chore: do not strip type cast

[tianzhou]

Fix #91

[Copilot]

Pull Request Overview

This PR refactors the expression parsing and formatting to preserve type casts in PostgreSQL default values and expressions instead of stripping them. The key changes ensure PostgreSQL's canonical representation is maintained by using pg_query.Deparse for expression handling.

• Replaced manual expression parsing with pg_query.Deparse to preserve type casts and handle edge cases
• Updated formatting to preserve canonical PostgreSQL representation with type casts
• Added normalization to handle case differences between inspector output and parser output

Reviewed Changes

Copilot reviewed 31 out of 31 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
ir/parser.go	Replaced manual expression parsing with pg_query.Deparse-based approach for better accuracy
ir/normalize.go	Added keyword normalization to handle case differences between PostgreSQL functions
ir/formatter.go	Removed type cast stripping logic to preserve canonical representation
internal/diff/table.go	Removed stripTypeQualifiers function and type cast removal from column defaults
testdata/diff/migrate/v5/*	Updated test data to reflect lowercase PostgreSQL keywords
testdata/diff/migrate/v3/*	Updated test data to reflect lowercase PostgreSQL keywords
testdata/diff/dependency/*	Updated test data to reflect lowercase PostgreSQL keywords
testdata/diff/create_table/*	Updated test data to reflect lowercase PostgreSQL keywords
cmd/dump/issue_82_view_logic_expr_*	Test files showing preserved vs stripped type casts in view expressions

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

Pull Request Overview

Copilot reviewed 16 out of 16 changed files in this pull request and generated 2 comments.

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The simple string replacement approach can cause incorrect replacements when the keyword appears as part of another word. For example, 'current_user_id' would be incorrectly transformed to 'CURRENT_USER_id'. Use word boundary matching with regex or more precise string matching to avoid partial word replacements.

Suggested change

	// Use simple case-insensitive replacement
	// This works because we're only processing deparsed SQL from pg_query,
	// which formats keywords consistently
	lowercase := strings.ToLower(keyword)
	result = strings.ReplaceAll(result, lowercase, keyword)
	// Use regex word boundary replacement, case-insensitive
	// This avoids replacing keywords inside longer identifiers
	pattern := `(?i)\b` + regexp.QuoteMeta(keyword) + `\b`
	re := regexp.MustCompile(pattern)
	result = re.ReplaceAllString(result, keyword)

[Copilot]

[nitpick] Unnecessary blank line added. This doesn't add any value and should be removed to maintain consistent code formatting.

Suggested change

[Copilot]

Pull Request Overview

Copilot reviewed 18 out of 18 changed files in this pull request and generated 2 comments.

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The regex compilation inside the loop is inefficient. Consider pre-compiling the regex patterns or using a more efficient string replacement approach for better performance when processing multiple keywords.

[Copilot]

The regex is compiled on every function call. Consider moving this to a package-level variable or using sync.Once to compile it only once for better performance.