refactor: restructure API, deduplicate code, expand test coverage
Some checks failed
CI / build (push) Failing after 2m4s
CI / test (push) Failing after 2m6s
CI / vet (push) Failing after 2m19s

- Extract shared DeferClose helper, removing 14 duplicate copies
- Rename PlayWright-prefixed types to cleaner names (BrowserOptions,
  BrowserSelection, NewBrowser, etc.)
- Rename fields: ServerAddress, RequireServer (was DontLaunchOnConnectFailure)
- Extract shared initBrowser/mergeOptions into browser_init.go,
  deduplicating ~120 lines between NewBrowser and NewInteractiveBrowser
- Remove unused locator field from document struct
- Add tests for all previously untested packages (archive, aislegopher,
  wegmans, useragents, powerball) and expand existing test suites
- Add MIGRATION.md documenting all breaking API changes

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
2026-02-09 13:59:47 -05:00
parent e7b7e78796
commit cb2ed10cfd
32 changed files with 667 additions and 417 deletions

View File

@@ -4,7 +4,6 @@ import (
"context"
"errors"
"fmt"
"io"
"log/slog"
"net/url"
"strings"
@@ -39,12 +38,6 @@ func (c Config) validate() Config {
var DefaultConfig = Config{}
func deferClose(cl io.Closer) {
if cl != nil {
_ = cl.Close()
}
}
// IsArchived checks if a url is archived. It returns the archived url if it is archived, or an empty string if it is not.
func (c Config) IsArchived(ctx context.Context, b extractor.Browser, target string) (extractor.Document, error) {
c = c.validate()

View File

@@ -0,0 +1,37 @@
package archive
import (
"testing"
"time"
)
func TestConfig_Validate_Defaults(t *testing.T) {
c := Config{}
c = c.validate()
if c.Endpoint != "https://archive.ph" {
t.Errorf("Endpoint = %q, want %q", c.Endpoint, "https://archive.ph")
}
if c.Timeout == nil {
t.Fatal("Timeout should not be nil after validate")
}
if *c.Timeout != 1*time.Hour {
t.Errorf("Timeout = %v, want %v", *c.Timeout, 1*time.Hour)
}
}
func TestConfig_Validate_Preserves(t *testing.T) {
timeout := 5 * time.Minute
c := Config{
Endpoint: "https://archive.org",
Timeout: &timeout,
}
c = c.validate()
if c.Endpoint != "https://archive.org" {
t.Errorf("Endpoint = %q, want %q", c.Endpoint, "https://archive.org")
}
if *c.Timeout != 5*time.Minute {
t.Errorf("Timeout = %v, want %v", *c.Timeout, 5*time.Minute)
}
}