schemadiff: supporting textual diff

[shlomi-noach]

Description

In this PR schemadiff is able to generate textual diffs, nuanced to diff hints and specific logic of schemadiff.

Consider the following example. Two schemas are diffed, and the logical diff is:

DROP TABLE `t1`
ALTER TABLE `t2` MODIFY COLUMN `id` bigint
CREATE TABLE `t4` (
	`id` int,
	PRIMARY KEY (`id`)
)

On top of this logical diff, schemadiff will now also be able to generate this textual diff:

-CREATE TABLE `t1` (
-	`id` int,
-	PRIMARY KEY (`id`)
-)
 CREATE TABLE `t2` (
-	`id` int,
+	`id` bigint,
 	PRIMARY KEY (`id`)
 )
+CREATE TABLE `t4` (
+	`id` int,
+	PRIMARY KEY (`id`)
+)

The above is a simple example. What sets schemadiff apart is that the textual diff only shows what schemadiff thinks are ntoeworthy changes. For example, schemadiff ignores index ordering: it considers two tables to be identical if all but order of indexes is the same. The standard textual diff of such two tables will show the indexes as + and - annotated entries, but the schemdiff textual diff will show no changes. Likewise, based on diff hints, constraint names can be ignored, etc.

It also takes care of trailing last comma issues, so common with e.g. git diff. For example, in the below:

CREATE TABLE `t4` (
	`id` int,
	PRIMARY KEY (`id`)
)

CREATE TABLE `t4` (
	`id` int,
	PRIMARY KEY (`id`),
	KEY `another_idx` (`id`)
)

A standard textual diff will capture the PRIMARY KEY line, due to the addition of the trailing comma. schemadiff, however, outputs:

 CREATE TABLE `t4` (
 	`id` int,
 	PRIMARY KEY (`id`),
+	KEY `another_idx` (`id`)
 )

Implementation-wise, a DiffEntity now has an Annotated() function, which returns 3 annotated diff texts:

1+2: The split from & to annotated forms.
3: The unified form (as shown in the above example).

Related Issue(s)

Fixes #15387

Checklist

• "Backport to:" labels have been added if this change should be back-ported to release branches
• If this change is to be back-ported to previous releases, a justification is included in the PR description
• Tests were added or are not required
• Did the new or modified tests pass consistently locally and on CI?
• Documentation was added or is not required

Deployment Notes

[vitess-bot]

Review Checklist

Hello reviewers! 👋 Please follow this checklist when reviewing this Pull Request.

General

• Ensure that the Pull Request has a descriptive title.
• Ensure there is a link to an issue (except for internal cleanup and flaky test fixes), new features should have an RFC that documents use cases and test cases.

Tests

• Bug fixes should have at least one unit or end-to-end test, enhancement and new features should have a sufficient number of tests.

Documentation

• Apply the release notes (needs details) label if users need to know about this change.
• New features should be documented.
• There should be some code comments as to why things are implemented the way they are.
• There should be a comment at the top of each new or modified test to explain what the test does.

New flags

• Is this flag really necessary?
• Flag names must be clear and intuitive, use dashes (-), and have a clear help text.

If a workflow is added or modified:

• Each item in Jobs should be named in order to mark it as required.
• If the workflow needs to be marked as required, the maintainer team must be notified.

Backward compatibility

• Protobuf changes should be wire-compatible.
• Changes to _vt tables and RPCs need to be backward compatible.
• RPC changes should be compatible with vitess-operator
• If a flag is removed, then it should also be removed from vitess-operator and arewefastyet, if used there.
• vtctl command output order should be stable and awk-able.

[codecov]

Codecov Report

Attention: Patch coverage is 98.36957% with 3 lines in your changes are missing coverage. Please review.

Project coverage is 65.69%. Comparing base (d5bd597) to head (90d6627).

Files	Patch %	Lines
go/vt/schemadiff/annotations.go	97.56%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #15388      +/-   ##
==========================================
+ Coverage   65.64%   65.69%   +0.05%     
==========================================
  Files        1563     1564       +1     
  Lines      194389   194558     +169     
==========================================
+ Hits       127602   127816     +214     
+ Misses      66787    66742      -45     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[shlomi-noach]

Will probably revert to b23bb2dca65ea9de007290cd320ccbc16ac1e306 and undo the HTML changes, as they're too naive. They would for example mark a column name that happens to also be a keyword.

[shlomi-noach]

HTML change reverted via force push.

[shlomi-noach]

For this PR, I suggest not implementing HTML diff yet, until we have more clarity on how we might want to generate it (with/out syntax highlighting, full/per-line HTML, etc.).

[mattlord]

Nice! Only a few very minor comments about potential additional unit test cases. If they're relatively easy to add, probably worth it. I'll leave that up to you though.

[mattlord]

Any reason not to add a test case for this in the unit test?

[shlomi-noach]

Added.

[mattlord]

Seems like it's probably relatively easy to cover the default in the unit test too?

[shlomi-noach]

Added.

[mattlord]

Looks like this isn't covered by the unit test.

[shlomi-noach]

Added,