TMS Migration (to develop) #367
Conversation
imashkanov
changed the title
imashkanov
force-pushed
the
feature/EPMRPP-migrate-tms-to-develop
branch
from
6f8b9c8 to
e72cc3f
Compare
imashkanov
force-pushed
the
feature/EPMRPP-migrate-tms-to-develop
branch
from
de2a21c to
8576ded
Compare
imashkanov
force-pushed
the
feature/EPMRPP-migrate-tms-to-develop
branch
from
d9f46fb to
bae82c6
Compare
imashkanov
force-pushed
the
feature/EPMRPP-migrate-tms-to-develop
branch
from
202a17d to
600f7a4
Compare
imashkanov
force-pushed
the
feature/EPMRPP-migrate-tms-to-develop
branch
from
600f7a4 to
21fe65d
Compare
WalkthroughAdds a comprehensive Test Management System schema (types, tables, FKs, indexes, triggers, functions, attachments, versioning, and full‑text search) and provides a down migration that removes those objects and the added Changes
Sequence Diagram(s)sequenceDiagram
autonumber
actor Client
participant DB as Database
participant TR_plan as Trigger tms_test_plan_search_vector_trigger
participant FN_plan as Function update_tms_test_plan_search_vector()
participant TR_case as Trigger tms_test_case_search_vector_trigger
participant FN_case as Function update_tms_test_case_search_vector()
Note over DB: UP migration creates tables, enums, functions, triggers, indexes
Client->>DB: INSERT/UPDATE tms_test_plan (name, description)
DB->>TR_plan: BEFORE INSERT/UPDATE fires
TR_plan->>FN_plan: compute tsvector (name, description, attributes)
FN_plan-->>DB: set NEW.search_vector
DB-->>Client: row persisted (GIN index updated)
Client->>DB: INSERT/UPDATE tms_test_case (name, description)
DB->>TR_case: BEFORE INSERT/UPDATE fires
TR_case->>FN_case: compute tsvector (name, description, attributes)
FN_case-->>DB: set NEW.search_vector
DB-->>Client: row persisted (GIN index updated)
Note over DB: DOWN migration removes triggers/functions/indexes then drops TMS objects
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Poem
Pre-merge checks and finishing touches❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✨ Finishing touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Nitpick comments (10)
migrations/103_tms_initial.up.sql (10)
202-217: Index FK tms_test_case_version.test_case_id.Partial unique index on defaults won’t help general joins. Add a regular index.
CREATE UNIQUE INDEX idx_tms_test_case_version_default ON tms_test_case_version (test_case_id) WHERE is_default = true; + +CREATE INDEX idx_tms_test_case_version_test_case_id + ON tms_test_case_version (test_case_id);
254-263: Make step → steps_manual_scenario mandatory.A step without a parent scenario is likely invalid; enforce NOT NULL.
- steps_manual_scenario_id bigint + steps_manual_scenario_id bigint NOT NULL CONSTRAINT tms_step_fk_steps_manual_scenario REFERENCES tms_steps_manual_scenario
86-101: Add date integrity check for milestones.Prevent inverted date ranges.
end_date TIMESTAMP, type varchar(255), @@ test_plan_id bigint CONSTRAINT tms_milestone_fk_test_plan REFERENCES tms_test_plan ); +ALTER TABLE tms_milestone + ADD CONSTRAINT chk_tms_milestone_dates + CHECK (end_date IS NULL OR start_date IS NULL OR end_date >= start_date);
279-296: Guard attachment file_size and avoid truncation risk.
- Enforce non-negative size.
- path_to_file can exceed 255; prefer text.
- file_size bigint, - path_to_file varchar(255), + file_size bigint CHECK (file_size IS NULL OR file_size >= 0), + path_to_file text,
66-84: Index foreign keys on tms_test_plan.Typical join/delete paths will benefit from indexes.
); +CREATE INDEX idx_tms_test_plan_project_id ON tms_test_plan (project_id); +CREATE INDEX idx_tms_test_plan_environment_id ON tms_test_plan (environment_id); +CREATE INDEX idx_tms_test_plan_product_version_id ON tms_test_plan (product_version_id); +CREATE INDEX idx_tms_test_plan_launch_id ON tms_test_plan (launch_id);
52-64: Cover common hierarchy and M2M filters.
- Add index on test_folder.parent_id for tree navigation.
- Add secondary index on tms_environment_dataset(dataset_id) to complement the unique (environment_id, dataset_id).
CREATE INDEX idx_tms_test_folder_project_id ON tms_test_folder (project_id, id); +CREATE INDEX idx_tms_test_folder_parent_id ON tms_test_folder (parent_id); @@ CONSTRAINT tms_environment_dataset_unique UNIQUE (environment_id, dataset_id) ); + +CREATE INDEX idx_tms_environment_dataset_dataset_id + ON tms_environment_dataset (dataset_id);Also applies to: 118-118
168-186: Auto-maintain updated_at; consider generated search_vector.
- Currently updated_at won’t refresh on UPDATE.
- Optional: replace trigger-managed tsvector with a generated column for simpler maintenance (PG12+).
Option A (minimal): add timestamp trigger.
CREATE OR REPLACE FUNCTION set_updated_at() RETURNS TRIGGER AS $$ BEGIN NEW.updated_at := now(); RETURN NEW; END; $$ LANGUAGE plpgsql; CREATE TRIGGER tms_test_case_set_updated_at BEFORE UPDATE ON tms_test_case FOR EACH ROW EXECUTE FUNCTION set_updated_at();Option B (generated column, replaces trigger/function):
-- In table definition: -- search_vector tsvector GENERATED ALWAYS AS ( -- to_tsvector('simple', -- coalesce(name,'')||' '||coalesce(description,'')||' '||coalesce(priority,'')) -- ) STORED -- and drop trigger/function.Also applies to: 138-139
8-17: Project-level uniqueness on human-readable names (confirm).Do you require uniqueness for:
- product_version.version within project,
- dataset.name within project,
- environment.name within project?
If yes, add unique constraints accordingly.-- Examples: ALTER TABLE tms_product_version ADD CONSTRAINT uq_tms_product_version_proj_ver UNIQUE (project_id, version); ALTER TABLE tms_dataset ADD CONSTRAINT uq_tms_dataset_proj_name UNIQUE (project_id, name); ALTER TABLE tms_environment ADD CONSTRAINT uq_tms_environment_proj_name UNIQUE (project_id, name);Also applies to: 21-29, 42-50
218-233: One-to-one manual scenario shape (confirm).Model enforces a single manual_scenario per test_case_version and either TEXT or STEPS via sub-tables. If that’s intended, great; if you need to allow both types simultaneously, remove the UNIQUE on test_case_version_id and add a composite unique (test_case_version_id, type).
- test_case_version_id bigint - UNIQUE + test_case_version_id bigint CONSTRAINT tms_manual_scenario_fk_test_case_version REFERENCES tms_test_case_version,-- Alternative: -- ALTER TABLE tms_manual_scenario -- ADD CONSTRAINT uq_manual_scenario_one_per_type -- UNIQUE (test_case_version_id, type);Also applies to: 234-250, 254-264, 265-275
1-6: Attributes key length/index (confirm).If tms_attribute.key is frequently filtered by prefix/exact match, consider adding a dedicated index (btree/trgm as applicable). 255 may be tight for some metadata keys; TEXT is safer.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
migrations/103_tms_initial.down.sql(1 hunks)migrations/103_tms_initial.up.sql(1 hunks)
🔇 Additional comments (1)
migrations/103_tms_initial.down.sql (1)
1-50: Down looks consistent; remember to mirror any new indexes/constraints you add.The drop order and IF EXISTS usage are solid. If you adopt the suggested new indexes/constraints in the up migration, add corresponding DROP statements here to keep rollbacks clean.
| CREATE TABLE tms_dataset_data | ||
| ( | ||
| id BIGSERIAL | ||
| CONSTRAINT tms_dataset_data_pk PRIMARY KEY, | ||
| key varchar(255), | ||
| value varchar(255), | ||
| dataset_id bigint NOT NULL | ||
| CONSTRAINT tms_dataset_attribute_fk_tms_dataset | ||
| REFERENCES tms_dataset | ||
| ); |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Ensure dataset key uniqueness per dataset and fix FK name typo.
- Prevent duplicate keys within the same dataset.
- Current FK name looks copy-pasted (“attribute”); align with table naming.
- dataset_id bigint NOT NULL
- CONSTRAINT tms_dataset_attribute_fk_tms_dataset
- REFERENCES tms_dataset
+ dataset_id bigint NOT NULL
+ CONSTRAINT tms_dataset_data_fk_dataset
+ REFERENCES tms_dataset,
+ CONSTRAINT tms_dataset_data_unique_key_per_dataset UNIQUE (dataset_id, key)📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| CREATE TABLE tms_dataset_data | |
| ( | |
| id BIGSERIAL | |
| CONSTRAINT tms_dataset_data_pk PRIMARY KEY, | |
| key varchar(255), | |
| value varchar(255), | |
| dataset_id bigint NOT NULL | |
| CONSTRAINT tms_dataset_attribute_fk_tms_dataset | |
| REFERENCES tms_dataset | |
| ); | |
| CREATE TABLE tms_dataset_data | |
| ( | |
| id BIGSERIAL | |
| CONSTRAINT tms_dataset_data_pk PRIMARY KEY, | |
| key varchar(255), | |
| value varchar(255), | |
| dataset_id bigint NOT NULL | |
| CONSTRAINT tms_dataset_data_fk_dataset | |
| REFERENCES tms_dataset, | |
| CONSTRAINT tms_dataset_data_unique_key_per_dataset UNIQUE (dataset_id, key) | |
| ); |
🤖 Prompt for AI Agents
migrations/103_tms_initial.up.sql around lines 31 to 40: the tms_dataset_data
table needs a per-dataset uniqueness guarantee for key and the foreign key
constraint name should reflect this table (not “attribute”); update the CREATE
TABLE so that key is NOT NULL (to ensure meaningful uniqueness), add a UNIQUE
constraint on (dataset_id, key) to prevent duplicate keys within the same
dataset, and rename the FK constraint to something like
tms_dataset_data_fk_dataset referencing tms_dataset(id).
| id BIGSERIAL | ||
| CONSTRAINT tms_test_case_pk PRIMARY KEY, | ||
| created_at TIMESTAMP DEFAULT now() NOT NULL, | ||
| updated_at TIMESTAMP DEFAULT now() NOT NULL, | ||
| name varchar(255), | ||
| description varchar(255), | ||
| priority varchar(255), | ||
| search_vector tsvector, | ||
| external_id varchar(255), | ||
| test_folder_id bigint NOT NULL | ||
| CONSTRAINT tms_test_case_fk_test_folder | ||
| REFERENCES tms_test_folder, | ||
| dataset_id bigint | ||
| CONSTRAINT tms_test_case_fk_dataset | ||
| REFERENCES tms_dataset, | ||
| CONSTRAINT tms_test_case_name_folder_unique UNIQUE (name, test_folder_id) | ||
| ); |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Add index for tms_test_case lookups by folder.
Queries “all cases in folder” will not use the (name, test_folder_id) unique index efficiently. Add a single-column index.
CREATE INDEX idx_tms_test_plan_test_case_test_plan_id ON tms_test_plan_test_case (test_plan_id);
CREATE INDEX idx_tms_test_plan_test_case_test_case_id ON tms_test_plan_test_case (test_case_id);
+
+-- Speed up folder-level listings and deletes
+CREATE INDEX idx_tms_test_case_test_folder_id ON tms_test_case (test_folder_id);Also applies to: 165-167
🤖 Prompt for AI Agents
In migrations/103_tms_initial.up.sql around lines 136-152 (and also apply the
same change for the table defined at lines 165-167), the schema lacks a
single-column index on test_folder_id which prevents efficient “all cases in
folder” lookups; add a standalone B-tree index for test_folder_id on the
tms_test_case table (e.g. CREATE INDEX ... ON tms_test_case(test_folder_id)) and
also create the equivalent single-column index on the table defined at lines
165-167 so folder-based queries use the index instead of scanning the unique
(name, test_folder_id) constraint.
| CREATE TYPE LAUNCH_TYPE_ENUM AS ENUM ('AUTOMATION', 'MANUAL'); | ||
| ALTER TABLE launch | ||
| ADD COLUMN launch_type LAUNCH_TYPE_ENUM; | ||
| ALTER TABLE launch | ||
| ALTER COLUMN launch_type SET DEFAULT 'AUTOMATION'; | ||
| UPDATE launch | ||
| SET launch_type = 'AUTOMATION'; |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Make launch.launch_type NOT NULL after backfill.
Nulls would complicate downstream logic and analytics. You already backfill and set default; lock in integrity with NOT NULL.
ALTER TABLE launch
ALTER COLUMN launch_type SET DEFAULT 'AUTOMATION';
UPDATE launch
SET launch_type = 'AUTOMATION';
+ALTER TABLE launch
+ ALTER COLUMN launch_type SET NOT NULL;📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| CREATE TYPE LAUNCH_TYPE_ENUM AS ENUM ('AUTOMATION', 'MANUAL'); | |
| ALTER TABLE launch | |
| ADD COLUMN launch_type LAUNCH_TYPE_ENUM; | |
| ALTER TABLE launch | |
| ALTER COLUMN launch_type SET DEFAULT 'AUTOMATION'; | |
| UPDATE launch | |
| SET launch_type = 'AUTOMATION'; | |
| CREATE TYPE LAUNCH_TYPE_ENUM AS ENUM ('AUTOMATION', 'MANUAL'); | |
| ALTER TABLE launch | |
| ADD COLUMN launch_type LAUNCH_TYPE_ENUM; | |
| ALTER TABLE launch | |
| ALTER COLUMN launch_type SET DEFAULT 'AUTOMATION'; | |
| UPDATE launch | |
| SET launch_type = 'AUTOMATION'; | |
| ALTER TABLE launch | |
| ALTER COLUMN launch_type SET NOT NULL; |
🤖 Prompt for AI Agents
In migrations/103_tms_initial.up.sql around lines 334 to 340, the launch_type
column is created, defaulted and backfilled but not constrained to NOT NULL; add
an ALTER TABLE launch ALTER COLUMN launch_type SET NOT NULL statement after the
backfill and default are applied so the column cannot contain NULLs going
forward and data integrity is enforced.
imashkanov
force-pushed
the
feature/EPMRPP-migrate-tms-to-develop
branch
2 times, most recently
from
df2e346 to
7e1c575
Compare
…nts-to-migrate-tms-to-develop Add tms-attachments
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (6)
migrations/103_tms_initial.up.sql (6)
247-259: Remove redundant unique index; the column constraint already creates it.manual_scenario_id has UNIQUE at column-level. The separate UNIQUE INDEX duplicates it.
-CREATE UNIQUE INDEX idx_tms_manual_scenario_preconditions_scenario_unique - ON tms_manual_scenario_preconditions(manual_scenario_id);
137-146: Add ON DELETE CASCADE to junction/mapping FKs to prevent orphans and ease cleanup.These are pure link tables; cascading deletions are expected.
Examples:
--- tms_test_folder_test_item - REFERENCES tms_test_folder, + REFERENCES tms_test_folder ON DELETE CASCADE, - REFERENCES test_item, + REFERENCES test_item ON DELETE CASCADE, --- tms_test_plan_test_case - REFERENCES tms_test_plan, + REFERENCES tms_test_plan ON DELETE CASCADE, - REFERENCES tms_test_case, + REFERENCES tms_test_case ON DELETE CASCADE, --- tms_test_case_test_item - REFERENCES tms_test_case, + REFERENCES tms_test_case ON DELETE CASCADE, - REFERENCES test_item, + REFERENCES test_item ON DELETE CASCADE, --- tms_step_test_item - REFERENCES tms_step, + REFERENCES tms_step ON DELETE CASCADE, - REFERENCES test_item, + REFERENCES test_item ON DELETE CASCADE, --- tms_step_attachment - REFERENCES tms_step, + REFERENCES tms_step ON DELETE CASCADE, - REFERENCES tms_attachment, + REFERENCES tms_attachment ON DELETE CASCADE, --- tms_text_manual_scenario_attachment - REFERENCES tms_text_manual_scenario, + REFERENCES tms_text_manual_scenario ON DELETE CASCADE, - REFERENCES tms_attachment, + REFERENCES tms_attachment ON DELETE CASCADE, --- tms_manual_scenario_preconditions_attachment - REFERENCES tms_manual_scenario_preconditions, + REFERENCES tms_manual_scenario_preconditions ON DELETE CASCADE, - REFERENCES tms_attachment, + REFERENCES tms_attachment ON DELETE CASCADE,Please confirm intended deletion semantics match this.
Also applies to: 171-180, 202-211, 291-300, 323-333, 338-348, 353-363
66-85: Consider nullifying optional references on parent delete.environment_id, product_version_id, launch_id look optional. Prefer ON DELETE SET NULL to avoid delete conflicts.
- CONSTRAINT tms_test_plan_fk_environment - REFERENCES tms_environment, + CONSTRAINT tms_test_plan_fk_environment + REFERENCES tms_environment ON DELETE SET NULL, ... - CONSTRAINT tms_test_plan_fk_product_version - REFERENCES tms_product_version, + CONSTRAINT tms_test_plan_fk_product_version + REFERENCES tms_product_version ON DELETE SET NULL, ... - CONSTRAINT tms_test_plan_fk_launch - REFERENCES launch + CONSTRAINT tms_test_plan_fk_launch + REFERENCES launch ON DELETE SET NULL
155-169: Keep updated_at accurate.Add a trigger to set updated_at = now() on UPDATE.
CREATE OR REPLACE FUNCTION set_updated_at() RETURNS TRIGGER AS $$ BEGIN NEW.updated_at := now(); RETURN NEW; END; $$ LANGUAGE plpgsql; CREATE TRIGGER tms_test_case_set_updated_at BEFORE UPDATE ON tms_test_case FOR EACH ROW EXECUTE FUNCTION set_updated_at();
305-318: Harden attachment metadata.Default created_at, and enforce non-negative file_size.
- file_size bigint, + file_size bigint, path_to_file varchar(255) NOT NULL, - created_at TIMESTAMP, + created_at TIMESTAMP DEFAULT now() NOT NULL, expires_at TIMESTAMP, ... + CONSTRAINT tms_attachment_file_size_nonnegative + CHECK (file_size IS NULL OR file_size >= 0),
103-119: Resolve the “many-to-many?” TODOs or track them.Milestone ↔ product_version / test_plan likely need junction tables. Create tasks or adjust now.
I can propose junction table DDLs if you confirm the intended relationships.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
migrations/103_tms_initial.down.sql(1 hunks)migrations/103_tms_initial.up.sql(1 hunks)
🔇 Additional comments (3)
migrations/103_tms_initial.up.sql (3)
151-169: Add index for folder-level test case lookups.Single-column index on test_folder_id avoids scans when listing/deleting cases by folder.
-- existing table definition... ); +-- Speed up folder-level listings and deletes +CREATE INDEX idx_tms_test_case_test_folder_id ON tms_test_case (test_folder_id);
31-40: Ensure per-dataset key uniqueness and correct FK name.Make key NOT NULL, enforce (dataset_id, key) uniqueness, and fix FK constraint name.
CREATE TABLE tms_dataset_data ( id BIGSERIAL CONSTRAINT tms_dataset_data_pk PRIMARY KEY, - key varchar(255), + key varchar(255) NOT NULL, value varchar(255), dataset_id bigint NOT NULL - CONSTRAINT tms_dataset_attribute_fk_tms_dataset - REFERENCES tms_dataset + CONSTRAINT tms_dataset_data_fk_dataset + REFERENCES tms_dataset(id), + CONSTRAINT tms_dataset_data_unique_key_per_dataset UNIQUE (dataset_id, key) );
405-411: Make launch.launch_type NOT NULL after backfill.You set default and backfill; lock integrity with NOT NULL.
ALTER TABLE launch ALTER COLUMN launch_type SET DEFAULT 'AUTOMATION'; UPDATE launch SET launch_type = 'AUTOMATION'; +ALTER TABLE launch + ALTER COLUMN launch_type SET NOT NULL;
…hes-to-migrate-tms-to-develop Feature/epmrpp manual launches to migrate tms to develop
There was a problem hiding this comment.
Actionable comments posted: 0
♻️ Duplicate comments (3)
migrations/103_tms_initial.up.sql (2)
452-463: Add NOT NULL constraint to launch.launch_type after backfill.The column is backfilled to 'AUTOMATION' and has a default, but is not constrained to NOT NULL. Nulls would complicate downstream logic and analytics.
Apply this diff:
CREATE TYPE LAUNCH_TYPE_ENUM AS ENUM ('AUTOMATION', 'MANUAL'); ALTER TABLE launch ADD COLUMN launch_type LAUNCH_TYPE_ENUM; ALTER TABLE launch ALTER COLUMN launch_type SET DEFAULT 'AUTOMATION'; UPDATE launch SET launch_type = 'AUTOMATION'; +ALTER TABLE launch + ALTER COLUMN launch_type SET NOT NULL; ALTER TABLE launch ADD COLUMN test_plan_id bigint;
46-55: Fix tms_dataset_data: add NOT NULL, composite uniqueness, and correct FK name.The table lacks data integrity constraints:
keyshould be NOT NULL to ensure meaningful uniqueness- Missing
UNIQUE (dataset_id, key)to prevent duplicate keys per dataset- FK name
tms_dataset_attribute_fk_tms_datasetis misnamed (should reflect the table being defined)Apply this diff:
CREATE TABLE tms_dataset_data ( id BIGSERIAL CONSTRAINT tms_dataset_data_pk PRIMARY KEY, - key varchar(255), + key varchar(255) NOT NULL, value varchar(255), dataset_id bigint NOT NULL - CONSTRAINT tms_dataset_attribute_fk_tms_dataset - REFERENCES tms_dataset + CONSTRAINT tms_dataset_data_fk_dataset + REFERENCES tms_dataset, + CONSTRAINT tms_dataset_data_unique_key_per_dataset UNIQUE (dataset_id, key) );migrations/103_tms_initial.down.sql (1)
1-9: Remove duplicate/incorrect index drops at lines 1-2 and consolidate.Lines 1-2 attempt to drop indexes with names that don't match the up migration. Specifically, the up migration creates indexes named
idx_preconditions_attachment_*(line 391-392), but lines 1-2 try to dropidx_tms_manual_scenario_preconditions_attachment_*. TheIF EXISTSguards prevent errors, but the unnecessary drops add noise and confusion.Compare against up migration:
- Up:
CREATE INDEX idx_preconditions_attachment_preconditions_id ON tms_manual_scenario_preconditions_attachment(preconditions_id);(line 391)- Down line 1 tries:
DROP INDEX IF EXISTS idx_tms_manual_scenario_preconditions_attachment_attachment_id;❌- Down line 5 correctly:
DROP INDEX IF EXISTS idx_preconditions_attachment_attachment_id;✓Apply this diff to consolidate:
-DROP INDEX IF EXISTS idx_tms_manual_scenario_preconditions_attachment_attachment_id; -DROP INDEX IF EXISTS idx_tms_manual_scenario_preconditions_attachment_preconditions_id; DROP TABLE IF EXISTS tms_manual_scenario_preconditions_attachment; -DROP INDEX IF EXISTS idx_preconditions_attachment_attachment_id; -DROP INDEX IF EXISTS idx_preconditions_attachment_preconditions_id; DROP INDEX IF EXISTS idx_tms_text_manual_scenario_attachment_scenario_id; DROP INDEX IF EXISTS idx_tms_text_manual_scenario_attachment_attachment_id; DROP TABLE IF EXISTS tms_text_manual_scenario_attachment; + +DROP INDEX IF EXISTS idx_preconditions_attachment_preconditions_id; +DROP INDEX IF EXISTS idx_preconditions_attachment_attachment_id; DROP INDEX IF EXISTS idx_tms_step_attachment_attachment_id;
🧹 Nitpick comments (1)
migrations/103_tms_initial.up.sql (1)
163-181: Add single-column index on test_folder_id for efficient folder queries.Queries like "all test cases in folder X" will scan the composite unique index
(name, test_folder_id)inefficiently. Add a dedicated B-tree index on the folder ID.Add this index after the table definition:
CREATE TABLE tms_test_case ( id BIGSERIAL CONSTRAINT tms_test_case_pk PRIMARY KEY, created_at TIMESTAMP DEFAULT now() NOT NULL, updated_at TIMESTAMP DEFAULT now() NOT NULL, name varchar(255), description varchar(255), priority varchar(255), search_vector tsvector, external_id varchar(255), test_folder_id bigint NOT NULL CONSTRAINT tms_test_case_fk_test_folder REFERENCES tms_test_folder, dataset_id bigint CONSTRAINT tms_test_case_fk_dataset REFERENCES tms_dataset, CONSTRAINT tms_test_case_name_folder_unique UNIQUE (name, test_folder_id) ); + +CREATE INDEX idx_tms_test_case_test_folder_id ON tms_test_case (test_folder_id);
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
migrations/103_tms_initial.down.sql(1 hunks)migrations/103_tms_initial.up.sql(1 hunks)
2 participants
Summary by CodeRabbit