As its name, this tool is implemented mainly to compare SQL schemas.
node src/app.js first_path sec_pathMarkdown file is exported. Check ./reports folders for example.
We use this sql-ddl-to-json-schema lib as a parser, but it is not perfect. We have to clean the input before feeding it in
the parser. Here are some regexps that we use to ignore some sections of the input file
const nonDDLRegex = /(--.*(\n)?)|(\/\*.*(\n?))|(DELIMITER.*(\n.*)*\nDELIMITER.*(\n)?)|(CONSTRAINT.*CHECK.*\n)|(INSERT INTO.*(\n.*)*)/g;- Ignore comments
--
-- Table structure for table `corebanking_fx_requests`
--/*!40101 SET @saved_cs_client = @@character_set_client */;
/*!40101 SET character_set_client = utf8 */;- Ignore functions (usually start and end with
DELIMITER
DELIMITER ;;
CREATE FUNCTION `DATE_TRUNC`(...
END ;;
DELIMITER ;- Ignore constraint with check
CONSTRAINT `token_address_check` CHECK (regexp_like(`token_address`,_utf8mb4'^0x[0-9a-fA-F]{40}$')),
CONSTRAINT `wallet_address_check` CHECK (regexp_like(`wallet_address`,_utf8mb4'^0x[0-9a-fA-F]{40}$'))- Ignore data generation part
INSERT INTO `schema_migrations` (version) VALUES
('20120129231330'),
('20120705005644'),
('20120711033715'),
('20120711044839'),
('20120711054138'),
...const edgeCasesRegex = /((?<!((PRIMARY|FOREIGN) )|\S)KEY.*\n)|(UNIQUE KEY.*\n)/g;- Ignore normal keys and unique keys (keep primary and foreign keys). If we parse these keys, the parser will take forever. This is probably a tool bug.
PRIMARY KEY (`id`),
UNIQUE KEY `index_xfers_transitions_on_sort_key_and_xfers_transaction_id` (`sort_key`,`xfers_transaction_id`), -- ignore
UNIQUE KEY `index_xfers_transitions_parent_most_recent` (`xfers_transaction_id`,`most_recent`), -- ignore
KEY `index_xfers_transitions_on_xfers_transaction_id` (`xfers_transaction_id`) -- ignore
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent