Check the spelling in our code

.github/gh_matrix_builder.py

@@ -337,9 +337,9 @@ def macos_config(overrides):
             tests = set()
             break
         basename = os.path.basename(f)
-        splitted = basename.split(".")
-        name = splitted[0]
-        ext = splitted[-1]
+        split = basename.split(".")
+        name = split[0]
+        ext = split[-1]
         if ext == "out":
             # Account for the version number.
             tests.add(name)

.github/workflows/changelog-check.yaml

@@ -5,7 +5,7 @@ name: Check for changelog entry file
     branches:
       - main
 jobs:
-  # Check if the PR creates a sperate file with changelog entry in the
+  # Check if the PR creates a separate file with changelog entry in the
   # ".unreleased"  folder
   #
   # This check can be disabled by adding the following line in the PR text

.github/workflows/code_style.yaml

@@ -5,6 +5,7 @@ name: Code style
       - main
       - prerelease_test
   pull_request:
+
 jobs:
   cmake_checks:
     name: Check CMake files
@@ -53,6 +54,20 @@ jobs:
         run: |
           find . -type f \( -name "*.yaml" -or -name "*.yml" \) -print -exec yamllint {} \+
 
+  spelling_checks:
+    name: Check spelling
+    runs-on: ubuntu-latest
+    steps:
+      - name: Install prerequisites
+        run: |
+          pip install codespell
+      - name: Checkout source
+        uses: actions/checkout@v3
+      - name: Run codespell
+        run: |
+          find . -type f \( -name "*.c" -or -name "*.h" -or -name "*.yaml" -or -name "*.sh" \) \
+            -exec codespell -L "inh,larg,inout" {} \+
+
   cc_checks:
     name: Check code formatting
     runs-on: ubuntu-22.04
@@ -100,7 +115,7 @@ jobs:
           git diff --exit-code
 
   misc_checks:
-    name: Check license, update scripts, git hooks, missing gitignore entries and unecessary template tests
+    name: Check license, update scripts, git hooks, missing gitignore entries and unnecessary template tests
     runs-on: ubuntu-22.04
     strategy:
       fail-fast: false
@@ -124,7 +139,7 @@ jobs:
     - name: Check for missing gitignore entries for template test files
       if: always()
       run: ./scripts/check_missing_gitignore_for_template_tests.sh
-    - name: Check for unecessary template test files
+    - name: Check for unnecessary template test files
       if: always()
       run: ./scripts/check_unecessary_template_tests.sh
 

.github/workflows/update-test.yaml

@@ -78,7 +78,7 @@ jobs:
         UPDATE_FROM_TAG=${DOWNGRADE_TO}-pg${PG_MAJOR}
         export UPDATE_FROM_TAG
         # We need to use same libssl version used in the latest official TimescaleDB container images.
-        # So we will use the fixed alpine version, this will guarantee that libssl version wont change.
+        # So we will use the fixed alpine version, this will guarantee that libssl version won't change.
         PG_IMAGE_TAG="${PG_VERSION}-alpine3.17" scripts/test_downgrade_from_tag.sh
 
     - name: Downgrade diff

scripts/check_file_license.sh

@@ -58,7 +58,7 @@ check_file() {
             return 0;
             ;;
         (*)
-            echo "Unkown flag" ${1}
+            echo "Unknown flag" ${1}
             return 1;
     esac
 

scripts/githooks/commit_msg.py

@@ -131,7 +131,7 @@ def check_body_limit(self):
 
     def check_body_uses_why(self):
         "Rule 7: Use the body to explain what and why vs. how"
-        # Not enforcable
+        # Not enforceable
         return True
 
     rule_funcs = [

scripts/test_downgrade_from_tag.sh

@@ -142,7 +142,7 @@ wait_for_pg() {
         if docker_exec $1 "pg_isready -U postgres"
         then
             # this makes the test less flaky, although not
-            # ideal. Apperently, pg_isready is not always a good
+            # ideal. Apparently, pg_isready is not always a good
             # indication of whether the DB is actually ready to accept
             # queries
             sleep 1
@@ -187,7 +187,7 @@ docker_pgcmd ${CONTAINER_ORIG} "CHECKPOINT;"
 
 # We need the previous version shared libraries as well, so we copy
 # all shared libraries out from the original container before stopping
-# it. We could limit it to just the preceeding version, but this is
+# it. We could limit it to just the preceding version, but this is
 # more straightforward.
 srcdir=$(docker exec ${CONTAINER_ORIG} /bin/bash -c 'pg_config --pkglibdir')
 FILES=$(docker exec ${CONTAINER_ORIG} /bin/bash -c "ls $srcdir/timescaledb*.so")

scripts/test_repair_from_tag.sh

@@ -88,7 +88,7 @@ wait_for_pg() {
 
         if docker_exec $1 "pg_isready -h localhost -U postgres"; then
             # this makes the test less flaky, although not
-            # ideal. Apperently, pg_isready is not always a good
+            # ideal. Apparently, pg_isready is not always a good
             # indication of whether the DB is actually ready to accept
             # queries
             sleep 5

scripts/test_update_from_tag.sh

@@ -144,7 +144,7 @@ wait_for_pg() {
         if docker_exec $1 "pg_isready -U postgres"
         then
             # this makes the test less flaky, although not
-            # ideal. Apperently, pg_isready is not always a good
+            # ideal. Apparently, pg_isready is not always a good
             # indication of whether the DB is actually ready to accept
             # queries
             sleep 1

src/cache.c

@@ -50,7 +50,7 @@ ts_cache_init(Cache *cache)
 
 	/*
 	 * We always want to be explicit about the memory context our hash table
-	 * ends up in to ensure it's not accidently put in TopMemoryContext.
+	 * ends up in to ensure it's not accidentally put in TopMemoryContext.
 	 */
 	Assert(cache->flags & HASH_CONTEXT);
 	cache->htab = hash_create(cache->name, cache->numelements, &cache->hctl, cache->flags);

src/chunk.c

@@ -509,7 +509,7 @@ chunk_collides(const Hypertable *ht, const Hypercube *hc)
 }
 
 /*-
- * Resolve collisions and perform alignmment.
+ * Resolve collisions and perform alignment.
  *
  * Chunks collide only if their hypercubes overlap in all dimensions. For
  * instance, the 2D chunks below collide because they overlap in both the X and
@@ -2122,7 +2122,7 @@ ts_chunk_show_chunks(PG_FUNCTION_ARGS)
 		funcctx = SRF_FIRSTCALL_INIT();
 		/*
 		 * For INTEGER type dimensions, we support querying using intervals or any
-		 * timetamp or date input. For such INTEGER dimensions, we get the chunks
+		 * timestamp or date input. For such INTEGER dimensions, we get the chunks
 		 * using their creation time values.
 		 */
 		if (IS_INTEGER_TYPE(time_type) && (arg_type == INTERVALOID || IS_TIMESTAMP_TYPE(arg_type)))
@@ -3497,7 +3497,7 @@ ts_chunk_is_frozen(Chunk *chunk)
 }
 
 /* only caller used to be ts_chunk_unset_frozen. This code was in PG14 block as we run into
- * defined but unsed error in CI/CD builds for PG < 14. But now called from recompress as well
+ * a "defined but unset" error in CI/CD builds for PG < 14. But now called from recompress as well
  */
 bool
 ts_chunk_clear_status(Chunk *chunk, int32 status)
@@ -3857,7 +3857,7 @@ ts_chunk_do_drop_chunks(Hypertable *ht, int64 older_than, int64 newer_than, int3
 	{
 		/*
 		 * For INTEGER type dimensions, we support querying using intervals or any
-		 * timetamp or date input. For such INTEGER dimensions, we get the chunks
+		 * timestamp or date input. For such INTEGER dimensions, we get the chunks
 		 * using their creation time values.
 		 */
 		if (IS_INTEGER_TYPE(time_type) && (arg_type == INTERVALOID || IS_TIMESTAMP_TYPE(arg_type)))
@@ -4206,7 +4206,7 @@ ts_chunk_drop_chunks(PG_FUNCTION_ARGS)
 
 	/*
 	 * For INTEGER type dimensions, we support querying using intervals or any
-	 * timetamp or date input. For such INTEGER dimensions, we get the chunks
+	 * timestamp or date input. For such INTEGER dimensions, we get the chunks
 	 * using their creation time values.
 	 */
 	if (IS_INTEGER_TYPE(time_type) && (arg_type == INTERVALOID || IS_TIMESTAMP_TYPE(arg_type)))
@@ -4645,7 +4645,7 @@ add_foreign_table_as_chunk(Oid relid, Hypertable *parent_ht)
 														   chunk->relkind,
 														   chunk->hypertable_relid);
 	chunk_create_table_constraints(parent_ht, chunk);
-	/* Add dimension constriants for the chunk */
+	/* Add dimension constraints for the chunk */
 	ts_chunk_constraints_add_dimension_constraints(chunk->constraints, chunk->fd.id, chunk->cube);
 	ts_chunk_constraints_insert_metadata(chunk->constraints);
 	chunk_add_inheritance(chunk, parent_ht);
@@ -4657,7 +4657,7 @@ add_foreign_table_as_chunk(Oid relid, Hypertable *parent_ht)
 	 * Noncontiguous flag should not be set since the chunk should be empty upon
 	 * creation, with an invalid range assigned, so ordered append should be allowed.
 	 * Once the data is moved into the OSM chunk, then our catalog should be
-	 * udpated with proper API calls from the OSM extension.
+	 * updated with proper API calls from the OSM extension.
 	 */
 	parent_ht->fd.status =
 		ts_set_flags_32(parent_ht->fd.status,

src/chunk_constraint.c

@@ -913,7 +913,7 @@ chunk_constraint_delete_metadata(TupleInfo *ti)
 
 		/*
 		 * If this is an index constraint, we need to cleanup the index
-		 * metadata. Don't drop the index though, since that will happend when
+		 * metadata. Don't drop the index though, since that will happen when
 		 * the constraint is dropped.
 		 */
 		if (OidIsValid(index_relid))

src/compat/compat.h

@@ -535,7 +535,7 @@ get_reindex_options(ReindexStmt *stmt)
 /*
  * PG15 added additional `force_flush` argument to shm_mq_send().
  *
- * Our _compat() version currently uses force_flush = true on PG15 to preseve
+ * Our _compat() version currently uses force_flush = true on PG15 to preserve
  * the same behaviour on all supported PostgreSQL versions.
  *
  * https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=46846433

src/constraint.h

@@ -11,7 +11,7 @@
 #include "export.h"
 
 /*
- * Return status for constraint processsing function.
+ * Return status for constraint processing function.
  *
  * PROCESSED - count the constraint as processed
  * IGNORED - the constraint wasn't processed

src/hypertable.c

@@ -1497,7 +1497,7 @@ ts_hypertable_create_internal(FunctionCallInfo fcinfo, Oid table_relid,
 		ts_cache_release(hcache);
 
 		/*
-		 * Validate create_hypertable arguments and use defaults accoring to the
+		 * Validate create_hypertable arguments and use defaults according to the
 		 * hypertable_distributed_default guc.
 		 *
 		 * Validate data nodes and check permissions on them if this is a

src/nodes/chunk_append/chunk_append.c

@@ -396,7 +396,7 @@ ts_chunk_append_path_create(PlannerInfo *root, RelOptInfo *rel, Hypertable *ht,
 		Assert(flat == NULL);
 
 		/*
-		 * if we do not have scans as direct childs of this
+		 * if we do not have scans as direct children of this
 		 * node we disable startup and runtime exclusion
 		 * in this node
 		 */

src/nodes/chunk_dispatch/chunk_insert_state.c

@@ -367,7 +367,7 @@ setup_on_conflict_state(ChunkInsertState *state, ChunkDispatch *dispatch,
 
 	/*
 	 * If the chunk's tuple descriptor matches exactly the hypertable
-	 * (the common case), we can re-use most of the parent's ON
+	 * (the common case), we can reuse most of the parent's ON
 	 * CONFLICT SET state, skipping a bunch of work.  Otherwise, we
 	 * need to create state specific to this partition.
 	 */

src/nodes/constraint_aware_append/constraint_aware_append.c

@@ -198,7 +198,7 @@ ca_append_begin(CustomScanState *node, EState *estate, int eflags)
 
 	/*
 	 * clauses should always have the same length as appendplans because
-	 * thats the base for building the lists
+	 * that's the base for building the lists
 	 */
 	Assert(list_length(old_appendplans) == list_length(chunk_ri_clauses));
 	Assert(list_length(chunk_relids) == list_length(chunk_ri_clauses));

src/partitioning.c

@@ -258,7 +258,7 @@ ts_partitioning_func_apply(PartitioningInfo *pinfo, Oid collation, Datum value)
  * Helper function to find the right partition value from a tuple,
  * for space partitioned hypertables. Since attributes in tuple can
  * be of different order when compared to physical table columns order,
- * we pass partition_col_idx which points to correct space parititioned
+ * we pass partition_col_idx which points to correct space partitioned
  * column in the given tuple.
  */
 TSDLLEXPORT Datum

src/planner/expand_hypertable.c

@@ -1319,7 +1319,7 @@ ts_plan_expand_hypertable_chunks(Hypertable *ht, PlannerInfo *root, RelOptInfo *
 #if PG16_LT
 		childrte->requiredPerms = 0;
 #else
-		/* Since PG16, the permission info is maintained separetely. Unlink
+		/* Since PG16, the permission info is maintained separately. Unlink
 		 * the old perminfo from the RTE to disable permission checking.
 		 */
 		childrte->perminfoindex = 0;

src/planner/planner.c

@@ -846,7 +846,7 @@ should_chunk_append(Hypertable *ht, PlannerInfo *root, RelOptInfo *rel, Path *pa
 				 * Even though ordered is true on the RelOptInfo we have to
 				 * double check that current Path fulfills requirements for
 				 * Ordered Append transformation because the RelOptInfo may
-				 * be used for multiple Pathes.
+				 * be used for multiple Paths.
 				 */
 				Expr *em_expr = find_em_expr_for_rel(pk->pk_eclass, rel);
 

src/process_utility.c

@@ -2447,7 +2447,7 @@ process_index_chunk_multitransaction(int32 hypertable_id, Oid chunk_relid, void
 	 * We grab a ShareLock on the chunk, because that's what CREATE INDEX
 	 * does. For the hypertable's index, we are ok using the weaker
 	 * AccessShareLock, since we only need to prevent the index itself from
-	 * being ALTERed or DROPed during this part of index creation.
+	 * being ALTERed or DROPped during this part of index creation.
 	 */
 	chunk_rel = table_open(chunk_relid, ShareLock);
 	chunk = ts_chunk_get_by_relid(chunk_relid, true);
@@ -2573,7 +2573,7 @@ process_index_start(ProcessUtilityArgs *args)
 				ts_cache_release(hcache);
 				ereport(ERROR,
 						(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
-						 errmsg("operation not supported on continuous aggreates that are not "
+						 errmsg("operation not supported on continuous aggregates that are not "
 								"finalized"),
 						 errhint("Recreate the continuous aggregate to allow index creation.")));
 			}
@@ -2722,7 +2722,7 @@ process_index_start(ProcessUtilityArgs *args)
 	 * Lock the index for the remainder of the command. Since we're using
 	 * multiple transactions for index creation, a regular
 	 * transaction-level lock won't prevent the index from being
-	 * concurrently ALTERed or DELETEed. Instead, we grab a session level
+	 * concurrently ALTERed or DELETEd. Instead, we grab a session level
 	 * lock on the index, which we'll release when the command is
 	 * finished. (This is the same strategy postgres uses in CREATE INDEX
 	 * CONCURRENTLY)
@@ -4423,7 +4423,7 @@ process_ddl_event_sql_drop(EventTriggerData *trigdata)
 TS_FUNCTION_INFO_V1(ts_timescaledb_process_ddl_event);
 
 /*
- * Event trigger hook for DDL commands that have alread been handled by
+ * Event trigger hook for DDL commands that have already been handled by
  * PostgreSQL (i.e., "ddl_command_end" and "sql_drop" events).
  */
 Datum

src/telemetry/functions.c

@@ -42,7 +42,7 @@ typedef struct AllowedFnHashEntry
 } AllowedFnHashEntry;
 
 // Get a HTAB of AllowedFnHashEntrys containing all and only those functions
-// that are withing visible_extensions. This function should be equivalent to
+// that are within visible_extensions. This function should be equivalent to
 // the SQL
 //     SELECT objid
 //     FROM pg_catalog.pg_depend, pg_catalog.pg_extension extension
@@ -222,7 +222,7 @@ ts_function_telemetry_read(const char **visible_extensions, int num_visible_exte
  * This function resets the shared function counts after we send back telemetry
  * in preparation for the next recording cycle. Note that there is no way to
  * atomically read and reset the counts in the shared hashmap, so writes that
- * occur between sending the old counts and reseting for the next cycle will be
+ * occur between sending the old counts and resetting for the next cycle will be
  * lost. Since this this telemetry is only ever an approximation of reality, we
  * believe this loss is acceptable considering that the alternatives are
  * resetting the counts whenever the telemetry is read (potentially even more

src/ts_catalog/continuous_aggs_watermark.c

@@ -250,7 +250,7 @@ TS_FUNCTION_INFO_V1(ts_continuous_agg_watermark_materialized);
  *
  * The difference between this function and `ts_continuous_agg_watermark` is
  * that this one get the max open dimension of the materialization hypertable
- * insted of get the stored value in the catalog table.
+ * instead of get the stored value in the catalog table.
  */
 Datum
 ts_continuous_agg_watermark_materialized(PG_FUNCTION_ARGS)

test/runner.sh

@@ -59,7 +59,7 @@ EOF
 trap cleanup EXIT
 
 # setup clusterwide settings on first run
-# we use mkdir here because it is an atomic operation unlike existance of a lockfile
+# we use mkdir here because it is an atomic operation unlike existence of a lockfile
 # where creating and checking are 2 separate operations
 if mkdir ${TEST_OUTPUT_DIR}/.pg_init 2>/dev/null; then
   ${PSQL} "$@" -U ${USER} -d template1 -v ECHO=none >/dev/null 2>&1 <<EOF

test/runner_shared.sh

@@ -36,7 +36,7 @@ TEST_ROLE_DEFAULT_PERM_USER_2=${TEST_ROLE_DEFAULT_PERM_USER_2:-default_perm_user
 shift
 
 # setup clusterwide settings on first run
-# we use mkdir here because it is an atomic operation unlike existance of a lockfile
+# we use mkdir here because it is an atomic operation unlike existence of a lockfile
 # where creating and checking are 2 separate operations
 if mkdir ${TEST_OUTPUT_DIR}/.pg_init 2>/dev/null; then
   ${PSQL} "$@" -U ${USER} -d postgres -v ECHO=none -c "ALTER USER ${TEST_ROLE_SUPERUSER} WITH SUPERUSER;" >/dev/null

test/src/bgw/log.c

@@ -47,7 +47,7 @@ bgw_log_insert_relation(Relation rel, char *msg)
 
 /* Insert a new entry into public.bgw_log
  * This table is used for testing as a way for mock background jobs
- * to insert messges into a log that could then be output into the golden file
+ * to insert messages into a log that could then be output into the golden file
  */
 static void
 bgw_log_insert(char *msg)

test/src/net/test_http.c

@@ -75,7 +75,7 @@ num_test_strings()
 	return sizeof(TEST_LENGTHS) / sizeof(TEST_LENGTHS[0]);
 }
 
-/*  Check we can succesfully parse partial by well-formed HTTP responses */
+/*  Check we can successfully parse partial by well-formed HTTP responses */
 Datum
 ts_test_http_parsing(PG_FUNCTION_ARGS)
 {