creating Err instances while using error_chain. That seems
not to be the case, but this change shows that we don't actually use
any of the Result helper methods, so there's no reason to overload
Result. This change might avoid some future confusion, so I'm going
to land it anyway.
Signed-off-by: Nick Alexander <nalexander@mozilla.com>
* Review comment: Don't preserve historical idents.
* Review comment: More prepared statements when updating materialized views.
* Post: Test altering :db/cardinality and :db/unique.
These tests fail due to a Datomic limitation, namely that the marker
flag :db.alter/attribute can only be asserted once for an attribute!
That is, [:db.part/db :db.alter/attribute :attribute] will only be
transacted at most once. Since older versions of Datomic required the
:db.alter/attribute flag, I can only imagine they either never wrote
:db.alter/attribute to the store, or they handled it specially. I'll
need to remove the marker flag system from Mentat in order to address
this fundamental limitation.
* Post: Remove some more single quotes from error output.
* Post: Add assert_transact! macro to unwrap safely.
I was finding it very difficult to track unwrapping errors while
making changes, due to an underlying Mac OS X symbolication issue that
makes running tests with RUST_BACKTRACE=1 so slow that they all time
out.
* Post: Don't expect or recognize :db.{install,alter}/attribute.
I had this all working... except we will never see a repeated
`[:db.part/db :db.alter/attribute :attribute]` assertion in the store!
That means my approach would let you alter an attribute at most one
time. It's not worth hacking around this; it's better to just stop
expecting (and recognizing) the marker flags. (We have all the data
to distinguish the various cases that we need without the marker
flags.)
This brings Mentat in line with the thrust of newer Datomic versions,
but isn't compatible with Datomic, because (if I understand correctly)
Datomic automatically adds :db.{install,alter}/attribute assertions to
transactions.
I haven't purged the corresponding :db/ident and schema fragments just
yet:
- we might want them back
- we might want them in order to upgrade v1 and v2 databases to the
new on-disk layout we're fleshing out (v3?).
* Post: Don't make :db/unique :db.unique/* imply :db/index true.
This patch avoids a potential bug with the "schema" materialized view.
If :db/unique :db.unique/value implies :db/index true, then what
happens when you _retract_ :db.unique/value? I think Datomic defines
this in some way, but I really want the "schema" materialized view to
be a slice of "datoms" and not have these sort of ambiguities and
persistent effects. Therefore, to ensure that we don't retract a
schema characteristic and accidentally change more than we intended
to, this patch stops having any schema characteristic imply any other
schema characteristic(s). To achieve that, I added an
Option<Unique::{Value,Identity}> type to Attribute; this helps with
this patch, and also looks ahead to when we allow to retract
:db/unique attributes.
* Post: Allow to retract :db/ident.
* Post: Include more details about invalid schema changes.
The tests use strings, so they hide the chained errors which do in
fact provide more detail.
* Review comment: Fix outdated comment.
* Review comment: s/_SET/_SQL_LIST/.
* Review comment: Use a sub-select for checking cardinality.
This might be faster in practice.
* Review comment: Put `attribute::Unique` into its own namespace.
2017-03-20 20:18:59 +00:00
let borrowed_schema = schema . borrow ( ) ;
2017-03-20 18:29:17 +00:00
let mut stmt : rusqlite ::Statement = conn . prepare ( " SELECT e, a, v, value_type_tag, tx, added FROM transactions WHERE tx > ? ORDER BY tx ASC, e ASC, a ASC, value_type_tag ASC, v ASC, added ASC " ) ? ;
2017-01-26 00:13:56 +00:00
2020-01-14 15:46:21 +00:00
let r : Result < Vec < _ > > = stmt
. query_and_then ( & [ & tx ] , | row | {
let e : i64 = row . get ( 0 ) ? ;
let a : i64 = row . get ( 1 ) ? ;
let v : rusqlite ::types ::Value = row . get ( 2 ) ? ;
let value_type_tag : i32 = row . get ( 3 ) ? ;
let attribute = borrowed_schema . require_attribute_for_entid ( a ) ? ;
let value_type_tag = if ! attribute . fulltext {
value_type_tag
} else {
ValueType ::Long . value_type_tag ( )
} ;
let typed_value =
TypedValue ::from_sql_value_pair ( v , value_type_tag ) ? . map_ident ( borrowed_schema ) ;
let ( value , _ ) = typed_value . to_edn_value_pair ( ) ;
let tx : i64 = row . get ( 4 ) ? ;
let added : bool = row . get ( 5 ) ? ;
Ok ( Datom {
e : EntidOrIdent ::Entid ( e ) ,
a : to_entid ( borrowed_schema , a ) ,
v : value ,
2020-01-23 18:16:19 +00:00
tx ,
2020-01-14 15:46:21 +00:00
added : Some ( added ) ,
} )
} ) ?
. collect ( ) ;
2017-02-08 22:04:32 +00:00
// Group by tx.
2020-01-14 15:46:21 +00:00
let r : Vec < Datoms > = r ?
. into_iter ( )
. group_by ( | x | x . tx )
. into_iter ( )
. map ( | ( _key , group ) | Datoms ( group . collect ( ) ) )
. collect ( ) ;
2017-02-08 22:04:32 +00:00
Ok ( Transactions ( r ) )
}
2017-03-21 20:12:10 +00:00
/// Return the set of fulltext values in the store, ordered by rowid.
2018-07-16 20:58:34 +00:00
pub fn fulltext_values ( conn : & rusqlite ::Connection ) -> Result < FulltextValues > {
2020-01-14 15:46:21 +00:00
let mut stmt : rusqlite ::Statement =
conn . prepare ( " SELECT rowid, text FROM fulltext_values ORDER BY rowid " ) ? ;
2017-03-21 20:12:10 +00:00
2020-01-14 15:46:21 +00:00
let r : Result < Vec < _ > > = stmt
2021-07-03 00:29:41 +00:00
. query_and_then ( [ ] , | row | {
2020-01-14 15:46:21 +00:00
let rowid : i64 = row . get ( 0 ) ? ;
let text : String = row . get ( 1 ) ? ;
Ok ( ( rowid , text ) )
} ) ?
. collect ( ) ;
2017-03-21 20:12:10 +00:00
r . map ( FulltextValues )
}
2017-02-08 22:04:32 +00:00
/// Execute the given `sql` query with the given `params` and format the results as a
/// tab-and-newline formatted string suitable for debug printing.
///
/// The query is printed followed by a newline, then the returned columns followed by a newline, and
/// then the data rows and columns. All columns are aligned.
2020-01-14 15:46:21 +00:00
pub fn dump_sql_query (
conn : & rusqlite ::Connection ,
sql : & str ,
params : & [ & dyn ToSql ] ,
) -> Result < String > {
2017-02-08 22:04:32 +00:00
let mut stmt : rusqlite ::Statement = conn . prepare ( sql ) ? ;
let mut tw = TabWriter ::new ( Vec ::new ( ) ) . padding ( 2 ) ;
2020-01-23 18:16:19 +00:00
writeln! ( & mut tw , " {} " , sql ) . unwrap ( ) ;
2017-02-08 22:04:32 +00:00
for column_name in stmt . column_names ( ) {
write! ( & mut tw , " {} \t " , column_name ) . unwrap ( ) ;
}
2020-01-23 18:16:19 +00:00
writeln! ( & mut tw ) . unwrap ( ) ;
2017-02-08 22:04:32 +00:00
2020-01-14 15:46:21 +00:00
let r : Result < Vec < _ > > = stmt
. query_and_then ( params , | row | {
2021-11-11 03:17:19 +00:00
for i in 0 .. row . as_ref ( ) . column_count ( ) {
2020-01-14 15:46:21 +00:00
let value : rusqlite ::types ::Value = row . get ( i ) ? ;
write! ( & mut tw , " {:?} \t " , value ) . unwrap ( ) ;
}
2020-01-23 18:16:19 +00:00
writeln! ( & mut tw ) . unwrap ( ) ;
2020-01-14 15:46:21 +00:00
Ok ( ( ) )
} ) ?
. collect ( ) ;
2017-02-08 22:04:32 +00:00
r ? ;
let dump = String ::from_utf8 ( tw . into_inner ( ) . unwrap ( ) ) . unwrap ( ) ;
Ok ( dump )
2017-01-26 00:13:56 +00:00
}
2018-07-16 20:58:34 +00:00
// A connection that doesn't try to be clever about possibly sharing its `Schema`. Compare to
// `mentat::Conn`.
pub struct TestConn {
pub sqlite : rusqlite ::Connection ,
pub partition_map : PartitionMap ,
pub schema : Schema ,
}
impl TestConn {
fn assert_materialized_views ( & self ) {
let materialized_ident_map = read_ident_map ( & self . sqlite ) . expect ( " ident map " ) ;
let materialized_attribute_map = read_attribute_map ( & self . sqlite ) . expect ( " schema map " ) ;
2020-01-14 15:46:21 +00:00
let materialized_schema = Schema ::from_ident_map_and_attribute_map (
materialized_ident_map ,
materialized_attribute_map ,
)
. expect ( " schema " ) ;
2018-07-16 20:58:34 +00:00
assert_eq! ( materialized_schema , self . schema ) ;
}
2020-01-14 15:46:21 +00:00
pub fn transact < I > ( & mut self , transaction : I ) -> Result < TxReport >
where
I : Borrow < str > ,
{
2018-07-16 20:58:34 +00:00
// Failure to parse the transaction is a coding error, so we unwrap.
2020-01-23 18:16:19 +00:00
let entities = edn ::parse ::entities ( transaction . borrow ( ) ) . unwrap_or_else ( | _ | {
panic! ( " to be able to parse {} into entities " , transaction . borrow ( ) )
} ) ;
2018-07-16 20:58:34 +00:00
let details = {
// The block scopes the borrow of self.sqlite.
// We're about to write, so go straight ahead and get an IMMEDIATE transaction.
2020-01-14 15:46:21 +00:00
let tx = self
. sqlite
. transaction_with_behavior ( TransactionBehavior ::Immediate ) ? ;
2018-07-16 20:58:34 +00:00
// Applying the transaction can fail, so we don't unwrap.
2020-01-14 15:46:21 +00:00
let details = transact (
& tx ,
self . partition_map . clone ( ) ,
& self . schema ,
& self . schema ,
NullWatcher ( ) ,
entities ,
) ? ;
2018-07-16 20:58:34 +00:00
tx . commit ( ) ? ;
details
} ;
let ( report , next_partition_map , next_schema , _watcher ) = details ;
self . partition_map = next_partition_map ;
if let Some ( next_schema ) = next_schema {
self . schema = next_schema ;
}
// Verify that we've updated the materialized views during transacting.
self . assert_materialized_views ( ) ;
Ok ( report )
}
2020-01-14 15:46:21 +00:00
pub fn transact_simple_terms < I > (
& mut self ,
terms : I ,
tempid_set : InternSet < TempId > ,
) -> Result < TxReport >
where
I : IntoIterator < Item = TermWithTempIds > ,
{
2018-07-16 20:58:34 +00:00
let details = {
// The block scopes the borrow of self.sqlite.
// We're about to write, so go straight ahead and get an IMMEDIATE transaction.
2020-01-14 15:46:21 +00:00
let tx = self
. sqlite
. transaction_with_behavior ( TransactionBehavior ::Immediate ) ? ;
2018-07-16 20:58:34 +00:00
// Applying the transaction can fail, so we don't unwrap.
2020-01-14 15:46:21 +00:00
let details = transact_terms (
& tx ,
self . partition_map . clone ( ) ,
& self . schema ,
& self . schema ,
NullWatcher ( ) ,
terms ,
tempid_set ,
) ? ;
2018-07-16 20:58:34 +00:00
tx . commit ( ) ? ;
details
} ;
let ( report , next_partition_map , next_schema , _watcher ) = details ;
self . partition_map = next_partition_map ;
if let Some ( next_schema ) = next_schema {
self . schema = next_schema ;
}
// Verify that we've updated the materialized views during transacting.
self . assert_materialized_views ( ) ;
Ok ( report )
}
pub fn last_tx_id ( & self ) -> Entid {
2020-01-14 15:46:21 +00:00
self . partition_map
. get ( & " :db.part/tx " . to_string ( ) )
. unwrap ( )
. next_entid ( )
- 1
2018-07-16 20:58:34 +00:00
}
pub fn last_transaction ( & self ) -> Datoms {
2020-01-14 15:46:21 +00:00
transactions_after ( & self . sqlite , & self . schema , self . last_tx_id ( ) - 1 )
. expect ( " last_transaction " )
. 0
. pop ( )
. unwrap ( )
2018-07-16 20:58:34 +00:00
}
pub fn transactions ( & self ) -> Transactions {
transactions_after ( & self . sqlite , & self . schema , bootstrap ::TX0 ) . expect ( " transactions " )
}
pub fn datoms ( & self ) -> Datoms {
datoms_after ( & self . sqlite , & self . schema , bootstrap ::TX0 ) . expect ( " datoms " )
}
pub fn fulltext_values ( & self ) -> FulltextValues {
fulltext_values ( & self . sqlite ) . expect ( " fulltext_values " )
}
pub fn with_sqlite ( mut conn : rusqlite ::Connection ) -> TestConn {
let db = ensure_current_version ( & mut conn ) . unwrap ( ) ;
// Does not include :db/txInstant.
let datoms = datoms_after ( & conn , & db . schema , 0 ) . unwrap ( ) ;
assert_eq! ( datoms . 0. len ( ) , 94 ) ;
// Includes :db/txInstant.
let transactions = transactions_after ( & conn , & db . schema , 0 ) . unwrap ( ) ;
assert_eq! ( transactions . 0. len ( ) , 1 ) ;
assert_eq! ( transactions . 0 [ 0 ] . 0. len ( ) , 95 ) ;
let mut parts = db . partition_map ;
// Add a fake partition to allow tests to do things like
// [:db/add 111 :foo/bar 222]
{
2018-07-12 21:42:59 +00:00
let fake_partition = Partition ::new ( 100 , 2000 , 1000 , true ) ;
2018-07-16 20:58:34 +00:00
parts . insert ( " :db.part/fake " . into ( ) , fake_partition ) ;
}
let test_conn = TestConn {
sqlite : conn ,
partition_map : parts ,
schema : db . schema ,
} ;
// Verify that we've created the materialized views during bootstrapping.
test_conn . assert_materialized_views ( ) ;
test_conn
}
2018-07-20 20:11:34 +00:00
pub fn sanitized_partition_map ( & mut self ) {
self . partition_map . remove ( " :db.part/fake " ) ;
}
2018-07-16 20:58:34 +00:00
}
impl Default for TestConn {
fn default ( ) -> TestConn {
TestConn ::with_sqlite ( new_connection ( " " ) . expect ( " Couldn't open in-memory db " ) )
}
}
pub struct TempIds ( edn ::Value ) ;
impl TempIds {
pub fn to_edn ( & self ) -> edn ::Value {
self . 0. clone ( )
}
}
pub fn tempids ( report : & TxReport ) -> TempIds {
let mut map : BTreeMap < edn ::Value , edn ::Value > = BTreeMap ::default ( ) ;
for ( tempid , & entid ) in report . tempids . iter ( ) {
map . insert ( edn ::Value ::Text ( tempid . clone ( ) ) , edn ::Value ::Integer ( entid ) ) ;
}
TempIds ( edn ::Value ::Map ( map ) )
}