Setup framework to do integration testing (#50)

.github/workflows/tests.yaml

@@ -17,6 +17,7 @@ jobs:
           profile: minimal
           toolchain: stable
           override: true
+      - uses: Swatinem/rust-cache@v1
       - uses: actions-rs/cargo@v1
         with:
           command: check
@@ -31,9 +32,12 @@ jobs:
           profile: minimal
           toolchain: stable
           override: true
+      - uses: Swatinem/rust-cache@v1
+      - run: cd compressor_integration_tests && docker-compose up -d
       - uses: actions-rs/cargo@v1
         with:
           command: test
+          args: --workspace
 
   fmt:
     name: Rustfmt
@@ -45,7 +49,8 @@ jobs:
           profile: minimal
           toolchain: stable
           override: true
-      - run: rustup component add rustfmt
+          components: rustfmt
+      - uses: Swatinem/rust-cache@v1
       - uses: actions-rs/cargo@v1
         with:
           command: fmt
@@ -61,7 +66,8 @@ jobs:
           profile: minimal
           toolchain: stable
           override: true
-      - run: rustup component add clippy
+          components: clippy
+      - uses: Swatinem/rust-cache@v1
       - uses: actions-rs/cargo@v1
         with:
           command: clippy

.gitignore

@@ -2,5 +2,5 @@
 **/*.rs.bk
 *.data
 *.old
-out.sql
+**.sql
 *.csv

Cargo.lock

@@ -99,6 +99,20 @@ dependencies = [
  "vec_map",
 ]
 
+[[package]]
+name = "compressor_integration_tests"
+version = "0.1.0"
+dependencies = [
+ "openssl",
+ "postgres",
+ "postgres-openssl",
+ "rand",
+ "serial_test",
+ "state-map",
+ "string_cache",
+ "synapse_compress_state",
+]
+
 [[package]]
 name = "console"
 version = "0.14.1"
@@ -909,6 +923,28 @@ version = "1.0.126"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "ec7505abeacaec74ae4778d9d9328fe5a5d04253220a85c4ee022239fc996d03"
 
+[[package]]
+name = "serial_test"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e0bccbcf40c8938196944a3da0e133e031a33f4d6b72db3bda3cc556e361905d"
+dependencies = [
+ "lazy_static",
+ "parking_lot",
+ "serial_test_derive",
+]
+
+[[package]]
+name = "serial_test_derive"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b2acd6defeddb41eb60bb468f8825d0cfd0c2a76bc03bfd235b6a1dc4f6a1ad5"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
 [[package]]
 name = "sha2"
 version = "0.9.5"

Cargo.toml

@@ -1,3 +1,6 @@
+[workspace]
+members = ["compressor_integration_tests"]
+
 [package]
 authors = ["Erik Johnston"]
 description = "A tool to compress some state in a Synapse instance's database"

compressor_integration_tests/Cargo.toml

@@ -0,0 +1,18 @@
+[package]
+name = "compressor_integration_tests"
+version = "0.1.0"
+edition = "2018"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+string_cache = "0.8.0"
+serial_test = "0.5.1"
+openssl = "0.10.32"
+postgres = "0.19.0"
+postgres-openssl = "0.5.0"
+rand = "0.8.0"
+synapse_compress_state = { path = "../" }
+
+[dependencies.state-map]
+git = "https://github.com/matrix-org/rust-matrix-state-map"

compressor_integration_tests/database_setup.sh

@@ -0,0 +1,28 @@
+#!/bin/sh
+
+#N.B. the database setup comes from:
+#https://github.com/matrix-org/synapse/blob/develop/synapse/storage/schema/state/full_schemas/54/full.sql
+
+# Setup the required tables for testing
+psql --username "$POSTGRES_USER" --dbname "$POSTGRES_DB" <<SQLCODE
+
+CREATE TABLE state_groups (
+    id BIGINT PRIMARY KEY,
+    room_id TEXT NOT NULL,
+    event_id TEXT NOT NULL
+);
+
+CREATE TABLE state_groups_state (
+    state_group BIGINT NOT NULL,
+    room_id TEXT NOT NULL,
+    type TEXT NOT NULL,
+    state_key TEXT NOT NULL,
+    event_id TEXT NOT NULL
+);
+
+CREATE TABLE state_group_edges (
+    state_group BIGINT NOT NULL,
+    prev_state_group BIGINT NOT NULL
+);
+
+SQLCODE

compressor_integration_tests/docker-compose.yaml

@@ -0,0 +1,20 @@
+version: '3'
+services:
+  postgres:
+    image: "postgres:latest"
+    
+    ports:
+      # N.B. format is [port on machine]:[port to expose from container]
+      - 5432:5432
+
+    environment: 
+      POSTGRES_USER: synapse_user
+      POSTGRES_PASSWORD: synapse_pass
+      POSTGRES_DB: synapse
+      PGDATA: /tmp/data
+
+    volumes:
+      - ./database_setup.sh:/docker-entrypoint-initdb.d/1_database_setup.sh 
+
+    tmpfs:
+      /tmp/data

compressor_integration_tests/src/lib.rs

@@ -0,0 +1,107 @@
+use openssl::ssl::{SslConnector, SslMethod, SslVerifyMode};
+use postgres::Client;
+use postgres_openssl::MakeTlsConnector;
+use rand::{distributions::Alphanumeric, thread_rng, Rng};
+use std::{borrow::Cow, collections::BTreeMap, fmt};
+
+use synapse_compress_state::StateGroupEntry;
+
+pub mod map_builder;
+
+pub static DB_URL: &str = "postgresql://synapse_user:synapse_pass@localhost/synapse";
+
+/// Adds the contents of a state group map to the testing database
+pub fn add_contents_to_database(room_id: &str, state_group_map: &BTreeMap<i64, StateGroupEntry>) {
+    // connect to the database
+    let mut builder = SslConnector::builder(SslMethod::tls()).unwrap();
+    builder.set_verify(SslVerifyMode::NONE);
+    let connector = MakeTlsConnector::new(builder.build());
+
+    let mut client = Client::connect(DB_URL, connector).unwrap();
+
+    // build up the query
+    let mut sql = "".to_string();
+
+    for (sg, entry) in state_group_map {
+        // create the entry for state_groups
+        sql.push_str(&format!(
+            "INSERT INTO state_groups (id, room_id, event_id) VALUES ({},{},{});\n",
+            sg,
+            PGEscape(room_id),
+            PGEscape("left_blank")
+        ));
+
+        // create the entry in state_group_edges IF exists
+        if let Some(prev_sg) = entry.prev_state_group {
+            sql.push_str(&format!(
+                "INSERT INTO state_group_edges (state_group, prev_state_group) VALUES ({}, {});\n",
+                sg, prev_sg
+            ));
+        }
+
+        // write entry for each row in delta
+        if !entry.state_map.is_empty() {
+            sql.push_str("INSERT INTO state_groups_state (state_group, room_id, type, state_key, event_id) VALUES");
+
+            let mut first = true;
+            for ((t, s), e) in entry.state_map.iter() {
+                if first {
+                    sql.push_str("     ");
+                    first = false;
+                } else {
+                    sql.push_str("    ,");
+                }
+                sql.push_str(&format!(
+                    "({}, {}, {}, {}, {})",
+                    sg,
+                    PGEscape(room_id),
+                    PGEscape(t),
+                    PGEscape(s),
+                    PGEscape(e)
+                ));
+            }
+            sql.push_str(";\n");
+        }
+    }
+
+    client.batch_execute(&sql).unwrap();
+}
+
+// Clears the contents of the testing database
+pub fn empty_database() {
+    // connect to the database
+    let mut builder = SslConnector::builder(SslMethod::tls()).unwrap();
+    builder.set_verify(SslVerifyMode::NONE);
+    let connector = MakeTlsConnector::new(builder.build());
+
+    let mut client = Client::connect(DB_URL, connector).unwrap();
+
+    // delete all the contents from all three tables
+    let sql = r"
+        DELETE FROM state_groups;
+        DELETE FROM state_group_edges;
+        DELETE FROM state_groups_state;
+    ";
+
+    client.batch_execute(sql).unwrap();
+}
+
+/// Safely escape the strings in sql queries
+struct PGEscape<'a>(pub &'a str);
+
+impl<'a> fmt::Display for PGEscape<'a> {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        let mut delim = Cow::from("$$");
+        while self.0.contains(&delim as &str) {
+            let s: String = thread_rng()
+                .sample_iter(&Alphanumeric)
+                .take(10)
+                .map(char::from)
+                .collect();
+
+            delim = format!("${}$", s).into();
+        }
+
+        write!(f, "{}{}{}", delim, self.0, delim)
+    }
+}

compressor_integration_tests/src/map_builder.rs

@@ -0,0 +1,36 @@
+use std::collections::BTreeMap;
+
+use state_map::StateMap;
+use synapse_compress_state::StateGroupEntry;
+
+/// Generates long chain of state groups each with state deltas
+///
+/// If called wiht start=0, end=13 this would build the following:
+///
+/// 0-1-2-3-4-5-6-7-8-9-10-11-12-13
+///
+/// Where each group i has state:
+///     ('node','is',      i)
+///     ('group',  j, 'seen') - for all j less than i
+pub fn line_with_state(start: i64, end: i64) -> BTreeMap<i64, StateGroupEntry> {
+    let mut initial: BTreeMap<i64, StateGroupEntry> = BTreeMap::new();
+    let mut prev = None;
+
+    for i in start..=end {
+        let mut entry = StateGroupEntry {
+            in_range: true,
+            prev_state_group: prev,
+            state_map: StateMap::new(),
+        };
+        entry
+            .state_map
+            .insert("group", &i.to_string(), "seen".into());
+        entry.state_map.insert("node", "is", i.to_string().into());
+
+        initial.insert(i, entry);
+
+        prev = Some(i)
+    }
+
+    initial
+}

compressor_integration_tests/tests/compressor_config_tests.rs

@@ -0,0 +1,55 @@
+use compressor_integration_tests::{
+    add_contents_to_database, empty_database, map_builder::line_with_state, DB_URL,
+};
+use serial_test::serial;
+use synapse_compress_state::{run, Config};
+
+// Remember to add #[serial(db)] before any test that access the database.
+// Only one test with this annotation can run at once - preventing
+// concurrency bugs.
+//
+// You will probably also want to use common::empty_database() at the start
+// of each test as well (since their order of execution is not guaranteed)
+
+#[test]
+#[serial(db)]
+fn run_succeeds_without_crashing() {
+    // This starts with the following structure
+    //
+    // 0-1-2-3-4-5-6-7-8-9-10-11-12-13
+    //
+    // Each group i has state:
+    //     ('node','is',      i)
+    //     ('group',  j, 'seen') - for all j less than i
+    let initial = line_with_state(0, 13);
+
+    empty_database();
+    add_contents_to_database("room1", &initial);
+
+    let db_url = DB_URL.to_string();
+    let room_id = "room1".to_string();
+    let output_file = Some("./tests/tmp/run_succeeds_without_crashing.sql".to_string());
+    let min_state_group = None;
+    let groups_to_compress = None;
+    let min_saved_rows = None;
+    let max_state_group = None;
+    let level_sizes = "3,3".to_string();
+    let transactions = true;
+    let graphs = false;
+
+    let config = Config::new(
+        db_url.clone(),
+        room_id.clone(),
+        output_file,
+        min_state_group,
+        groups_to_compress,
+        min_saved_rows,
+        max_state_group,
+        level_sizes,
+        transactions,
+        graphs,
+    )
+    .unwrap();
+
+    run(config);
+}

compressor_integration_tests/tests/tmp/README.md

@@ -0,0 +1 @@
+This folder is where sql files generated by the integration tests are saved

src/lib.rs

@@ -44,9 +44,9 @@ use database::PGEscape;
 /// delta from that previous group (or the full state if no previous group)
 #[derive(Default, Debug, Clone, PartialEq, Eq)]
 pub struct StateGroupEntry {
-    in_range: bool,
+    pub in_range: bool,
-    prev_state_group: Option<i64>,
+    pub prev_state_group: Option<i64>,
-    state_map: StateMap<Atom>,
+    pub state_map: StateMap<Atom>,
 }
 
 /// Helper struct for parsing the `level_sizes` argument.