Update license and readme.

.gitignore

@@ -1,10 +1,17 @@
-.idea
-.goroot
 **/.#*
 **/*~
+/.idea
+/.goroot
+/blacklist.txt
+/check.sh
+/debug.sh
+/run.sh
 /.go
 /cmd
 /db/initdb.sql
 /gemini-grc
 run*.sh
 /main
+/db/migration*/**
+/db/populate/**
+/db/sql/**

COPYING

@@ -1,14 +0,0 @@
-
- Copyright (c) Antanst
-
- Permission to use, copy, modify, and distribute this software for any
- purpose with or without fee is hereby granted, provided that the above
- copyright notice and this permission notice appear in all copies.
-
- THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
- WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
- MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
- ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
- WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
- ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
- OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

LICENSE

@@ -0,0 +1,15 @@
+ISC License
+
+Copyright (c) Antanst 2014-2015
+
+Permission to use, copy, modify, and distribute this software for any
+purpose with or without fee is hereby granted, provided that the above
+copyright notice and this permission notice appear in all copies.
+
+THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

Makefile

@@ -1,10 +1,10 @@
-SHELL := /usr/local/bin/oksh
+SHELL := /bin/env oksh
 export PATH := $(PATH)
 
-all: fmt lintfix test clean build
+all: fmt lintfix tidy test clean build
 
 clean:
-	rm ./main
+	rm -f ./gemini-grc
 
 debug:
 	@echo "PATH: $(PATH)"
@@ -16,7 +16,10 @@ debug:
 
 # Test
 test:
-	go test ./...
+	go test -race ./...
+
+tidy:
+	go mod tidy
 
 # Format code
 fmt:
@@ -32,4 +35,13 @@ lintfix: fmt
 	golangci-lint run --fix
 
 build:
-	go build ./main.go
+	go build -race -o gemini-grc ./main.go
+
+show-updates:
+	go list -m -u all
+
+update:
+	go get -u all
+
+update-patch:
+	go get -u=patch all

README.md

@@ -1,27 +1,83 @@
 # gemini-grc
 
-A crawler for the [Gemini](https://en.wikipedia.org/wiki/Gemini_(protocol)) network. Easily extendable as a "wayback machine" of Gemini.
+A crawler for the [Gemini](https://en.wikipedia.org/wiki/Gemini_(protocol)) network.
+Easily extendable as a "wayback machine" of Gemini.
 
-## Features done
+## Features
-- [x] URL normalization
-- [x] Handle redirects (3X status codes)
 - [x] Follow robots.txt, see gemini://geminiprotocol.net/docs/companion/robots.gmi
 - [x] Save image/* and text/* files
-- [x] Concurrent downloading with workers
+- [x] Concurrent downloading with configurable number of workers
 - [x] Connection limit per host
 - [x] URL Blacklist
 - [x] Configuration via environment variables
-- [x] Storing snapshots in PostgreSQL
+- [x] Storing capsule snapshots in PostgreSQL
 - [x] Proper response header & body UTF-8 and format validation
+- [x] Proper URL normalization
+- [x] Handle redirects (3X status codes)
+
+## How to run
+
+Spin up a PostgreSQL, check `db/sql/initdb.sql` to create the tables and start the crawler.
+All configuration is done via environment variables.
+
+## Configuration
+
+Bool can be `true`,`false` or `0`,`1`.
+
+```text
+	LogLevel               string // Logging level (debug, info, warn, error)
+	MaxResponseSize        int // Maximum size of response in bytes
+	NumOfWorkers           int // Number of concurrent workers
+	ResponseTimeout        int // Timeout for responses in seconds
+	WorkerBatchSize        int // Batch size for worker processing
+	PanicOnUnexpectedError bool // Panic on unexpected errors when visiting a URL
+	BlacklistPath          string // File that has blacklisted strings of "host:port"
+	DryRun                 bool // If false, don't write to disk
+	PrintWorkerStatus      bool // If false, print logs and not worker status table
+```
+
+Example:
+
+```shell
+LOG_LEVEL=info \
+NUM_OF_WORKERS=10 \
+WORKER_BATCH_SIZE=10 \
+BLACKLIST_PATH="./blacklist.txt" \ # one url per line, can be empty
+MAX_RESPONSE_SIZE=10485760 \
+RESPONSE_TIMEOUT=10 \
+PANIC_ON_UNEXPECTED_ERROR=true \
+PG_DATABASE=test \
+PG_HOST=127.0.0.1 \
+PG_MAX_OPEN_CONNECTIONS=100 \
+PG_PORT=5434 \
+PG_USER=test \
+PG_PASSWORD=test \
+DRY_RUN=false \
+./gemini-grc
+```
+
+## Development
+
+Install linters. Check the versions first.
+```shell
+go install mvdan.cc/gofumpt@v0.7.0
+go install github.com/golangci/golangci-lint/cmd/golangci-lint@v1.63.4
+```
 
 ## TODO
 - [ ] Add snapshot history
 - [ ] Add a web interface
 - [ ] Provide to servers a TLS cert for sites that require it, like Astrobotany
+- [ ] Use pledge/unveil in OpenBSD hosts
 
 ## TODO (lower priority)
 - [ ] Gopher
-- [ ] Scroll gemini://auragem.letz.dev/devlog/20240316.gmi
+- [ ] More? http://dbohdan.sdf.org/smolnet/
-- [ ] Spartan
+
-- [ ] Nex
+## Notes
-- [ ] SuperTXT https://supertxt.net/00-intro.html
+Good starting points:
+
+gemini://warmedal.se/~antenna/
+gemini://tlgs.one/
+gopher://i-logout.cz:70/1/bongusta/
+gopher://gopher.quux.org:70/

bin/gemget/main.go

@@ -0,0 +1,47 @@
+package main
+
+import (
+	"encoding/json"
+	"fmt"
+	"os"
+
+	"gemini-grc/common/snapshot"
+	_url "gemini-grc/common/url"
+	"gemini-grc/config"
+	"gemini-grc/errors"
+	"gemini-grc/gemini"
+	"gemini-grc/gopher"
+	"gemini-grc/logging"
+)
+
+func main() {
+	config.CONFIG = *config.GetConfig()
+	err := runApp()
+	if err != nil {
+		fmt.Printf("%v\n", err)
+		logging.LogError("%v", err)
+		os.Exit(1)
+	}
+}
+
+func runApp() error {
+	if len(os.Args) != 2 {
+		return errors.NewError(fmt.Errorf("missing URL to visit"))
+	}
+	url := os.Args[1]
+	var s *snapshot.Snapshot
+	var err error
+	if _url.IsGeminiUrl(url) {
+		s, err = gemini.Visit(url)
+	} else if _url.IsGopherURL(url) {
+		s, err = gopher.Visit(url)
+	} else {
+		return errors.NewFatalError(fmt.Errorf("not a Gemini or Gopher URL"))
+	}
+	if err != nil {
+		return err
+	}
+	_json, _ := json.MarshalIndent(s, "", "  ")
+	fmt.Printf("%s\n", _json)
+	return err
+}

bin/normalizeSnapshot/main.go

@@ -4,7 +4,9 @@ import (
 	"fmt"
 	"os"
 
-	"gemini-grc/gemini"
+	"gemini-grc/common/snapshot"
+	"gemini-grc/common/url"
+	main2 "gemini-grc/db"
 	_ "github.com/jackc/pgx/v5/stdlib" // PGX driver for PostgreSQL
 	"github.com/jmoiron/sqlx"
 )
@@ -21,7 +23,7 @@ func main() {
         ORDER BY id
         LIMIT 10000 OFFSET $1
     `
-		var snapshots []gemini.Snapshot
+		var snapshots []snapshot.Snapshot
 		err := tx.Select(&snapshots, query, count)
 		if err != nil {
 			printErrorAndExit(tx, err)
@@ -32,8 +34,8 @@ func main() {
 		}
 		for _, s := range snapshots {
 			count++
-			escaped := gemini.EscapeURL(s.URL.String())
+			escaped := url.EscapeURL(s.URL.String())
-			normalizedGeminiURL, err := gemini.ParseURL(escaped, "")
+			normalizedGeminiURL, err := url.ParseURL(escaped, "", true)
 			if err != nil {
 				fmt.Println(s.URL.String())
 				fmt.Println(escaped)
@@ -47,7 +49,7 @@ func main() {
 			}
 			// If a snapshot already exists with the normalized
 			// URL, delete the current snapshot and leave the other.
-			var ss []gemini.Snapshot
+			var ss []snapshot.Snapshot
 			err = tx.Select(&ss, "SELECT * FROM snapshots WHERE URL=$1", normalizedURLString)
 			if err != nil {
 				printErrorAndExit(tx, err)
@@ -69,7 +71,7 @@ func main() {
 			// Saves the snapshot with the normalized URL
 			tx.MustExec("DELETE FROM snapshots WHERE id=$1", s.ID)
 			s.URL = *normalizedGeminiURL
-			err = gemini.UpsertSnapshot(0, tx, &s)
+			err = main2.OverwriteSnapshot(tx, &s)
 			if err != nil {
 				printErrorAndExit(tx, err)
 			}

common/blackList/blacklist.go

@@ -0,0 +1,55 @@
+package blackList
+
+import (
+	"fmt"
+	"os"
+	"regexp"
+	"strings"
+
+	"gemini-grc/config"
+	"gemini-grc/errors"
+	"gemini-grc/logging"
+)
+
+var Blacklist []regexp.Regexp //nolint:gochecknoglobals
+
+func LoadBlacklist() error {
+	if config.CONFIG.BlacklistPath == "" {
+		return nil
+	}
+	if Blacklist == nil {
+		data, err := os.ReadFile(config.CONFIG.BlacklistPath)
+		if err != nil {
+			Blacklist = []regexp.Regexp{}
+			return errors.NewError(fmt.Errorf("could not load Blacklist file: %w", err))
+		}
+
+		lines := strings.Split(string(data), "\n")
+
+		for _, line := range lines {
+			if line == "" || strings.HasPrefix(line, "#") {
+				continue
+			}
+			regex, err := regexp.Compile(line)
+			if err != nil {
+				return errors.NewError(fmt.Errorf("could not compile Blacklist line %s: %w", line, err))
+			}
+			Blacklist = append(Blacklist, *regex)
+
+		}
+
+		if len(lines) > 0 {
+			logging.LogInfo("Loaded %d blacklist entries", len(Blacklist))
+		}
+	}
+	return nil
+}
+
+func IsBlacklisted(u string) bool {
+	for _, v := range Blacklist {
+		if v.MatchString(u) {
+			return true
+		}
+	}
+	return false
+}

common/blackList/blacklist_test.go

@@ -0,0 +1,295 @@
+package blackList
+
+import (
+	"os"
+	"regexp"
+	"testing"
+
+	"gemini-grc/config"
+)
+
+func TestIsBlacklisted(t *testing.T) {
+	// Save original blacklist to restore after test
+	originalBlacklist := Blacklist
+	defer func() {
+		Blacklist = originalBlacklist
+	}()
+
+	tests := []struct {
+		name     string
+		setup    func()
+		url      string
+		expected bool
+	}{
+		{
+			name: "empty blacklist",
+			setup: func() {
+				Blacklist = []regexp.Regexp{}
+			},
+			url:      "https://example.com",
+			expected: false,
+		},
+		{
+			name: "exact hostname match",
+			setup: func() {
+				regex, _ := regexp.Compile(`example\.com`)
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "example.com",
+			expected: true,
+		},
+		{
+			name: "hostname in URL match",
+			setup: func() {
+				regex, _ := regexp.Compile(`example\.com`)
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "https://example.com/path",
+			expected: true,
+		},
+		{
+			name: "partial hostname match",
+			setup: func() {
+				regex, _ := regexp.Compile(`example\.com`)
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "https://safe-example.com",
+			expected: true,
+		},
+		{
+			name: "full URL match",
+			setup: func() {
+				regex, _ := regexp.Compile(`https://example\.com/bad-path`)
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "https://example.com/bad-path",
+			expected: true,
+		},
+		{
+			name: "path match",
+			setup: func() {
+				regex, _ := regexp.Compile("/malicious-path")
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "https://example.com/malicious-path",
+			expected: true,
+		},
+		{
+			name: "subdomain match with word boundary",
+			setup: func() {
+				regex, _ := regexp.Compile(`bad\.example\.com`)
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "https://bad.example.com/path",
+			expected: true,
+		},
+		{
+			name: "multiple patterns, one match",
+			setup: func() {
+				regex1, _ := regexp.Compile(`badsite\.com`)
+				regex2, _ := regexp.Compile(`malicious\.org`)
+				regex3, _ := regexp.Compile(`example\.com/sensitive`)
+				Blacklist = []regexp.Regexp{*regex1, *regex2, *regex3}
+			},
+			url:      "https://example.com/sensitive/data",
+			expected: true,
+		},
+		{
+			name: "multiple patterns, no match",
+			setup: func() {
+				regex1, _ := regexp.Compile(`badsite\.com`)
+				regex2, _ := regexp.Compile(`malicious\.org`)
+				regex3, _ := regexp.Compile(`example\.com/sensitive`)
+				Blacklist = []regexp.Regexp{*regex1, *regex2, *regex3}
+			},
+			url:      "https://example.com/safe/data",
+			expected: false,
+		},
+		{
+			name: "pattern with wildcard",
+			setup: func() {
+				regex, _ := regexp.Compile(`.*\.evil\.com`)
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "https://subdomain.evil.com/path",
+			expected: true,
+		},
+		{
+			name: "pattern with special characters",
+			setup: func() {
+				regex, _ := regexp.Compile(`example\.com/path\?id=[0-9]+`)
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "https://example.com/path?id=12345",
+			expected: true,
+		},
+		{
+			name: "unicode character support",
+			setup: func() {
+				regex, _ := regexp.Compile(`example\.com/[\p{L}]+`)
+				Blacklist = []regexp.Regexp{*regex}
+			},
+			url:      "https://example.com/café",
+			expected: true,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			tt.setup()
+			result := IsBlacklisted(tt.url)
+			if result != tt.expected {
+				t.Errorf("IsBlacklisted(%q) = %v, want %v", tt.url, result, tt.expected)
+			}
+		})
+	}
+}
+
+func TestLoadBlacklist(t *testing.T) {
+	// Save original blacklist to restore after test
+	originalBlacklist := Blacklist
+	originalConfigPath := config.CONFIG.BlacklistPath
+	defer func() {
+		Blacklist = originalBlacklist
+		config.CONFIG.BlacklistPath = originalConfigPath
+	}()
+
+	// Create a temporary blacklist file for testing
+	tmpFile, err := os.CreateTemp("", "blacklist-*.txt")
+	if err != nil {
+		t.Fatalf("Failed to create temporary file: %v", err)
+	}
+	defer os.Remove(tmpFile.Name())
+
+	// Test cases for LoadBlacklist
+	tests := []struct {
+		name           string
+		blacklistLines []string
+		configPath     string
+		wantErr        bool
+		expectedLen    int
+	}{
+		{
+			name:           "empty path",
+			blacklistLines: []string{},
+			configPath:     "",
+			wantErr:        false,
+			expectedLen:    0,
+		},
+		{
+			name:           "valid blacklist with comments",
+			blacklistLines: []string{"example\\.com", "# This is a comment", "malicious\\.org"},
+			configPath:     tmpFile.Name(),
+			wantErr:        false,
+			expectedLen:    2,
+		},
+		{
+			name:           "invalid regex",
+			blacklistLines: []string{"example\\.com", "[invalid regex"},
+			configPath:     tmpFile.Name(),
+			wantErr:        true,
+			expectedLen:    0,
+		},
+		{
+			name:           "nonexistent file",
+			blacklistLines: []string{},
+			configPath:     "nonexistent-file.txt",
+			wantErr:        true,
+			expectedLen:    0,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			// Reset blacklist
+			Blacklist = nil
+
+			// Set config path
+			config.CONFIG.BlacklistPath = tt.configPath
+
+			// Write test data to file if needed
+			if tt.configPath == tmpFile.Name() {
+				content := ""
+				for _, line := range tt.blacklistLines {
+					content += line + "\n"
+				}
+				if err := os.WriteFile(tmpFile.Name(), []byte(content), 0o644); err != nil {
+					t.Fatalf("Failed to write to temporary file: %v", err)
+				}
+			}
+
+			// Call the function
+			err := LoadBlacklist()
+
+			// Check results
+			if (err != nil) != tt.wantErr {
+				t.Errorf("LoadBlacklist() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+
+			if !tt.wantErr && len(Blacklist) != tt.expectedLen {
+				t.Errorf("LoadBlacklist() loaded %d entries, want %d", len(Blacklist), tt.expectedLen)
+			}
+		})
+	}
+}
+
+// TestIsBlacklistedIntegration tests the integration between LoadBlacklist and IsBlacklisted
+func TestIsBlacklistedIntegration(t *testing.T) {
+	// Save original blacklist to restore after test
+	originalBlacklist := Blacklist
+	originalConfigPath := config.CONFIG.BlacklistPath
+	defer func() {
+		Blacklist = originalBlacklist
+		config.CONFIG.BlacklistPath = originalConfigPath
+	}()
+
+	// Create a temporary blacklist file for testing
+	tmpFile, err := os.CreateTemp("", "blacklist-*.txt")
+	if err != nil {
+		t.Fatalf("Failed to create temporary file: %v", err)
+	}
+	defer os.Remove(tmpFile.Name())
+
+	// Write test patterns to the blacklist file
+	blacklistContent := `# Test blacklist file
+example\.com
+malicious\.org
+/phishing
+.*\.evil\.com
+\w+@spam\.com
+`
+	if err := os.WriteFile(tmpFile.Name(), []byte(blacklistContent), 0o644); err != nil {
+		t.Fatalf("Failed to write to temporary file: %v", err)
+	}
+
+	// Set up the test
+	Blacklist = nil
+	config.CONFIG.BlacklistPath = tmpFile.Name()
+
+	// Load the blacklist
+	if err := LoadBlacklist(); err != nil {
+		t.Fatalf("LoadBlacklist() failed: %v", err)
+	}
+
+	// Test URLs against the loaded blacklist
+	tests := []struct {
+		url      string
+		expected bool
+	}{
+		{"https://example.com", true},
+		{"https://safe-site.com", false},
+		{"https://malicious.org/path", true},
+		{"https://example.org/phishing", true},
+		{"https://subdomain.evil.com", true},
+		{"https://safe-site.com/safe-path", false},
+		{"mailto:user@spam.com", true},
+	}
+
+	for _, tt := range tests {
+		result := IsBlacklisted(tt.url)
+		if result != tt.expected {
+			t.Errorf("IsBlacklisted(%q) = %v, want %v", tt.url, result, tt.expected)
+		}
+	}
+}

common/errors.go

@@ -1,106 +0,0 @@
-package common
-
-import (
-	"errors"
-	"fmt"
-)
-
-type GeminiError struct {
-	Msg    string
-	Code   int
-	Header string
-}
-
-func (e *GeminiError) Error() string {
-	return fmt.Sprintf("%s: %s", e.Msg, e.Header)
-}
-
-func NewErrGeminiStatusCode(code int, header string) error {
-	var msg string
-	switch {
-	case code >= 10 && code < 20:
-		msg = "needs input"
-	case code >= 30 && code < 40:
-		msg = "redirect"
-	case code >= 40 && code < 50:
-		msg = "bad request"
-	case code >= 50 && code < 60:
-		msg = "server error"
-	case code >= 60 && code < 70:
-		msg = "TLS error"
-	default:
-		msg = "unexpected Status code"
-	}
-	return &GeminiError{
-		Msg:    msg,
-		Code:   code,
-		Header: header,
-	}
-}
-
-var (
-	ErrGeminiRobotsParse      = errors.New("gemini robots.txt parse error")
-	ErrGeminiRobotsDisallowed = errors.New("gemini robots.txt disallowed")
-	ErrGeminiResponseHeader   = errors.New("gemini response header error")
-	ErrGeminiRedirect         = errors.New("gemini redirection error")
-	ErrGeminiLinkLineParse    = errors.New("gemini link line parse error")
-
-	ErrURLParse     = errors.New("URL parse error")
-	ErrURLNotGemini = errors.New("not a Gemini URL")
-	ErrURLDecode    = errors.New("URL decode error")
-	ErrUTF8Parse    = errors.New("UTF-8 parse error")
-	ErrTextParse    = errors.New("text parse error")
-
-	ErrBlacklistMatches = errors.New("url matches blacklist")
-
-	ErrNetwork                        = errors.New("network error")
-	ErrNetworkDNS                     = errors.New("network DNS error")
-	ErrNetworkTLS                     = errors.New("network TLS error")
-	ErrNetworkSetConnectionDeadline   = errors.New("network error - cannot set connection deadline")
-	ErrNetworkCannotWrite             = errors.New("network error - cannot write")
-	ErrNetworkResponseSizeExceededMax = errors.New("network error - response size exceeded maximum size")
-
-	ErrDatabase     = errors.New("database error")
-	ErrDatabaseScan = errors.New("database scan error")
-)
-
-// We could have used a map for speed, but
-// we would lose ability to check wrapped
-// errors via errors.Is().
-
-var errGemini *GeminiError
-
-var knownErrors = []error{ //nolint:gochecknoglobals
-	errGemini,
-	ErrGeminiLinkLineParse,
-	ErrGeminiRobotsParse,
-	ErrGeminiRobotsDisallowed,
-	ErrGeminiResponseHeader,
-	ErrGeminiRedirect,
-
-	ErrBlacklistMatches,
-
-	ErrURLParse,
-	ErrURLDecode,
-	ErrUTF8Parse,
-	ErrTextParse,
-
-	ErrNetwork,
-	ErrNetworkDNS,
-	ErrNetworkTLS,
-	ErrNetworkSetConnectionDeadline,
-	ErrNetworkCannotWrite,
-	ErrNetworkResponseSizeExceededMax,
-
-	ErrDatabase,
-	ErrDatabaseScan,
-}
-
-func IsKnownError(err error) bool {
-	for _, known := range knownErrors {
-		if errors.Is(err, known) {
-			return true
-		}
-	}
-	return errors.As(err, new(*GeminiError))
-}

common/errors/errors.go

@@ -0,0 +1,41 @@
+package errors
+
+import (
+	"fmt"
+
+	"gemini-grc/errors"
+)
+
+// HostError is an error encountered while
+// visiting a host, and should be recorded
+// to the snapshot.
+type HostError struct {
+	Err error
+}
+
+func (e *HostError) Error() string {
+	return e.Err.Error()
+}
+
+func (e *HostError) Unwrap() error {
+	return e.Err
+}
+
+func NewHostError(err error) error {
+	return &HostError{Err: err}
+}
+
+func IsHostError(err error) bool {
+	if err == nil {
+		return false
+	}
+	var asError *HostError
+	return errors.As(err, &asError)
+}
+
+// Sentinel errors used for their string message primarily.
+// Do not use them by themselves, to be embedded to HostError.
+var (
+	ErrBlacklistMatch = fmt.Errorf("black list match")
+	ErrRobotsMatch    = fmt.Errorf("robots match")
+)

common/errors/errors_test.go

@@ -0,0 +1,38 @@
+package errors_test
+
+import (
+	"errors"
+	"fmt"
+	"testing"
+
+	"gemini-grc/gemini"
+)
+
+func TestErrGemini(t *testing.T) {
+	t.Parallel()
+	err := gemini.NewGeminiError(50, "50 server error")
+	if !errors.As(err, new(*gemini.GeminiError)) {
+		t.Errorf("TestErrGemini fail")
+	}
+}
+
+func TestErrGeminiWrapped(t *testing.T) {
+	t.Parallel()
+	err := gemini.NewGeminiError(50, "50 server error")
+	errWrapped := fmt.Errorf("%w wrapped", err)
+	if !errors.As(errWrapped, new(*gemini.GeminiError)) {
+		t.Errorf("TestErrGeminiWrapped fail")
+	}
+}
+
+func TestIsGeminiError(t *testing.T) {
+	t.Parallel()
+	err1 := gemini.NewGeminiError(50, "50 server error")
+	if !gemini.IsGeminiError(err1) {
+		t.Errorf("TestGeminiError fail #1")
+	}
+	wrappedErr1 := fmt.Errorf("wrapped %w", err1)
+	if !gemini.IsGeminiError(wrappedErr1) {
+		t.Errorf("TestGeminiError fail #2")
+	}
+}

common/errors_test.go

@@ -1,25 +0,0 @@
-package common_test
-
-import (
-	"errors"
-	"fmt"
-	"gemini-grc/common"
-	"testing"
-)
-
-func TestErrGemini(t *testing.T) {
-	t.Parallel()
-	err := common.NewErrGeminiStatusCode(50, "50 server error")
-	if !errors.As(err, new(*common.GeminiError)) {
-		t.Errorf("TestErrGemini fail")
-	}
-}
-
-func TestErrGeminiWrapped(t *testing.T) {
-	t.Parallel()
-	err := common.NewErrGeminiStatusCode(50, "50 server error")
-	errWrapped := fmt.Errorf("%w wrapped", err)
-	if !errors.As(errWrapped, new(*common.GeminiError)) {
-		t.Errorf("TestErrGeminiWrapped fail")
-	}
-}

common/gemini_url_test.go

@@ -1,251 +0,0 @@
-package common_test
-
-import (
-	"gemini-grc/common"
-	"reflect"
-	"testing"
-)
-
-func TestParseURL(t *testing.T) {
-	t.Parallel()
-	input := "gemini://caolan.uk/cgi-bin/weather.py/wxfcs/3162"
-	parsed, err := common.ParseURL(input, "")
-	value, _ := parsed.Value()
-	if err != nil || !(value == "gemini://caolan.uk:1965/cgi-bin/weather.py/wxfcs/3162") {
-		t.Errorf("fail: %s", parsed)
-	}
-}
-
-func TestDeriveAbsoluteURL_abs_url_input(t *testing.T) {
-	t.Parallel()
-	currentURL := common.URL{
-		Protocol: "gemini",
-		Hostname: "smol.gr",
-		Port:     1965,
-		Path:     "/a/b",
-		Descr:    "Nothing",
-		Full:     "gemini://smol.gr:1965/a/b",
-	}
-	input := "gemini://a.b/c"
-	output, err := common.DeriveAbsoluteURL(currentURL, input)
-	if err != nil {
-		t.Errorf("fail: %v", err)
-	}
-	expected := &common.URL{
-		Protocol: "gemini",
-		Hostname: "a.b",
-		Port:     1965,
-		Path:     "/c",
-		Descr:    "",
-		Full:     "gemini://a.b:1965/c",
-	}
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestDeriveAbsoluteURL_abs_path_input(t *testing.T) {
-	t.Parallel()
-	currentURL := common.URL{
-		Protocol: "gemini",
-		Hostname: "smol.gr",
-		Port:     1965,
-		Path:     "/a/b",
-		Descr:    "Nothing",
-		Full:     "gemini://smol.gr:1965/a/b",
-	}
-	input := "/c"
-	output, err := common.DeriveAbsoluteURL(currentURL, input)
-	if err != nil {
-		t.Errorf("fail: %v", err)
-	}
-	expected := &common.URL{
-		Protocol: "gemini",
-		Hostname: "smol.gr",
-		Port:     1965,
-		Path:     "/c",
-		Descr:    "",
-		Full:     "gemini://smol.gr:1965/c",
-	}
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestDeriveAbsoluteURL_rel_path_input(t *testing.T) {
-	t.Parallel()
-	currentURL := common.URL{
-		Protocol: "gemini",
-		Hostname: "smol.gr",
-		Port:     1965,
-		Path:     "/a/b",
-		Descr:    "Nothing",
-		Full:     "gemini://smol.gr:1965/a/b",
-	}
-	input := "c/d"
-	output, err := common.DeriveAbsoluteURL(currentURL, input)
-	if err != nil {
-		t.Errorf("fail: %v", err)
-	}
-	expected := &common.URL{
-		Protocol: "gemini",
-		Hostname: "smol.gr",
-		Port:     1965,
-		Path:     "/a/b/c/d",
-		Descr:    "",
-		Full:     "gemini://smol.gr:1965/a/b/c/d",
-	}
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeURLSlash(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net/retro-computing/magazines/"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := input
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeURLNoSlash(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net/retro-computing/magazines"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := input
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeMultiSlash(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net/retro-computing/////////a///magazines"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := "gemini://uscoffings.net/retro-computing/a/magazines"
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeTrailingSlash(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net/"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := "gemini://uscoffings.net/"
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeNoTrailingSlash(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := "gemini://uscoffings.net"
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeTrailingSlashPath(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net/a/"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := "gemini://uscoffings.net/a/"
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeNoTrailingSlashPath(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net/a"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := "gemini://uscoffings.net/a"
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeDot(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net/retro-computing/./././////a///magazines"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := "gemini://uscoffings.net/retro-computing/a/magazines"
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizePort(t *testing.T) {
-	t.Parallel()
-	input := "gemini://uscoffings.net:1965/a"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := "gemini://uscoffings.net/a"
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}
-
-func TestNormalizeURL(t *testing.T) {
-	t.Parallel()
-	input := "gemini://chat.gemini.lehmann.cx:11965/"
-	normalized, _ := common.NormalizeURL(input)
-	output := normalized.String()
-	expected := "gemini://chat.gemini.lehmann.cx:11965/"
-	pass := reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-
-	input = "gemini://chat.gemini.lehmann.cx:11965/index?a=1&b=c"
-	normalized, _ = common.NormalizeURL(input)
-	output = normalized.String()
-	expected = "gemini://chat.gemini.lehmann.cx:11965/index?a=1&b=c"
-	pass = reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-
-	input = "gemini://chat.gemini.lehmann.cx:11965/index#1"
-	normalized, _ = common.NormalizeURL(input)
-	output = normalized.String()
-	expected = "gemini://chat.gemini.lehmann.cx:11965/index#1"
-	pass = reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-
-	input = "gemini://gemi.dev/cgi-bin/xkcd.cgi?1494"
-	normalized, _ = common.NormalizeURL(input)
-	output = normalized.String()
-	expected = "gemini://gemi.dev/cgi-bin/xkcd.cgi?1494"
-	pass = reflect.DeepEqual(output, expected)
-	if !pass {
-		t.Errorf("fail: %#v != %#v", output, expected)
-	}
-}

common/linkList/linkList.go

@@ -0,0 +1,27 @@
+package linkList
+
+import (
+	"database/sql/driver"
+	"encoding/json"
+	"fmt"
+
+	"gemini-grc/common/url"
+)
+
+type LinkList []url.URL
+
+func (l *LinkList) Value() (driver.Value, error) {
+	return json.Marshal(l)
+}
+
+func (l *LinkList) Scan(value interface{}) error {
+	if value == nil {
+		*l = nil
+		return nil
+	}
+	b, ok := value.([]byte) // Type assertion! Converts to []byte
+	if !ok {
+		return fmt.Errorf("failed to scan LinkList: expected []byte, got %T", value)
+	}
+	return json.Unmarshal(b, l)
+}

common/shared.go

@@ -0,0 +1,13 @@
+package common
+
+var (
+	StatusChan chan WorkerStatus
+	// ErrorsChan accepts errors from workers.
+	// In case of fatal error, gracefully
+	// exits the application.
+	ErrorsChan chan error
+)
+
+const VERSION string = "0.0.1"
+
+const CtxKeyLogger string = "CtxKeyLogger"

common/snapshot.go

@@ -1,56 +0,0 @@
-package common
-
-import (
-	"database/sql/driver"
-	"encoding/json"
-	"fmt"
-	"time"
-
-	"github.com/guregu/null/v5"
-)
-
-type LinkList []URL
-
-func (l *LinkList) Value() (driver.Value, error) {
-	return json.Marshal(l)
-}
-
-func (l *LinkList) Scan(value interface{}) error {
-	if value == nil {
-		*l = nil
-		return nil
-	}
-	b, ok := value.([]byte) // Type assertion! Converts to []byte
-	if !ok {
-		return fmt.Errorf("failed to scan LinkList: expected []byte, got %T", value)
-	}
-	return json.Unmarshal(b, l)
-}
-
-type Snapshot struct {
-	ID           int                  `db:"id" json:"id,omitempty"`
-	URL          URL                  `db:"url" json:"url,omitempty"`
-	Host         string               `db:"host" json:"host,omitempty"`
-	Timestamp    null.Time            `db:"timestamp" json:"timestamp,omitempty"`
-	MimeType     null.String          `db:"mimetype" json:"mimetype,omitempty"`
-	Data         null.Value[[]byte]   `db:"data" json:"data,omitempty"`       // For non text/gemini files.
-	GemText      null.String          `db:"gemtext" json:"gemtext,omitempty"` // For text/gemini files.
-	Header       null.String          `db:"header" json:"header,omitempty"`   // Response header.
-	Links        null.Value[LinkList] `db:"links" json:"links,omitempty"`
-	Lang         null.String          `db:"lang" json:"lang,omitempty"`
-	ResponseCode null.Int             `db:"response_code" json:"code,omitempty"` // Gemini response status code.
-	Error        null.String          `db:"error" json:"error,omitempty"`        // On network errors only
-}
-
-func SnapshotFromURL(u string) *Snapshot {
-	url, err := ParseURL(u, "")
-	if err != nil {
-		return nil
-	}
-	newSnapshot := Snapshot{
-		URL:       *url,
-		Host:      url.Hostname,
-		Timestamp: null.TimeFrom(time.Now()),
-	}
-	return &newSnapshot
-}

common/snapshot/snapshot.go

@@ -0,0 +1,38 @@
+package snapshot
+
+import (
+	"time"
+
+	"gemini-grc/common/linkList"
+	commonUrl "gemini-grc/common/url"
+	"gemini-grc/errors"
+	"github.com/guregu/null/v5"
+)
+
+type Snapshot struct {
+	ID           int                           `db:"ID" json:"ID,omitempty"`
+	URL          commonUrl.URL                 `db:"url" json:"url,omitempty"`
+	Host         string                        `db:"host" json:"host,omitempty"`
+	Timestamp    null.Time                     `db:"timestamp" json:"timestamp,omitempty"`
+	MimeType     null.String                   `db:"mimetype" json:"mimetype,omitempty"`
+	Data         null.Value[[]byte]            `db:"data" json:"data,omitempty"`       // For non text/gemini files.
+	GemText      null.String                   `db:"gemtext" json:"gemtext,omitempty"` // For text/gemini files.
+	Header       null.String                   `db:"header" json:"header,omitempty"`   // Response header.
+	Links        null.Value[linkList.LinkList] `db:"links" json:"links,omitempty"`
+	Lang         null.String                   `db:"lang" json:"lang,omitempty"`
+	ResponseCode null.Int                      `db:"response_code" json:"code,omitempty"` // Gemini response Status code.
+	Error        null.String                   `db:"error" json:"error,omitempty"`        // On network errors only
+}
+
+func SnapshotFromURL(u string, normalize bool) (*Snapshot, error) {
+	url, err := commonUrl.ParseURL(u, "", normalize)
+	if err != nil {
+		return nil, errors.NewError(err)
+	}
+	newSnapshot := Snapshot{
+		URL:       *url,
+		Host:      url.Hostname,
+		Timestamp: null.TimeFrom(time.Now()),
+	}
+	return &newSnapshot, nil
+}

common/url/url.go

@@ -1,12 +1,15 @@
-package common
+package url
 
 import (
 	"database/sql/driver"
 	"fmt"
 	"net/url"
 	"path"
+	"regexp"
 	"strconv"
 	"strings"
+
+	"gemini-grc/errors"
 )
 
 type URL struct {
@@ -26,11 +29,10 @@ func (u *URL) Scan(value interface{}) error {
 	}
 	b, ok := value.(string)
 	if !ok {
-		return fmt.Errorf("%w: expected string, got %T", ErrDatabaseScan, value)
+		return errors.NewFatalError(fmt.Errorf("database scan error: expected string, got %T", value))
 	}
-	parsedURL, err := ParseURLNoNormalize(b, "")
+	parsedURL, err := ParseURL(b, "", false)
 	if err != nil {
-		err = fmt.Errorf("%w: failed to scan GeminiUrl %s: %v", ErrDatabaseScan, b, err)
 		return err
 	}
 	*u = *parsedURL
@@ -42,8 +44,14 @@ func (u URL) String() string {
 }
 
 func (u URL) StringNoDefaultPort() string {
-	if u.Port == 1965 {
+	if IsGeminiUrl(u.String()) {
-		return fmt.Sprintf("%s://%s%s", u.Protocol, u.Hostname, u.Path)
+		if u.Port == 1965 {
+			return fmt.Sprintf("%s://%s%s", u.Protocol, u.Hostname, u.Path)
+		}
+	} else {
+		if u.Port == 70 {
+			return fmt.Sprintf("%s://%s%s", u.Protocol, u.Hostname, u.Path)
+		}
 	}
 	return u.Full
 }
@@ -55,54 +63,43 @@ func (u URL) Value() (driver.Value, error) {
 	return u.Full, nil
 }
 
-func ParseURLNoNormalize(input string, descr string) (*URL, error) {
+func IsGeminiUrl(url string) bool {
-	u, err := url.Parse(input)
+	return strings.HasPrefix(url, "gemini://")
-	if err != nil {
-		return nil, fmt.Errorf("%w: Input %s URL Parse Error: %w", ErrURLParse, input, err)
-	}
-	if u.Scheme != "gemini" {
-		return nil, fmt.Errorf("%w: URL scheme '%s' is not supported", ErrURLNotGemini, u.Scheme)
-	}
-	protocol := u.Scheme
-	hostname := u.Hostname()
-	strPort := u.Port()
-	urlPath := u.Path
-	if strPort == "" {
-		strPort = "1965"
-	}
-	port, err := strconv.Atoi(strPort)
-	if err != nil {
-		return nil, fmt.Errorf("%w: Input %s GeminiError %w", ErrURLParse, input, err)
-	}
-	full := fmt.Sprintf("%s://%s:%d%s", protocol, hostname, port, urlPath)
-	// full field should also contain query params and url fragments
-	if u.RawQuery != "" {
-		full += "?" + u.RawQuery
-	}
-	if u.Fragment != "" {
-		full += "#" + u.Fragment
-	}
-	return &URL{Protocol: protocol, Hostname: hostname, Port: port, Path: urlPath, Descr: descr, Full: full}, nil
 }
 
-func ParseURL(input string, descr string) (*URL, error) {
+func IsGopherURL(s string) bool {
-	u, err := NormalizeURL(input)
+	return strings.HasPrefix(s, "gopher://")
-	if err != nil {
+}
-		return nil, fmt.Errorf("%w: Input %s URL Parse Error: %w", ErrURLParse, input, err)
+
-	}
+func ParseURL(input string, descr string, normalize bool) (*URL, error) {
-	if u.Scheme != "gemini" {
+	var u *url.URL
-		return nil, fmt.Errorf("%w: URL scheme '%s' is not supported", ErrURLNotGemini, u.Scheme)
+	var err error
+	if normalize {
+		u, err = NormalizeURL(input)
+		if err != nil {
+			return nil, err
+		}
+	} else {
+		u, err = url.Parse(input)
+		if err != nil {
+			return nil, errors.NewError(fmt.Errorf("error parsing URL: %w: %s", err, input))
+		}
 	}
 	protocol := u.Scheme
 	hostname := u.Hostname()
 	strPort := u.Port()
+	// urlPath := u.EscapedPath()
 	urlPath := u.Path
 	if strPort == "" {
-		strPort = "1965"
+		if u.Scheme == "gemini" {
+			strPort = "1965" // default Gemini port
+		} else {
+			strPort = "70" // default Gopher port
+		}
 	}
 	port, err := strconv.Atoi(strPort)
 	if err != nil {
-		return nil, fmt.Errorf("%w: Input %s GeminiError %w", ErrURLParse, input, err)
+		return nil, errors.NewError(fmt.Errorf("error parsing URL: %w: %s", err, input))
 	}
 	full := fmt.Sprintf("%s://%s:%d%s", protocol, hostname, port, urlPath)
 	// full field should also contain query params and url fragments
@@ -121,7 +118,7 @@ func ParseURL(input string, descr string) (*URL, error) {
 func DeriveAbsoluteURL(currentURL URL, input string) (*URL, error) {
 	// If target URL is absolute, return just it
 	if strings.Contains(input, "://") {
-		return ParseURL(input, "")
+		return ParseURL(input, "", true)
 	}
 	// input is a relative path. Clean it and construct absolute.
 	var newPath string
@@ -134,10 +131,10 @@ func DeriveAbsoluteURL(currentURL URL, input string) (*URL, error) {
 		newPath = path.Join(currentURL.Path, "/", path.Clean(input))
 	}
 	strURL := fmt.Sprintf("%s://%s:%d%s", currentURL.Protocol, currentURL.Hostname, currentURL.Port, newPath)
-	return ParseURL(strURL, "")
+	return ParseURL(strURL, "", true)
 }
 
-// NormalizeURL takes a URL string and returns a normalized version.
+// NormalizeURL takes a URL string and returns a normalized version
 // Normalized meaning:
 // - Path normalization (removing redundant slashes, . and .. segments)
 // - Proper escaping of special characters
@@ -148,7 +145,13 @@ func NormalizeURL(rawURL string) (*url.URL, error) {
 	// Parse the URL
 	u, err := url.Parse(rawURL)
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", ErrURLParse, err)
+		return nil, errors.NewError(fmt.Errorf("error normalizing URL: %w: %s", err, rawURL))
+	}
+	if u.Scheme == "" {
+		return nil, errors.NewError(fmt.Errorf("error normalizing URL: No scheme: %s", rawURL))
+	}
+	if u.Host == "" {
+		return nil, errors.NewError(fmt.Errorf("error normalizing URL: No host: %s", rawURL))
 	}
 
 	// Convert scheme to lowercase
@@ -159,7 +162,7 @@ func NormalizeURL(rawURL string) (*url.URL, error) {
 		u.Host = strings.ToLower(u.Host)
 	}
 
-	// Remove default ports
+	// remove default ports
 	if u.Port() != "" {
 		switch {
 		case u.Scheme == "http" && u.Port() == "80":
@@ -168,6 +171,8 @@ func NormalizeURL(rawURL string) (*url.URL, error) {
 			u.Host = u.Hostname()
 		case u.Scheme == "gemini" && u.Port() == "1965":
 			u.Host = u.Hostname()
+		case u.Scheme == "gopher" && u.Port() == "70":
+			u.Host = u.Hostname()
 		}
 	}
 
@@ -176,7 +181,7 @@ func NormalizeURL(rawURL string) (*url.URL, error) {
 		// Check if there was a trailing slash before cleaning
 		hadTrailingSlash := strings.HasSuffix(u.Path, "/")
 
-		u.Path = path.Clean(u.Path)
+		u.Path = path.Clean(u.EscapedPath())
 		// If path was "/", path.Clean() will return "."
 		if u.Path == "." {
 			u.Path = "/"
@@ -186,20 +191,25 @@ func NormalizeURL(rawURL string) (*url.URL, error) {
 		}
 	}
 
-	// Properly escape the path
+	// Properly escape the path, but only for unescaped parts
-	// First split on '/' to avoid escaping them
 	parts := strings.Split(u.Path, "/")
 	for i, part := range parts {
-		parts[i] = url.PathEscape(part)
+		// Try to unescape to check if it's already escaped
+		unescaped, err := url.PathUnescape(part)
+		if err != nil || unescaped == part {
+			// Part is not escaped, so escape it
+			parts[i] = url.PathEscape(part)
+		}
+		// If already escaped, leave as is
 	}
 	u.Path = strings.Join(parts, "/")
 
-	// Remove trailing fragment if empty
+	// remove trailing fragment if empty
 	if u.Fragment == "" {
 		u.Fragment = ""
 	}
 
-	// Remove trailing query if empty
+	// remove trailing query if empty
 	if u.RawQuery == "" {
 		u.RawQuery = ""
 	}
@@ -212,7 +222,7 @@ func EscapeURL(input string) string {
 	if strings.Contains(input, "%") && !strings.Contains(input, "% ") {
 		return input
 	}
-	// Split URL into parts (protocol, host, path)
+	// Split URL into parts (protocol, host, p)
 	parts := strings.SplitN(input, "://", 2)
 	if len(parts) != 2 {
 		return input
@@ -226,18 +236,50 @@ func EscapeURL(input string) string {
 		return input
 	}
 
-	// Split host and path
+	// Split host and p
 	parts = strings.SplitN(remainder, "/", 2)
 	host := parts[0]
 	if len(parts) == 1 {
 		return protocol + "://" + host
 	}
 
-	path := parts[1]
-
 	// Escape the path portion
-	escapedPath := url.PathEscape(path)
+	escapedPath := url.PathEscape(parts[1])
 
 	// Reconstruct the URL
 	return protocol + "://" + host + "/" + escapedPath
 }
+
+// TrimTrailingPathSlash trims trailing slash and handles empty path
+func TrimTrailingPathSlash(path string) string {
+	// Handle empty path (e.g., "http://example.com" -> treat as root)
+	if path == "" {
+		return "/"
+	}
+
+	// Trim trailing slash while preserving root slash
+	path = strings.TrimSuffix(path, "/")
+	if path == "" { // This happens if path was just "/"
+		return "/"
+	}
+	return path
+}
+
+// ExtractRedirectTargetFromHeader returns the redirection
+// URL by parsing the header (or error message)
+func ExtractRedirectTargetFromHeader(currentURL URL, input string) (*URL, error) {
+	// \d+ - matches one or more digits
+	// \s+ - matches one or more whitespace
+	// ([^\r]+) - captures everything until it hits a \r (or end of string)
+	pattern := `\d+\s+([^\r]+)`
+	re := regexp.MustCompile(pattern)
+	matches := re.FindStringSubmatch(input)
+	if len(matches) < 2 {
+		return nil, errors.NewError(fmt.Errorf("error extracting redirect target from string %s", input))
+	}
+	newURL, err := DeriveAbsoluteURL(currentURL, matches[1])
+	if err != nil {
+		return nil, err
+	}
+	return newURL, nil
+}

common/url/url_test.go

@@ -0,0 +1,420 @@
+package url
+
+import (
+	"reflect"
+	"testing"
+)
+
+func TestURLOperations(t *testing.T) {
+	t.Parallel()
+
+	t.Run("ParseURL", func(t *testing.T) {
+		t.Parallel()
+		tests := []struct {
+			name     string
+			input    string
+			base     string
+			absolute bool
+			want     string
+			wantErr  bool
+		}{
+			{
+				name:     "parse CGI URL",
+				input:    "gemini://caolan.uk/cgi-bin/weather.py/wxfcs/3162",
+				base:     "",
+				absolute: true,
+				want:     "gemini://caolan.uk:1965/cgi-bin/weather.py/wxfcs/3162",
+			},
+		}
+
+		for _, tt := range tests {
+			tt := tt
+			t.Run(tt.name, func(t *testing.T) {
+				t.Parallel()
+				parsed, err := ParseURL(tt.input, tt.base, tt.absolute)
+				if (err != nil) != tt.wantErr {
+					t.Errorf("ParseURL() error = %v, wantErr %v", err, tt.wantErr)
+					return
+				}
+				if !tt.wantErr {
+					value, _ := parsed.Value()
+					if value != tt.want {
+						t.Errorf("ParseURL() = %v, want %v", value, tt.want)
+					}
+				}
+			})
+		}
+	})
+
+	t.Run("DeriveAbsoluteURL", func(t *testing.T) {
+		t.Parallel()
+
+		baseURL := URL{
+			Protocol: "gemini",
+			Hostname: "smol.gr",
+			Port:     1965,
+			Path:     "/a/b",
+			Descr:    "Nothing",
+			Full:     "gemini://smol.gr:1965/a/b",
+		}
+
+		tests := []struct {
+			name     string
+			current  URL
+			input    string
+			expected *URL
+		}{
+			{
+				name:    "absolute URL input",
+				current: baseURL,
+				input:   "gemini://a.b/c",
+				expected: &URL{
+					Protocol: "gemini",
+					Hostname: "a.b",
+					Port:     1965,
+					Path:     "/c",
+					Full:     "gemini://a.b:1965/c",
+				},
+			},
+			{
+				name:    "absolute path input",
+				current: baseURL,
+				input:   "/c",
+				expected: &URL{
+					Protocol: "gemini",
+					Hostname: "smol.gr",
+					Port:     1965,
+					Path:     "/c",
+					Full:     "gemini://smol.gr:1965/c",
+				},
+			},
+			{
+				name:    "relative path input",
+				current: baseURL,
+				input:   "c/d",
+				expected: &URL{
+					Protocol: "gemini",
+					Hostname: "smol.gr",
+					Port:     1965,
+					Path:     "/a/b/c/d",
+					Full:     "gemini://smol.gr:1965/a/b/c/d",
+				},
+			},
+		}
+
+		for _, tt := range tests {
+			tt := tt
+			t.Run(tt.name, func(t *testing.T) {
+				t.Parallel()
+				output, err := DeriveAbsoluteURL(tt.current, tt.input)
+				if err != nil {
+					t.Fatalf("unexpected error: %v", err)
+				}
+				if !reflect.DeepEqual(output, tt.expected) {
+					t.Errorf("got %#v, want %#v", output, tt.expected)
+				}
+			})
+		}
+	})
+
+	t.Run("NormalizeURL", func(t *testing.T) {
+		t.Parallel()
+
+		tests := []struct {
+			name     string
+			input    string
+			expected string
+		}{
+			{
+				name:     "with trailing slash",
+				input:    "gemini://uscoffings.net/retro-computing/magazines/",
+				expected: "gemini://uscoffings.net/retro-computing/magazines/",
+			},
+			{
+				name:     "without trailing slash",
+				input:    "gemini://uscoffings.net/retro-computing/magazines",
+				expected: "gemini://uscoffings.net/retro-computing/magazines",
+			},
+			{
+				name:     "multiple slashes",
+				input:    "gemini://uscoffings.net/retro-computing/////////a///magazines",
+				expected: "gemini://uscoffings.net/retro-computing/a/magazines",
+			},
+			{
+				name:     "root with trailing slash",
+				input:    "gemini://uscoffings.net/",
+				expected: "gemini://uscoffings.net/",
+			},
+			{
+				name:     "root without trailing slash",
+				input:    "gemini://uscoffings.net",
+				expected: "gemini://uscoffings.net",
+			},
+			{
+				name:     "path with trailing slash",
+				input:    "gemini://uscoffings.net/a/",
+				expected: "gemini://uscoffings.net/a/",
+			},
+			{
+				name:     "path without trailing slash",
+				input:    "gemini://uscoffings.net/a",
+				expected: "gemini://uscoffings.net/a",
+			},
+			{
+				name:     "with dot segments",
+				input:    "gemini://uscoffings.net/retro-computing/./././////a///magazines",
+				expected: "gemini://uscoffings.net/retro-computing/a/magazines",
+			},
+			{
+				name:     "with default port",
+				input:    "gemini://uscoffings.net:1965/a",
+				expected: "gemini://uscoffings.net/a",
+			},
+		}
+
+		for _, tt := range tests {
+			tt := tt
+			t.Run(tt.name, func(t *testing.T) {
+				t.Parallel()
+				normalized, err := NormalizeURL(tt.input)
+				if err != nil {
+					t.Fatalf("unexpected error: %v", err)
+				}
+				output := normalized.String()
+				if output != tt.expected {
+					t.Errorf("got %#v, want %#v", output, tt.expected)
+				}
+			})
+		}
+	})
+}
+
+func TestNormalizeURL(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name     string
+		input    string
+		expected string
+	}{
+		{
+			name:     "URL with non-default port",
+			input:    "gemini://chat.gemini.lehmann.cx:11965/",
+			expected: "gemini://chat.gemini.lehmann.cx:11965/",
+		},
+		{
+			name:     "URL with query parameters",
+			input:    "gemini://chat.gemini.lehmann.cx:11965/index?a=1&b=c",
+			expected: "gemini://chat.gemini.lehmann.cx:11965/index?a=1&b=c",
+		},
+		{
+			name:     "URL with fragment",
+			input:    "gemini://chat.gemini.lehmann.cx:11965/index#1",
+			expected: "gemini://chat.gemini.lehmann.cx:11965/index#1",
+		},
+		{
+			name:     "URL with CGI script and query",
+			input:    "gemini://gemi.dev/cgi-bin/xkcd.cgi?1494",
+			expected: "gemini://gemi.dev/cgi-bin/xkcd.cgi?1494",
+		},
+	}
+
+	for _, tt := range tests {
+		tt := tt // capture range variable for parallel testing
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+			normalized, err := NormalizeURL(tt.input)
+			if err != nil {
+				t.Fatalf("unexpected error: %v", err)
+			}
+			output := normalized.String()
+			if output != tt.expected {
+				t.Errorf("got %#v, want %#v", output, tt.expected)
+			}
+		})
+	}
+}
+
+func TestNormalizePath(t *testing.T) {
+	t.Parallel()
+	tests := []struct {
+		name     string
+		input    string // URL string to parse
+		expected string // Expected normalized path
+	}{
+		// Basic cases
+		{
+			name:     "empty_path",
+			input:    "http://example.com",
+			expected: "",
+		},
+		{
+			name:     "root_path",
+			input:    "http://example.com/",
+			expected: "/",
+		},
+		{
+			name:     "single_trailing_slash",
+			input:    "http://example.com/test/",
+			expected: "/test/",
+		},
+		{
+			name:     "no_trailing_slash",
+			input:    "http://example.com/test",
+			expected: "/test",
+		},
+
+		// Edge cases with slashes
+		{
+			name:     "multiple_trailing_slashes",
+			input:    "http://example.com/test//",
+			expected: "/test/",
+		},
+		{
+			name:     "multiple_consecutive_slashes",
+			input:    "http://example.com//test//",
+			expected: "/test/",
+		},
+		{
+			name:     "only_slashes",
+			input:    "http://example.com////",
+			expected: "/",
+		},
+
+		// Encoded characters
+		{
+			name:     "encoded_spaces",
+			input:    "http://example.com/foo%20bar/",
+			expected: "/foo%20bar/",
+		},
+		{
+			name:     "encoded_special_chars",
+			input:    "http://example.com/foo%2Fbar/",
+			expected: "/foo%2Fbar/",
+		},
+
+		// Query parameters and fragments
+		{
+			name:     "with_query_parameters",
+			input:    "http://example.com/path?query=param",
+			expected: "/path",
+		},
+		{
+			name:     "with_fragment",
+			input:    "http://example.com/path#fragment",
+			expected: "/path",
+		},
+		{
+			name:     "with_both_query_and_fragment",
+			input:    "http://example.com/path?query=param#fragment",
+			expected: "/path",
+		},
+
+		// Unicode paths
+		{
+			name:     "unicode_characters",
+			input:    "http://example.com/über/path/",
+			expected: "/%C3%BCber/path/",
+		},
+		{
+			name:     "unicode_encoded",
+			input:    "http://example.com/%C3%BCber/path/",
+			expected: "/%C3%BCber/path/",
+		},
+
+		// Weird but valid cases
+		{
+			name:     "dot_in_path",
+			input:    "http://example.com/./path/",
+			expected: "/path/",
+		},
+		{
+			name:     "double_dot_in_path",
+			input:    "http://example.com/../path/",
+			expected: "/path/",
+		},
+		{
+			name:     "mixed_case",
+			input:    "http://example.com/PaTh/",
+			expected: "/PaTh/",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+			u, err := ParseURL(tt.input, "", true)
+			if err != nil {
+				t.Fatalf("Failed to parse URL %q: %v", tt.input, err)
+			}
+
+			result := u.Path
+			if result != tt.expected {
+				t.Errorf("Input: %s\nExpected: %q\nGot: %q",
+					u.Path, tt.expected, result)
+			}
+		})
+	}
+}
+
+func TestExtractRedirectTargetFullURL(t *testing.T) {
+	t.Parallel()
+	currentURL, _ := ParseURL("gemini://smol.gr", "", true)
+	input := "redirect: 31 gemini://target.gr"
+	result, err := ExtractRedirectTargetFromHeader(*currentURL, input)
+	expected := "gemini://target.gr:1965"
+	if err != nil || (result.String() != expected) {
+		t.Errorf("fail: Expected %s got %s", expected, result)
+	}
+}
+
+func TestExtractRedirectTargetFullURLSlash(t *testing.T) {
+	t.Parallel()
+	currentURL, _ := ParseURL("gemini://smol.gr", "", true)
+	input := "redirect: 31 gemini://target.gr/"
+	result, err := ExtractRedirectTargetFromHeader(*currentURL, input)
+	expected := "gemini://target.gr:1965/"
+	if err != nil || (result.String() != expected) {
+		t.Errorf("fail: Expected %s got %s", expected, result)
+	}
+}
+
+func TestExtractRedirectTargetRelativeURL(t *testing.T) {
+	t.Parallel()
+	currentURL, _ := ParseURL("gemini://smol.gr", "", true)
+	input := "redirect: 31 /a/b"
+	result, err := ExtractRedirectTargetFromHeader(*currentURL, input)
+	if err != nil || (result.String() != "gemini://smol.gr:1965/a/b") {
+		t.Errorf("fail: %s", result)
+	}
+}
+
+func TestExtractRedirectTargetRelativeURL2(t *testing.T) {
+	t.Parallel()
+	currentURL, _ := ParseURL("gemini://nox.im:1965", "", true)
+	input := "redirect: 31 ./"
+	result, err := ExtractRedirectTargetFromHeader(*currentURL, input)
+	if err != nil || (result.String() != "gemini://nox.im:1965/") {
+		t.Errorf("fail: %s", result)
+	}
+}
+
+func TestExtractRedirectTargetRelativeURL3(t *testing.T) {
+	t.Parallel()
+	currentURL, _ := ParseURL("gemini://status.zvava.org:1965", "", true)
+	input := "redirect: 31 index.gmi"
+	result, err := ExtractRedirectTargetFromHeader(*currentURL, input)
+	if err != nil || (result.String() != "gemini://status.zvava.org:1965/index.gmi") {
+		t.Errorf("fail: %s", result)
+	}
+}
+
+func TestExtractRedirectTargetWrong(t *testing.T) {
+	t.Parallel()
+	currentURL, _ := ParseURL("gemini://smol.gr", "", true)
+	input := "redirect: 31"
+	result, err := ExtractRedirectTargetFromHeader(*currentURL, input)
+	if result != nil || err == nil {
+		t.Errorf("fail: result should be nil, err is %s", err)
+	}
+}

common/worker.go

@@ -0,0 +1,320 @@
+package common
+
+import (
+	"fmt"
+	"time"
+
+	"gemini-grc/common/blackList"
+	errors2 "gemini-grc/common/errors"
+	"gemini-grc/common/snapshot"
+	url2 "gemini-grc/common/url"
+	_db "gemini-grc/db"
+	"gemini-grc/errors"
+	"gemini-grc/gemini"
+	"gemini-grc/gopher"
+	"gemini-grc/hostPool"
+	"gemini-grc/logging"
+	"github.com/guregu/null/v5"
+	"github.com/jmoiron/sqlx"
+)
+
+func CrawlOneURL(db *sqlx.DB, url *string) error {
+	parsedURL, err := url2.ParseURL(*url, "", true)
+	if err != nil {
+		return err
+	}
+
+	if !url2.IsGeminiUrl(parsedURL.String()) && !url2.IsGopherURL(parsedURL.String()) {
+		return errors.NewError(fmt.Errorf("error parsing URL: not a Gemini or Gopher URL: %s", parsedURL.String()))
+	}
+
+	tx, err := db.Beginx()
+	if err != nil {
+		return errors.NewFatalError(err)
+	}
+
+	err = _db.InsertURL(tx, parsedURL.Full)
+	if err != nil {
+		return err
+	}
+
+	err = workOnUrl(0, tx, parsedURL.Full)
+	if err != nil {
+		return err
+	}
+
+	err = tx.Commit()
+	if err != nil {
+		//if _db.IsDeadlockError(err) {
+		//	logging.LogError("Deadlock detected. Rolling back")
+		//	time.Sleep(time.Duration(10) * time.Second)
+		//	err := tx.Rollback()
+		//	return errors.NewFatalError(err)
+		//}
+		return errors.NewFatalError(err)
+	}
+	logging.LogInfo("Done")
+	return nil
+}
+
+func SpawnWorkers(numOfWorkers int, db *sqlx.DB) {
+	logging.LogInfo("Spawning %d workers", numOfWorkers)
+	go PrintWorkerStatus(numOfWorkers, StatusChan)
+
+	for i := range numOfWorkers {
+		go func(i int) {
+			UpdateWorkerStatus(i, "Waiting to start")
+			// Jitter to avoid starting everything at the same time
+			time.Sleep(time.Duration(i+2) * time.Second)
+			for {
+				// TODO: Use cancellable context with tx, logger & worker ID.
+				// ctx := context.WithCancel()
+				// ctx = context.WithValue(ctx, common.CtxKeyLogger, &RequestLogger{r: r})
+				RunWorkerWithTx(i, db)
+			}
+		}(i)
+	}
+}
+
+func RunWorkerWithTx(workerID int, db *sqlx.DB) {
+	defer func() {
+		UpdateWorkerStatus(workerID, "Done")
+	}()
+
+	tx, err := db.Beginx()
+	if err != nil {
+		ErrorsChan <- err
+		return
+	}
+
+	err = runWorker(workerID, tx)
+	if err != nil {
+		// TODO: Rollback in this case?
+		ErrorsChan <- err
+		return
+	}
+
+	logging.LogDebug("[%3d] Committing transaction", workerID)
+	err = tx.Commit()
+	// On deadlock errors, rollback and return, otherwise panic.
+	if err != nil {
+		logging.LogError("[%3d] Failed to commit transaction: %w", workerID, err)
+		if _db.IsDeadlockError(err) {
+			logging.LogError("[%3d] Deadlock detected. Rolling back", workerID)
+			time.Sleep(time.Duration(10) * time.Second)
+			err := tx.Rollback()
+			if err != nil {
+				panic(fmt.Sprintf("[%3d] Failed to roll back transaction: %v", workerID, err))
+			}
+			return
+		}
+		panic(fmt.Sprintf("[%3d] Failed to commit transaction: %v", workerID, err))
+	}
+	logging.LogDebug("[%3d] Worker done!", workerID)
+}
+
+func runWorker(workerID int, tx *sqlx.Tx) error {
+	var urls []string
+	var err error
+
+	UpdateWorkerStatus(workerID, "Getting URLs from DB")
+	urls, err = _db.GetRandomUrls(tx)
+	// urls, err = _db.GetRandomUrlsWithBasePath(tx)
+	if err != nil {
+		return err
+	} else if len(urls) == 0 {
+		logging.LogInfo("[%3d] No URLs to visit, sleeping...", workerID)
+		UpdateWorkerStatus(workerID, "No URLs to visit, sleeping...")
+		time.Sleep(1 * time.Minute)
+		return nil
+	}
+
+	// Start visiting URLs.
+	total := len(urls)
+	for i, u := range urls {
+		logging.LogInfo("[%3d] Starting %d/%d %s", workerID, i+1, total, u)
+		UpdateWorkerStatus(workerID, fmt.Sprintf("Starting %d/%d %s", i+1, total, u))
+		err := workOnUrl(workerID, tx, u)
+		if err != nil {
+			return err
+		}
+		logging.LogDebug("[%3d] Done %d/%d.", workerID, i+1, total)
+		UpdateWorkerStatus(workerID, fmt.Sprintf("Done %d/%d %s", i+1, total, u))
+	}
+	return nil
+}
+
+// workOnUrl visits a URL and stores the result.
+// unexpected errors are returned.
+// expected errors are stored within the snapshot.
+func workOnUrl(workerID int, tx *sqlx.Tx, url string) (err error) {
+	s, err := snapshot.SnapshotFromURL(url, false)
+	if err != nil {
+		return err
+	}
+
+	isGemini := url2.IsGeminiUrl(s.URL.String())
+	isGopher := url2.IsGopherURL(s.URL.String())
+	if !isGemini && !isGopher {
+		return errors.NewError(fmt.Errorf("not a Gopher or Gemini URL: %s", s.URL.String()))
+	}
+
+	if blackList.IsBlacklisted(s.URL.String()) {
+		logging.LogInfo("[%3d] URL matches blacklist, ignoring", workerID)
+		s.Error = null.StringFrom(errors2.ErrBlacklistMatch.Error())
+		return saveSnapshotAndRemoveURL(tx, s)
+	}
+
+	if isGemini {
+		// If URL matches a robots.txt disallow line,
+		// add it as an error and remove url
+		robotMatch, err := gemini.RobotMatch(s.URL.String())
+		if err != nil {
+			// robotMatch returns only network errors!
+			// we stop because we don't want to hit
+			// the server with another request on this case.
+			return err
+		}
+		if robotMatch {
+			logging.LogInfo("[%3d] URL matches robots.txt, ignoring", workerID)
+			s.Error = null.StringFrom(errors2.ErrRobotsMatch.Error())
+			return saveSnapshotAndRemoveURL(tx, s)
+		}
+	}
+
+	logging.LogDebug("[%3d] Adding to pool %s", workerID, s.URL.String())
+	UpdateWorkerStatus(workerID, fmt.Sprintf("Adding to pool %s", s.URL.String()))
+	hostPool.AddHostToHostPool(s.Host)
+	defer func(s string) {
+		hostPool.RemoveHostFromPool(s)
+	}(s.Host)
+
+	logging.LogDebug("[%3d] Visiting %s", workerID, s.URL.String())
+	UpdateWorkerStatus(workerID, fmt.Sprintf("Visiting %s", s.URL.String()))
+
+	if isGopher {
+		s, err = gopher.Visit(s.URL.String())
+	} else {
+		s, err = gemini.Visit(s.URL.String())
+	}
+
+	if err != nil {
+		return err
+	}
+
+	// Handle Gemini redirection.
+	if isGemini &&
+		s.ResponseCode.ValueOrZero() >= 30 &&
+		s.ResponseCode.ValueOrZero() < 40 {
+		err = handleRedirection(workerID, tx, s)
+		if err != nil {
+			return fmt.Errorf("error while handling redirection: %s", err)
+		}
+	}
+
+	// Store links
+	if len(s.Links.ValueOrZero()) > 0 {
+		logging.LogDebug("[%3d] Found %d links", workerID, len(s.Links.ValueOrZero()))
+		err = storeLinks(tx, s)
+		if err != nil {
+			return err
+		}
+	}
+
+	logging.LogInfo("[%3d] %2d %s", workerID, s.ResponseCode.ValueOrZero(), s.URL.String())
+	return saveSnapshotAndRemoveURL(tx, s)
+}
+
+func storeLinks(tx *sqlx.Tx, s *snapshot.Snapshot) error {
+	if s.Links.Valid { //nolint:nestif
+		for _, link := range s.Links.ValueOrZero() {
+			if shouldPersistURL(&link) {
+				visited, err := haveWeVisitedURL(tx, link.Full)
+				if err != nil {
+					return err
+				}
+				if !visited {
+					err := _db.InsertURL(tx, link.Full)
+					if err != nil {
+						return err
+					}
+				} else {
+					logging.LogDebug("Link already persisted: %s", link.Full)
+				}
+			}
+		}
+	}
+	return nil
+}
+
+func saveSnapshotAndRemoveURL(tx *sqlx.Tx, s *snapshot.Snapshot) error {
+	err := _db.OverwriteSnapshot(tx, s)
+	if err != nil {
+		return err
+	}
+	err = _db.DeleteURL(tx, s.URL.String())
+	if err != nil {
+		return err
+	}
+	return nil
+}
+
+// shouldPersistURL returns true if we
+// should save the URL in the _db.
+// Only gemini:// urls are saved.
+func shouldPersistURL(u *url2.URL) bool {
+	return url2.IsGeminiUrl(u.String()) || url2.IsGopherURL(u.String())
+}
+
+func haveWeVisitedURL(tx *sqlx.Tx, u string) (bool, error) {
+	var result []bool
+	err := tx.Select(&result, `SELECT TRUE FROM urls WHERE url=$1`, u)
+	if err != nil {
+		return false, errors.NewFatalError(fmt.Errorf("database error: %w", err))
+	}
+	if len(result) > 0 {
+		return result[0], nil
+	}
+	err = tx.Select(&result, `SELECT TRUE FROM snapshots WHERE snapshots.url=$1`, u)
+	if err != nil {
+		return false, errors.NewFatalError(fmt.Errorf("database error: %w", err))
+	}
+	if len(result) > 0 {
+		return result[0], nil
+	}
+	return false, nil
+}
+
+// handleRedirection saves redirection URL.
+func handleRedirection(workerID int, tx *sqlx.Tx, s *snapshot.Snapshot) error {
+	newURL, err := url2.ExtractRedirectTargetFromHeader(s.URL, s.Error.ValueOrZero())
+	if err != nil {
+		return err
+	}
+	logging.LogDebug("[%3d] Page redirects to %s", workerID, newURL)
+
+	haveWeVisited, _ := haveWeVisitedURL(tx, newURL.String())
+	if shouldPersistURL(newURL) && !haveWeVisited {
+		err = _db.InsertURL(tx, newURL.Full)
+		if err != nil {
+			return err
+		}
+		logging.LogDebug("[%3d] Saved redirection URL %s", workerID, newURL.String())
+	}
+	return nil
+}
+
+func GetSnapshotFromURL(tx *sqlx.Tx, url string) ([]snapshot.Snapshot, error) {
+	query := `
+	SELECT *
+	FROM snapshots
+	WHERE url=$1
+	LIMIT 1
+	`
+	var snapshots []snapshot.Snapshot
+	err := tx.Select(&snapshots, query, url)
+	if err != nil {
+		return nil, err
+	}
+	return snapshots, nil
+}

common/workerStatus.go

@@ -1,19 +1,35 @@
-package gemini
+package common
 
 import (
 	"fmt"
 	"strings"
+
+	"gemini-grc/config"
 )
 
 type WorkerStatus struct {
-	id     int
+	ID     int
-	status string
+	Status string
 }
 
-var statusChan chan WorkerStatus
+func UpdateWorkerStatus(workerID int, status string) {
+	if !config.GetConfig().PrintWorkerStatus {
+		return
+	}
+	if config.CONFIG.NumOfWorkers > 1 {
+		StatusChan <- WorkerStatus{
+			ID:     workerID,
+			Status: status,
+		}
+	}
+}
 
 func PrintWorkerStatus(totalWorkers int, statusChan chan WorkerStatus) {
-	// Create a slice to store current status of each worker
+	if !config.GetConfig().PrintWorkerStatus {
+		return
+	}
+
+	// Create a slice to store current Status of each worker
 	statuses := make([]string, totalWorkers)
 
 	// Initialize empty statuses
@@ -32,14 +48,14 @@ func PrintWorkerStatus(totalWorkers int, statusChan chan WorkerStatus) {
 	}
 	fmt.Print(output.String())
 
-	// Continuously receive status updates
+	// Continuously receive Status updates
 	for update := range statusChan {
-		if update.id >= totalWorkers {
+		if update.ID >= totalWorkers {
 			continue
 		}
 
-		// Update the status
+		// Update the Status
-		statuses[update.id] = update.status
+		statuses[update.ID] = update.Status
 
 		// Build the complete output string
 		output.Reset()
@@ -48,7 +64,7 @@ func PrintWorkerStatus(totalWorkers int, statusChan chan WorkerStatus) {
 			output.WriteString(fmt.Sprintf("[%2d] %.100s\n", i, status))
 		}
 
-		// Print the entire status
+		// Print the entire Status
 		fmt.Print(output.String())
 	}
 }

db/db.go

@@ -2,20 +2,22 @@ package db
 
 import (
 	"encoding/json"
-	"errors"
 	"fmt"
-	"gemini-grc/common"
 	"os"
 	"strconv"
+	"time"
 
+	"gemini-grc/common/snapshot"
+	commonUrl "gemini-grc/common/url"
 	"gemini-grc/config"
+	"gemini-grc/errors"
 	"gemini-grc/logging"
 	_ "github.com/jackc/pgx/v5/stdlib" // PGX driver for PostgreSQL
 	"github.com/jmoiron/sqlx"
 	"github.com/lib/pq"
 )
 
-func ConnectToDB() *sqlx.DB {
+func ConnectToDB() (*sqlx.DB, error) {
 	connStr := fmt.Sprintf("postgres://%s:%s@%s:%s/%s", //nolint:nosprintfhostport
 		os.Getenv("PG_USER"),
 		os.Getenv("PG_PASSWORD"),
@@ -27,25 +29,26 @@ func ConnectToDB() *sqlx.DB {
 	// Create a connection pool
 	db, err := sqlx.Open("pgx", connStr)
 	if err != nil {
-		panic(fmt.Sprintf("Unable to connect to database with URL %s: %v\n", connStr, err))
+		return nil, errors.NewFatalError(fmt.Errorf("unable to connect to database with URL %s: %w", connStr, err))
 	}
 	// TODO move PG_MAX_OPEN_CONNECTIONS to config env variables
 	maxConnections, err := strconv.Atoi(os.Getenv("PG_MAX_OPEN_CONNECTIONS"))
 	if err != nil {
-		panic(fmt.Sprintf("Unable to set max DB connections: %s\n", err))
+		return nil, errors.NewFatalError(fmt.Errorf("unable to set DB max connections: %w", err))
 	}
 	db.SetMaxOpenConns(maxConnections)
 	err = db.Ping()
 	if err != nil {
-		panic(fmt.Sprintf("Unable to ping database: %v\n", err))
+		return nil, errors.NewFatalError(fmt.Errorf("unable to ping database: %w", err))
 	}
 
 	logging.LogDebug("Connected to database")
-	return db
+	return db, nil
 }
 
-// IsDeadlockError checks if the error is a PostgreSQL deadlock error
+// IsDeadlockError checks if the error is a PostgreSQL deadlock error.
 func IsDeadlockError(err error) bool {
+	err = errors.Unwrap(err)
 	var pqErr *pq.Error
 	if errors.As(err, &pqErr) {
 		return pqErr.Code == "40P01" // PostgreSQL deadlock error code
@@ -53,134 +56,85 @@ func IsDeadlockError(err error) bool {
 	return false
 }
 
-func GetURLsToVisit(tx *sqlx.Tx) ([]string, error) {
+func GetRandomUrls(tx *sqlx.Tx) ([]string, error) {
 	var urls []string
-	err := tx.Select(&urls, SQL_SELECT_RANDOM_URLS_UNIQUE_HOSTS, config.CONFIG.WorkerBatchSize)
+	err := tx.Select(&urls, SQL_SELECT_RANDOM_URLS, config.CONFIG.WorkerBatchSize)
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", common.ErrDatabase, err)
+		return nil, errors.NewFatalError(err)
+	}
+	return urls, nil
+}
+
+func GetRandomUrlsWithBasePath(tx *sqlx.Tx) ([]string, error) {
+	SqlQuery := `SELECT url FROM snapshots WHERE url ~ '^[^:]+://[^/]+/?$' ORDER BY RANDOM() LIMIT $1`
+	var urls []string
+	err := tx.Select(&urls, SqlQuery, config.CONFIG.WorkerBatchSize)
+	if err != nil {
+		return nil, errors.NewFatalError(err)
 	}
 	return urls, nil
 }
 
 func InsertURL(tx *sqlx.Tx, url string) error {
+	logging.LogDebug("Inserting URL %s", url)
 	query := SQL_INSERT_URL
-	_, err := tx.NamedExec(query, url)
+	normalizedURL, err := commonUrl.ParseURL(url, "", true)
 	if err != nil {
-		return fmt.Errorf("%w inserting URL: %w", common.ErrDatabase, err)
+		return err
+	}
+	a := struct {
+		Url       string
+		Host      string
+		Timestamp time.Time
+	}{
+		Url:       normalizedURL.Full,
+		Host:      normalizedURL.Hostname,
+		Timestamp: time.Now(),
+	}
+	_, err = tx.NamedExec(query, a)
+	if err != nil {
+		return errors.NewFatalError(fmt.Errorf("cannot insert URL: database error %w URL %s", err, url))
 	}
 	return nil
 }
 
-func SaveSnapshotIfNew(tx *sqlx.Tx, s *common.Snapshot) error {
+func DeleteURL(tx *sqlx.Tx, url string) error {
+	logging.LogDebug("Deleting URL %s", url)
+	query := SQL_DELETE_URL
+	_, err := tx.Exec(query, url)
+	if err != nil {
+		return errors.NewFatalError(fmt.Errorf("cannot delete URL: database error %w URL %s", err, url))
+	}
+	return nil
+}
+
+func OverwriteSnapshot(tx *sqlx.Tx, s *snapshot.Snapshot) (err error) {
 	if config.CONFIG.DryRun {
 		marshalled, err := json.MarshalIndent(s, "", "  ")
 		if err != nil {
-			panic(fmt.Sprintf("JSON serialization error for %v", s))
+			return errors.NewFatalError(fmt.Errorf("JSON serialization error for %v", s))
 		}
-		logging.LogDebug("Would insert (if new) snapshot %s", marshalled)
+		logging.LogDebug("Would upsert snapshot %s", marshalled)
 		return nil
 	}
-	query := SQL_INSERT_SNAPSHOT_IF_NEW
-	_, err := tx.NamedExec(query, s)
-	if err != nil {
-		return fmt.Errorf("[%s] GeminiError inserting snapshot: %w", s.URL, err)
-	}
-	return nil
-}
-
-func OverwriteSnapshot(workedID int, tx *sqlx.Tx, s *common.Snapshot) (err error) {
-	//	if config.CONFIG.DryRun {
-	//marshalled, err := json.MarshalIndent(s, "", "  ")
-	//if err != nil {
-	//	panic(fmt.Sprintf("JSON serialization error for %v", s))
-	//}
-	//logging.LogDebug("[%d] Would upsert snapshot %s", workedID, marshalled)
-	//		return nil
-	//	}
 	query := SQL_UPSERT_SNAPSHOT
 	rows, err := tx.NamedQuery(query, s)
 	if err != nil {
-		return fmt.Errorf("[%d] %w while upserting snapshot: %w", workedID, common.ErrDatabase, err)
+		return errors.NewFatalError(fmt.Errorf("cannot overwrite snapshot: %w", err))
 	}
 	defer func() {
 		_err := rows.Close()
-		if _err != nil {
+		if err == nil && _err != nil {
-			err = fmt.Errorf("[%d] %w error closing rows: %w", workedID, common.ErrDatabase, _err)
+			err = errors.NewFatalError(fmt.Errorf("cannot overwrite snapshot: error closing rows: %w", err))
 		}
 	}()
 	if rows.Next() {
 		var returnedID int
 		err = rows.Scan(&returnedID)
 		if err != nil {
-			return fmt.Errorf("[%d] %w error scanning returned id: %w", workedID, common.ErrDatabase, err)
+			return errors.NewFatalError(fmt.Errorf("cannot overwrite snapshot: error scanning rows: %w", err))
 		}
 		s.ID = returnedID
-		// logging.LogDebug("[%d] Upserted snapshot with ID %d", workedID, returnedID)
-	}
-	return nil
-}
-
-func UpdateSnapshot(workedID int, tx *sqlx.Tx, s *common.Snapshot) (err error) {
-	//	if config.CONFIG.DryRun {
-	//marshalled, err := json.MarshalIndent(s, "", "  ")
-	//if err != nil {
-	//	panic(fmt.Sprintf("JSON serialization error for %v", s))
-	//}
-	//logging.LogDebug("[%d] Would upsert snapshot %s", workedID, marshalled)
-	//		return nil
-	//	}
-	query := SQL_UPDATE_SNAPSHOT
-	rows, err := tx.NamedQuery(query, s)
-	if err != nil {
-		return fmt.Errorf("[%d] %w while updating snapshot: %w", workedID, common.ErrDatabase, err)
-	}
-	defer func() {
-		_err := rows.Close()
-		if _err != nil {
-			err = fmt.Errorf("[%d] %w error closing rows: %w", workedID, common.ErrDatabase, _err)
-		}
-	}()
-	if rows.Next() {
-		var returnedID int
-		err = rows.Scan(&returnedID)
-		if err != nil {
-			return fmt.Errorf("[%d] %w error scanning returned id: %w", workedID, common.ErrDatabase, err)
-		}
-		s.ID = returnedID
-		// logging.LogDebug("[%d] Updated snapshot with ID %d", workedID, returnedID)
-	}
-	return nil
-}
-
-func SaveLinksToDBinBatches(tx *sqlx.Tx, snapshots []*common.Snapshot) error {
-	if config.CONFIG.DryRun {
-		return nil
-	}
-	const batchSize = 5000
-	query := SQL_INSERT_SNAPSHOT_IF_NEW
-	for i := 0; i < len(snapshots); i += batchSize {
-		end := i + batchSize
-		if end > len(snapshots) {
-			end = len(snapshots)
-		}
-		batch := snapshots[i:end]
-		_, err := tx.NamedExec(query, batch)
-		if err != nil {
-			return fmt.Errorf("%w: While saving links in batches: %w", common.ErrDatabase, err)
-		}
-	}
-	return nil
-}
-
-func SaveLinksToDB(tx *sqlx.Tx, snapshots []*common.Snapshot) error {
-	if config.CONFIG.DryRun {
-		return nil
-	}
-	query := SQL_INSERT_SNAPSHOT_IF_NEW
-	_, err := tx.NamedExec(query, snapshots)
-	if err != nil {
-		logging.LogError("GeminiError batch inserting snapshots: %w", err)
-		return fmt.Errorf("DB error: %w", err)
 	}
 	return nil
 }

db/db_queries.go

@@ -1,53 +1,24 @@
 package db
 
 const (
-	SQL_SELECT_RANDOM_UNVISITED_SNAPSHOTS = `
-SELECT *
-FROM snapshots
-WHERE response_code IS NULL
-  AND error IS NULL
-ORDER BY RANDOM()
-FOR UPDATE SKIP LOCKED
-LIMIT $1
-	`
 	SQL_SELECT_RANDOM_URLS_UNIQUE_HOSTS = `
 SELECT url
 FROM urls u
 WHERE u.id IN (
-      SELECT MIN(id)
+    SELECT id FROM (
-      FROM urls
+        SELECT id, ROW_NUMBER() OVER (PARTITION BY host ORDER BY id) as rn
-      GROUP BY host
+        FROM urls
+    ) t 
+    WHERE rn <= 3
 )
-LIMIT $1
-`
-	SQL_SELECT_RANDOM_UNVISITED_SNAPSHOTS_UNIQUE_HOSTS = `
-SELECT *
-FROM snapshots s
-WHERE response_code IS NULL
-  AND error IS NULL
-  AND s.id IN (
-      SELECT MIN(id)
-      FROM snapshots
-      WHERE response_code IS NULL 
-        AND error IS NULL
-      GROUP BY host
-  )
 ORDER BY RANDOM()
 FOR UPDATE SKIP LOCKED
 LIMIT $1
-`
+	`
-	SQL_SELECT_UNVISITED_SNAPSHOTS_UNIQUE_HOSTS = `
+	SQL_SELECT_RANDOM_URLS = `
-SELECT *
+SELECT url
-FROM snapshots s
+FROM urls u
-WHERE response_code IS NULL
+ORDER BY RANDOM()
-  AND error IS NULL
-  AND s.id IN (
-      SELECT MIN(id)
-      FROM snapshots
-      WHERE response_code IS NULL
-        AND error IS NULL
-      GROUP BY host
-  )
 FOR UPDATE SKIP LOCKED
 LIMIT $1
 `
@@ -90,4 +61,7 @@ RETURNING id
         VALUES (:url, :host, :timestamp)
         ON CONFLICT (url) DO NOTHING
     `
+	SQL_DELETE_URL = `
+        DELETE FROM urls WHERE url=$1
+    `
 )

errors/errors.go

@@ -0,0 +1,104 @@
+package errors
+
+import (
+	"errors"
+	"fmt"
+	"runtime"
+	"strings"
+)
+
+type fatal interface {
+	Fatal() bool
+}
+
+func IsFatal(err error) bool {
+	var e fatal
+	ok := As(err, &e)
+	return ok && e.Fatal()
+}
+
+func As(err error, target any) bool {
+	return errors.As(err, target)
+}
+
+func Is(err, target error) bool {
+	return errors.Is(err, target)
+}
+
+func Unwrap(err error) error {
+	return errors.Unwrap(err)
+}
+
+type Error struct {
+	Err   error
+	Stack string
+	fatal bool
+}
+
+func (e *Error) Error() string {
+	var sb strings.Builder
+	sb.WriteString(fmt.Sprintf("%v\n", e.Err))
+	sb.WriteString(fmt.Sprintf("Stack Trace:\n%s", e.Stack))
+	return sb.String()
+}
+
+func (e *Error) Fatal() bool {
+	return e.fatal
+}
+
+func (e *Error) Unwrap() error {
+	return e.Err
+}
+
+func NewError(err error) error {
+	if err == nil {
+		return nil
+	}
+
+	// Check if it's already of our own
+	// Error type, so we don't add stack twice.
+	var asError *Error
+	if errors.As(err, &asError) {
+		return err
+	}
+
+	// has the stack trace
+	var stack strings.Builder
+	buf := make([]uintptr, 50)
+	n := runtime.Callers(2, buf)
+	frames := runtime.CallersFrames(buf[:n])
+
+	// Format the stack trace
+	for {
+		frame, more := frames.Next()
+		// Skip runtime and standard library frames
+		if !strings.Contains(frame.File, "runtime/") {
+			stack.WriteString(fmt.Sprintf("\t%s:%d - %s\n", frame.File, frame.Line, frame.Function))
+		}
+		if !more {
+			break
+		}
+	}
+
+	return &Error{
+		Err:   err,
+		Stack: stack.String(),
+	}
+}
+
+func NewFatalError(err error) error {
+	if err == nil {
+		return nil
+	}
+
+	// Check if it's already of our own
+	// Error type.
+	var asError *Error
+	if errors.As(err, &asError) {
+		asError.fatal = true // Set fatal even for existing Error types
+		return err
+	}
+	err2 := NewError(err)
+	err2.(*Error).fatal = true
+	return err2
+}

errors/errors_test.go

@@ -0,0 +1,184 @@
+package errors
+
+import (
+	"errors"
+	"fmt"
+	"testing"
+)
+
+type CustomError struct {
+	Err error
+}
+
+func (e *CustomError) Error() string { return e.Err.Error() }
+
+func IsCustomError(err error) bool {
+	var asError *CustomError
+	return errors.As(err, &asError)
+}
+
+func TestWrapping(t *testing.T) {
+	t.Parallel()
+	originalErr := errors.New("original error")
+	err1 := NewError(originalErr)
+	if !errors.Is(err1, originalErr) {
+		t.Errorf("original error is not wrapped")
+	}
+	if !Is(err1, originalErr) {
+		t.Errorf("original error is not wrapped")
+	}
+	unwrappedErr := errors.Unwrap(err1)
+	if !errors.Is(unwrappedErr, originalErr) {
+		t.Errorf("original error is not wrapped")
+	}
+	if !Is(unwrappedErr, originalErr) {
+		t.Errorf("original error is not wrapped")
+	}
+	unwrappedErr = Unwrap(err1)
+	if !errors.Is(unwrappedErr, originalErr) {
+		t.Errorf("original error is not wrapped")
+	}
+	if !Is(unwrappedErr, originalErr) {
+		t.Errorf("original error is not wrapped")
+	}
+	wrappedErr := fmt.Errorf("wrapped: %w", originalErr)
+	if !errors.Is(wrappedErr, originalErr) {
+		t.Errorf("original error is not wrapped")
+	}
+	if !Is(wrappedErr, originalErr) {
+		t.Errorf("original error is not wrapped")
+	}
+}
+
+func TestNewError(t *testing.T) {
+	t.Parallel()
+	originalErr := &CustomError{errors.New("err1")}
+	if !IsCustomError(originalErr) {
+		t.Errorf("TestNewError fail #1")
+	}
+	err1 := NewError(originalErr)
+	if !IsCustomError(err1) {
+		t.Errorf("TestNewError fail #2")
+	}
+	wrappedErr1 := fmt.Errorf("wrapped %w", err1)
+	if !IsCustomError(wrappedErr1) {
+		t.Errorf("TestNewError fail #3")
+	}
+	unwrappedErr1 := Unwrap(wrappedErr1)
+	if !IsCustomError(unwrappedErr1) {
+		t.Errorf("TestNewError fail #4")
+	}
+}
+
+func TestIsFatal(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name string
+		err  error
+		want bool
+	}{
+		{
+			name: "nil error",
+			err:  nil,
+			want: false,
+		},
+		{
+			name: "simple non-fatal error",
+			err:  fmt.Errorf("regular error"),
+			want: false,
+		},
+		{
+			name: "direct fatal error",
+			err:  NewFatalError(fmt.Errorf("fatal error")),
+			want: true,
+		},
+		{
+			name: "non-fatal Error type",
+			err:  NewError(fmt.Errorf("non-fatal error")),
+			want: false,
+		},
+		{
+			name: "wrapped fatal error - one level",
+			err:  fmt.Errorf("outer: %w", NewFatalError(fmt.Errorf("inner fatal"))),
+			want: true,
+		},
+		{
+			name: "wrapped fatal error - two levels",
+			err: fmt.Errorf("outer: %w",
+				fmt.Errorf("middle: %w",
+					NewFatalError(fmt.Errorf("inner fatal")))),
+			want: true,
+		},
+		{
+			name: "wrapped fatal error - three levels",
+			err: fmt.Errorf("outer: %w",
+				fmt.Errorf("middle1: %w",
+					fmt.Errorf("middle2: %w",
+						NewFatalError(fmt.Errorf("inner fatal"))))),
+			want: true,
+		},
+		{
+			name: "multiple wrapped errors - non-fatal",
+			err: fmt.Errorf("outer: %w",
+				fmt.Errorf("middle: %w",
+					fmt.Errorf("inner: %w",
+						NewError(fmt.Errorf("base error"))))),
+			want: false,
+		},
+		{
+			name: "wrapped non-fatal Error type",
+			err:  fmt.Errorf("outer: %w", NewError(fmt.Errorf("inner"))),
+			want: false,
+		},
+		{
+			name: "wrapped basic error",
+			err:  fmt.Errorf("outer: %w", fmt.Errorf("inner")),
+			want: false,
+		},
+		{
+			name: "fatal error wrapping fatal error",
+			err:  NewFatalError(NewFatalError(fmt.Errorf("double fatal"))),
+			want: true,
+		},
+		{
+			name: "fatal error wrapping non-fatal Error",
+			err:  NewFatalError(NewError(fmt.Errorf("mixed"))),
+			want: true,
+		},
+		{
+			name: "non-fatal Error wrapping fatal error",
+			err:  NewError(NewFatalError(fmt.Errorf("mixed"))),
+			want: true,
+		},
+		{
+			name: "Error wrapping Error",
+			err:  NewError(NewError(fmt.Errorf("double wrapped"))),
+			want: false,
+		},
+		{
+			name: "wrapped nil error",
+			err:  fmt.Errorf("outer: %w", nil),
+			want: false,
+		},
+		{
+			name: "fatal wrapping nil",
+			err:  NewFatalError(nil),
+			want: false,
+		},
+	}
+
+	for _, tt := range tests {
+		tt := tt
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+			got := IsFatal(tt.err)
+			if got != tt.want {
+				t.Errorf("IsFatal() = %v, want %v", got, tt.want)
+				if tt.err != nil {
+					t.Errorf("Error was: %v", tt.err)
+				}
+			}
+		})
+	}
+}

gemget.sh

@@ -0,0 +1,16 @@
+#!/bin/env bash
+set -eu
+set -o pipefail
+
+# Max response size 10MiB
+LOG_LEVEL=debug \
+PRINT_WORKER_STATUS=false \
+DRY_RUN=false \
+NUM_OF_WORKERS=1 \
+WORKER_BATCH_SIZE=1 \
+BLACKLIST_PATH="$(pwd)/blacklist.txt" \
+MAX_RESPONSE_SIZE=10485760 \
+RESPONSE_TIMEOUT=10 \
+PANIC_ON_UNEXPECTED_ERROR=true \
+go run ./bin/gemget/main.go "$@"
+

gemini/blacklist.go

@@ -1,55 +0,0 @@
-package gemini
-
-import (
-	"fmt"
-	"gemini-grc/common"
-	"os"
-	"strings"
-
-	"gemini-grc/config"
-	"gemini-grc/logging"
-)
-
-var Blacklist *[]string //nolint:gochecknoglobals
-
-func LoadBlacklist() {
-	if Blacklist == nil {
-		data, err := os.ReadFile(config.CONFIG.BlacklistPath)
-		if err != nil {
-			Blacklist = &[]string{}
-			logging.LogWarn("Could not load Blacklist file: %v", err)
-			return
-		}
-		lines := strings.Split(string(data), "\n")
-
-		// Ignore lines starting with '#' (comments)
-		filteredLines := func() []string {
-			out := make([]string, 0, len(lines))
-			for _, line := range lines {
-				if !strings.HasPrefix(line, "#") {
-					out = append(out, line)
-				}
-			}
-			return out
-		}()
-
-		if len(lines) > 0 {
-			Blacklist = &filteredLines
-			logging.LogInfo("Blacklist has %d entries", len(*Blacklist))
-		}
-	}
-}
-
-func IsBlacklisted(u string) bool {
-	url, err := common.ParseURL(u, "")
-	if err != nil {
-		return false
-	}
-	hostWithPort := fmt.Sprintf("%s:%d", url.Hostname, url.Port)
-	for _, v := range *Blacklist {
-		if v == url.Hostname || v == hostWithPort {
-			return true
-		}
-	}
-	return false
-}

gemini/errors.go

@@ -0,0 +1,52 @@
+package gemini
+
+import (
+	"fmt"
+
+	"gemini-grc/errors"
+)
+
+// GeminiError is used to represent
+// Gemini network protocol errors only.
+// Should be recorded to the snapshot.
+// See https://geminiprotocol.net/docs/protocol-specification.gmi
+type GeminiError struct {
+	Msg    string
+	Code   int
+	Header string
+}
+
+func (e *GeminiError) Error() string {
+	return fmt.Sprintf("gemini error: code %d %s", e.Code, e.Msg)
+}
+
+func NewGeminiError(code int, header string) error {
+	var msg string
+	switch {
+	case code >= 10 && code < 20:
+		msg = "needs input"
+	case code >= 30 && code < 40:
+		msg = "redirect"
+	case code >= 40 && code < 50:
+		msg = "bad request"
+	case code >= 50 && code < 60:
+		msg = "server error"
+	case code >= 60 && code < 70:
+		msg = "TLS error"
+	default:
+		msg = "unexpected Status code"
+	}
+	return &GeminiError{
+		Msg:    msg,
+		Code:   code,
+		Header: header,
+	}
+}
+
+func IsGeminiError(err error) bool {
+	if err == nil {
+		return false
+	}
+	var asError *GeminiError
+	return errors.As(err, &asError)
+}

gemini/files.go

@@ -2,13 +2,13 @@ package gemini
 
 import (
 	"fmt"
-	"gemini-grc/common"
 	"net/url"
 	"os"
 	"path"
 	"path/filepath"
 	"strings"
 
+	"gemini-grc/common/snapshot"
 	"gemini-grc/logging"
 )
 
@@ -64,7 +64,7 @@ func calcFilePath(rootPath, urlPath string) (string, error) {
 	return finalPath, nil
 }
 
-func SaveToFile(rootPath string, s *common.Snapshot, done chan struct{}) {
+func SaveToFile(rootPath string, s *snapshot.Snapshot, done chan struct{}) {
 	parentPath := path.Join(rootPath, s.URL.Hostname)
 	urlPath := s.URL.Path
 	// If path is empty, add `index.gmi` as the file to save
@@ -105,7 +105,7 @@ func ReadLines(path string) []string {
 		panic(fmt.Sprintf("Failed to read file: %s", err))
 	}
 	lines := strings.Split(string(data), "\n")
-	// Remove last line if empty
+	// remove last line if empty
 	// (happens when file ends with '\n')
 	if lines[len(lines)-1] == "" {
 		lines = lines[:len(lines)-1]

gemini/gemini.go

@@ -1,49 +0,0 @@
-package gemini
-
-import (
-	"fmt"
-	"regexp"
-	"strconv"
-
-	"gemini-grc/common"
-)
-
-// ParseFirstTwoDigits takes a string and returns the first one or two digits as an int.
-// If no valid digits are found, it returns an error.
-func ParseFirstTwoDigits(input string) (int, error) {
-	// Define the regular expression pattern to match one or two leading digits
-	re := regexp.MustCompile(`^(\d{1,2})`)
-
-	// Find the first match in the string
-	matches := re.FindStringSubmatch(input)
-	if len(matches) == 0 {
-		return 0, fmt.Errorf("%w", common.ErrGeminiResponseHeader)
-	}
-
-	// Parse the captured match as an integer
-	snapshot, err := strconv.Atoi(matches[1])
-	if err != nil {
-		return 0, fmt.Errorf("%w: %w", common.ErrTextParse, err)
-	}
-
-	return snapshot, nil
-}
-
-// extractRedirectTarget returns the redirection
-// URL by parsing the header (or error message)
-func extractRedirectTarget(currentURL common.URL, input string) (*common.URL, error) {
-	// \d+ - matches one or more digits
-	// \s+ - matches one or more whitespace
-	// ([^\r]+) - captures everything until it hits a \r (or end of string)
-	pattern := `\d+\s+([^\r]+)`
-	re := regexp.MustCompile(pattern)
-	matches := re.FindStringSubmatch(input)
-	if len(matches) < 2 {
-		return nil, fmt.Errorf("%w: %s", common.ErrGeminiRedirect, input)
-	}
-	newURL, err := common.DeriveAbsoluteURL(currentURL, matches[1])
-	if err != nil {
-		return nil, fmt.Errorf("%w: %w: %s", common.ErrGeminiRedirect, err, input)
-	}
-	return newURL, nil
-}

gemini/geminiLinks.go

@@ -5,22 +5,24 @@ import (
 	"net/url"
 	"regexp"
 
-	"gemini-grc/common"
+	"gemini-grc/common/linkList"
+	url2 "gemini-grc/common/url"
+	"gemini-grc/errors"
 	"gemini-grc/logging"
 	"gemini-grc/util"
 )
 
-func GetPageLinks(currentURL common.URL, gemtext string) common.LinkList {
+func GetPageLinks(currentURL url2.URL, gemtext string) linkList.LinkList {
 	linkLines := util.GetLinesMatchingRegex(gemtext, `(?m)^=>[ \t]+.*`)
 	if len(linkLines) == 0 {
 		return nil
 	}
-	var linkURLs common.LinkList
+	var linkURLs linkList.LinkList
 	// Normalize URLs in links
 	for _, line := range linkLines {
 		linkUrl, err := ParseGeminiLinkLine(line, currentURL.String())
 		if err != nil {
-			logging.LogDebug("%s: %s", common.ErrGeminiLinkLineParse, err)
+			logging.LogDebug("error parsing gemini link line: %s", err)
 			continue
 		}
 		linkURLs = append(linkURLs, *linkUrl)
@@ -31,19 +33,18 @@ func GetPageLinks(currentURL common.URL, gemtext string) common.LinkList {
 // ParseGeminiLinkLine takes a single link line and the current URL,
 // return the URL converted to an absolute URL
 // and its description.
-func ParseGeminiLinkLine(linkLine string, currentURL string) (*common.URL, error) {
+func ParseGeminiLinkLine(linkLine string, currentURL string) (*url2.URL, error) {
 	// Check: currentURL is parseable
 	baseURL, err := url.Parse(currentURL)
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", common.ErrURLParse, err)
+		return nil, errors.NewError(fmt.Errorf("error parsing link line: %w input '%s'", err, linkLine))
 	}
 
 	// Extract the actual URL and the description
 	re := regexp.MustCompile(`^=>[ \t]+(\S+)([ \t]+.*)?`)
 	matches := re.FindStringSubmatch(linkLine)
 	if len(matches) == 0 {
-		// If the line doesn't match the expected format, return it unchanged
+		return nil, errors.NewError(fmt.Errorf("error parsing link line: no regexp match for line %s", linkLine))
-		return nil, fmt.Errorf("%w could not parse gemini link %s", common.ErrGeminiLinkLineParse, linkLine)
 	}
 
 	originalURLStr := matches[1]
@@ -51,7 +52,7 @@ func ParseGeminiLinkLine(linkLine string, currentURL string) (*common.URL, error
 	// Check: Unescape the URL if escaped
 	_, err = url.QueryUnescape(originalURLStr)
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", common.ErrURLDecode, err)
+		return nil, errors.NewError(fmt.Errorf("error parsing link line: %w input '%s'", err, linkLine))
 	}
 
 	description := ""
@@ -62,8 +63,7 @@ func ParseGeminiLinkLine(linkLine string, currentURL string) (*common.URL, error
 	// Parse the URL from the link line
 	parsedURL, err := url.Parse(originalURLStr)
 	if err != nil {
-		// If URL parsing fails, return an error
+		return nil, errors.NewError(fmt.Errorf("error parsing link line: %w input '%s'", err, linkLine))
-		return nil, fmt.Errorf("%w: %w", common.ErrURLParse, err)
 	}
 
 	// If link URL is relative, resolve full URL
@@ -71,17 +71,16 @@ func ParseGeminiLinkLine(linkLine string, currentURL string) (*common.URL, error
 		parsedURL = baseURL.ResolveReference(parsedURL)
 	}
 
-	// Remove usual first space from URL description:
+	// remove usual first space from URL description:
 	// => URL description
 	//       ^^^^^^^^^^^^
 	if len(description) > 0 && description[0] == ' ' {
 		description = description[1:]
 	}
 
-	finalURL, err := common.ParseURL(parsedURL.String(), description)
+	finalURL, err := url2.ParseURL(parsedURL.String(), description, true)
 	if err != nil {
-		// If URL parsing fails, return an error
+		return nil, errors.NewError(fmt.Errorf("error parsing link line: %w input '%s'", err, linkLine))
-		return nil, fmt.Errorf("%w: %w", common.ErrURLParse, err)
 	}
 
 	return finalURL, nil

gemini/geminiLinks_test.go

@@ -1,18 +1,18 @@
 package gemini
 
 import (
-	"errors"
 	"reflect"
+	"strings"
 	"testing"
 
-	"gemini-grc/common"
+	"gemini-grc/common/url"
 )
 
 type TestData struct {
 	currentURL string
 	link       string
-	value      *common.URL
+	value      *url.URL
-	error      error
+	error      string
 }
 
 var data = []TestData{
@@ -20,12 +20,12 @@ var data = []TestData{
 		currentURL: "https://gemini.com/",
 		link:       "https://gemini.com/",
 		value:      nil,
-		error:      common.ErrGeminiLinkLineParse,
+		error:      "error parsing link line",
 	},
 	{
 		currentURL: "gemini://gemi.dev/cgi-bin/xkcd/",
 		link:       "=> archive/ Complete Archive",
-		value: &common.URL{
+		value: &url.URL{
 			Protocol: "gemini",
 			Hostname: "gemi.dev",
 			Port:     1965,
@@ -33,12 +33,12 @@ var data = []TestData{
 			Descr:    "Complete Archive",
 			Full:     "gemini://gemi.dev:1965/cgi-bin/xkcd/archive/",
 		},
-		error: nil,
+		error: "",
 	},
 	{
 		currentURL: "gemini://gemi.dev/cgi-bin/xkcd/",
 		link:       "=> /cgi-bin/xkcd.cgi?a=5&b=6 Example",
-		value: &common.URL{
+		value: &url.URL{
 			Protocol: "gemini",
 			Hostname: "gemi.dev",
 			Port:     1965,
@@ -46,12 +46,12 @@ var data = []TestData{
 			Descr:    "Example",
 			Full:     "gemini://gemi.dev:1965/cgi-bin/xkcd.cgi?a=5&b=6",
 		},
-		error: nil,
+		error: "",
 	},
 	{
 		currentURL: "gemini://gemi.dev/cgi-bin/xkcd/",
 		link:       "=> /cgi-bin/xkcd.cgi?1494 XKCD 1494: Insurance",
-		value: &common.URL{
+		value: &url.URL{
 			Protocol: "gemini",
 			Hostname: "gemi.dev",
 			Port:     1965,
@@ -59,12 +59,12 @@ var data = []TestData{
 			Descr:    "XKCD 1494: Insurance",
 			Full:     "gemini://gemi.dev:1965/cgi-bin/xkcd.cgi?1494",
 		},
-		error: nil,
+		error: "",
 	},
 	{
 		currentURL: "gemini://gemi.dev/cgi-bin/xkcd/",
 		link:       "=> /cgi-bin/xkcd.cgi?1494#f XKCD 1494: Insurance",
-		value: &common.URL{
+		value: &url.URL{
 			Protocol: "gemini",
 			Hostname: "gemi.dev",
 			Port:     1965,
@@ -72,12 +72,12 @@ var data = []TestData{
 			Descr:    "XKCD 1494: Insurance",
 			Full:     "gemini://gemi.dev:1965/cgi-bin/xkcd.cgi?1494#f",
 		},
-		error: nil,
+		error: "",
 	},
 	{
 		currentURL: "gemini://gemi.dev/cgi-bin/xkcd/",
 		link:       "=> /cgi-bin/xkcd.cgi?c=5#d XKCD 1494: Insurance",
-		value: &common.URL{
+		value: &url.URL{
 			Protocol: "gemini",
 			Hostname: "gemi.dev",
 			Port:     1965,
@@ -85,12 +85,12 @@ var data = []TestData{
 			Descr:    "XKCD 1494: Insurance",
 			Full:     "gemini://gemi.dev:1965/cgi-bin/xkcd.cgi?c=5#d",
 		},
-		error: nil,
+		error: "",
 	},
 	{
 		currentURL: "gemini://a.b/c#d",
 		link:       "=> /d/e#f",
-		value: &common.URL{
+		value: &url.URL{
 			Protocol: "gemini",
 			Hostname: "a.b",
 			Port:     1965,
@@ -98,7 +98,7 @@ var data = []TestData{
 			Descr:    "",
 			Full:     "gemini://a.b:1965/d/e#f",
 		},
-		error: nil,
+		error: "",
 	},
 }
 
@@ -110,13 +110,10 @@ func Test(t *testing.T) {
 			if expected.value != nil {
 				t.Errorf("data[%d]: Expected value %v, got %v", i, nil, expected.value)
 			}
-			if !errors.Is(err, common.ErrGeminiLinkLineParse) {
+			if !strings.HasPrefix(err.Error(), expected.error) {
 				t.Errorf("data[%d]: expected error %v, got %v", i, expected.error, err)
 			}
 		} else {
-			if expected.error != nil {
-				t.Errorf("data[%d]: Expected error %v, got %v", i, nil, expected.error)
-			}
 			if !(reflect.DeepEqual(result, expected.value)) {
 				t.Errorf("data[%d]: expected %#v, got %#v", i, expected.value, result)
 			}

gemini/gemini_test.go

@@ -1,69 +0,0 @@
-package gemini
-
-import (
-	"testing"
-
-	"gemini-grc/common"
-)
-
-func TestExtractRedirectTargetFullURL(t *testing.T) {
-	t.Parallel()
-	currentURL, _ := common.ParseURL("gemini://smol.gr", "")
-	input := "redirect: 31 gemini://target.gr"
-	result, err := extractRedirectTarget(*currentURL, input)
-	expected := "gemini://target.gr:1965"
-	if err != nil || (result.String() != expected) {
-		t.Errorf("fail: Expected %s got %s", expected, result)
-	}
-}
-
-func TestExtractRedirectTargetFullURLSlash(t *testing.T) {
-	t.Parallel()
-	currentURL, _ := common.ParseURL("gemini://smol.gr", "")
-	input := "redirect: 31 gemini://target.gr/"
-	result, err := extractRedirectTarget(*currentURL, input)
-	expected := "gemini://target.gr:1965/"
-	if err != nil || (result.String() != expected) {
-		t.Errorf("fail: Expected %s got %s", expected, result)
-	}
-}
-
-func TestExtractRedirectTargetRelativeURL(t *testing.T) {
-	t.Parallel()
-	currentURL, _ := common.ParseURL("gemini://smol.gr", "")
-	input := "redirect: 31 /a/b"
-	result, err := extractRedirectTarget(*currentURL, input)
-	if err != nil || (result.String() != "gemini://smol.gr:1965/a/b") {
-		t.Errorf("fail: %s", result)
-	}
-}
-
-func TestExtractRedirectTargetRelativeURL2(t *testing.T) {
-	t.Parallel()
-	currentURL, _ := common.ParseURL("gemini://nox.im:1965", "")
-	input := "redirect: 31 ./"
-	result, err := extractRedirectTarget(*currentURL, input)
-	if err != nil || (result.String() != "gemini://nox.im:1965/") {
-		t.Errorf("fail: %s", result)
-	}
-}
-
-func TestExtractRedirectTargetRelativeURL3(t *testing.T) {
-	t.Parallel()
-	currentURL, _ := common.ParseURL("gemini://status.zvava.org:1965", "")
-	input := "redirect: 31 index.gmi"
-	result, err := extractRedirectTarget(*currentURL, input)
-	if err != nil || (result.String() != "gemini://status.zvava.org:1965/index.gmi") {
-		t.Errorf("fail: %s", result)
-	}
-}
-
-func TestExtractRedirectTargetWrong(t *testing.T) {
-	t.Parallel()
-	currentURL, _ := common.ParseURL("gemini://smol.gr", "")
-	input := "redirect: 31"
-	result, err := extractRedirectTarget(*currentURL, input)
-	if result != nil || err == nil {
-		t.Errorf("fail: result should be nil, err is %s", err)
-	}
-}

gemini/ip-address-pool.go

@@ -1,54 +0,0 @@
-package gemini
-
-import "sync"
-
-// Used to limit requests per
-// IP address. Maps IP address
-// to number of active connections.
-type IpAddressPool struct {
-	IPs  map[string]int
-	Lock sync.RWMutex
-}
-
-func (p *IpAddressPool) Set(key string, value int) {
-	p.Lock.Lock()         // Lock for writing
-	defer p.Lock.Unlock() // Ensure mutex is unlocked after the write
-	p.IPs[key] = value
-}
-
-func (p *IpAddressPool) Get(key string) int {
-	p.Lock.RLock()         // Lock for reading
-	defer p.Lock.RUnlock() // Ensure mutex is unlocked after reading
-	if value, ok := p.IPs[key]; !ok {
-		return 0
-	} else {
-		return value
-	}
-}
-
-func (p *IpAddressPool) Delete(key string) {
-	p.Lock.Lock()
-	defer p.Lock.Unlock()
-	delete(p.IPs, key)
-}
-
-func (p *IpAddressPool) Incr(key string) {
-	p.Lock.Lock()
-	defer p.Lock.Unlock()
-	if _, ok := p.IPs[key]; !ok {
-		p.IPs[key] = 1
-	} else {
-		p.IPs[key] = p.IPs[key] + 1
-	}
-}
-
-func (p *IpAddressPool) Decr(key string) {
-	p.Lock.Lock()
-	defer p.Lock.Unlock()
-	if val, ok := p.IPs[key]; ok {
-		p.IPs[key] = val - 1
-		if p.IPs[key] == 0 {
-			delete(p.IPs, key)
-		}
-	}
-}

gemini/network.go

@@ -2,44 +2,78 @@ package gemini
 
 import (
 	"crypto/tls"
-	"errors"
 	"fmt"
 	"io"
 	"net"
-	gourl "net/url"
+	stdurl "net/url"
 	"regexp"
 	"slices"
 	"strconv"
 	"strings"
 	"time"
 
-	"gemini-grc/common"
+	errors2 "gemini-grc/common/errors"
+	"gemini-grc/common/snapshot"
+	_url "gemini-grc/common/url"
 	"gemini-grc/config"
+	"gemini-grc/errors"
 	"gemini-grc/logging"
 	"github.com/guregu/null/v5"
 )
 
-type PageData struct {
+// Visit given URL, using the Gemini protocol.
-	ResponseCode   int
+// Mutates given Snapshot with the data.
-	ResponseHeader string
+// In case of error, we store the error string
-	MimeType       string
+// inside snapshot and return the error.
-	Lang           string
+func Visit(url string) (s *snapshot.Snapshot, err error) {
-	GemText        string
+	s, err = snapshot.SnapshotFromURL(url, true)
-	Data           []byte
-}
-
-func getHostIPAddresses(hostname string) ([]string, error) {
-	addrs, err := net.LookupHost(hostname)
 	if err != nil {
-		return nil, fmt.Errorf("%w:%w", common.ErrNetworkDNS, err)
+		return nil, err
 	}
-	return addrs, nil
+
+	defer func() {
+		if err != nil {
+			// GeminiError and HostError should
+			// be stored in the snapshot. Other
+			// errors are returned.
+			if errors2.IsHostError(err) {
+				s.Error = null.StringFrom(err.Error())
+				err = nil
+			} else if IsGeminiError(err) {
+				s.Error = null.StringFrom(err.Error())
+				s.Header = null.StringFrom(errors.Unwrap(err).(*GeminiError).Header)
+				s.ResponseCode = null.IntFrom(int64(errors.Unwrap(err).(*GeminiError).Code))
+				err = nil
+			} else {
+				s = nil
+			}
+		}
+	}()
+
+	data, err := ConnectAndGetData(s.URL.String())
+	if err != nil {
+		return s, err
+	}
+
+	s, err = processData(*s, data)
+	if err != nil {
+		return s, err
+	}
+
+	if isGeminiCapsule(s) {
+		links := GetPageLinks(s.URL, s.GemText.String)
+		if len(links) > 0 {
+			logging.LogDebug("Found %d links", len(links))
+			s.Links = null.ValueFrom(links)
+		}
+	}
+	return s, nil
 }
 
 func ConnectAndGetData(url string) ([]byte, error) {
-	parsedURL, err := gourl.Parse(url)
+	parsedURL, err := stdurl.Parse(url)
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", common.ErrURLParse, err)
+		return nil, errors.NewError(err)
 	}
 	hostname := parsedURL.Hostname()
 	port := parsedURL.Port()
@@ -47,29 +81,28 @@ func ConnectAndGetData(url string) ([]byte, error) {
 		port = "1965"
 	}
 	host := fmt.Sprintf("%s:%s", hostname, port)
+	timeoutDuration := time.Duration(config.CONFIG.ResponseTimeout) * time.Second
 	// Establish the underlying TCP connection.
 	dialer := &net.Dialer{
-		Timeout: time.Duration(config.CONFIG.ResponseTimeout) * time.Second,
+		Timeout: timeoutDuration,
 	}
 	conn, err := dialer.Dial("tcp", host)
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", common.ErrNetwork, err)
+		return nil, errors2.NewHostError(err)
 	}
 	// Make sure we always close the connection.
 	defer func() {
-		// No need to handle error:
-		// Connection will time out eventually if still open somehow.
 		_ = conn.Close()
 	}()
 
 	// Set read and write timeouts on the TCP connection.
-	err = conn.SetReadDeadline(time.Now().Add(time.Duration(config.CONFIG.ResponseTimeout) * time.Second))
+	err = conn.SetReadDeadline(time.Now().Add(timeoutDuration))
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", common.ErrNetworkSetConnectionDeadline, err)
+		return nil, errors2.NewHostError(err)
 	}
-	err = conn.SetWriteDeadline(time.Now().Add(time.Duration(config.CONFIG.ResponseTimeout) * time.Second))
+	err = conn.SetWriteDeadline(time.Now().Add(timeoutDuration))
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", common.ErrNetworkSetConnectionDeadline, err)
+		return nil, errors2.NewHostError(err)
 	}
 
 	// Perform the TLS handshake
@@ -79,8 +112,17 @@ func ConnectAndGetData(url string) ([]byte, error) {
 		// MinVersion:         tls.VersionTLS12, // Use a minimum TLS version. Warning breaks a lot of sites.
 	}
 	tlsConn := tls.Client(conn, tlsConfig)
-	if err := tlsConn.Handshake(); err != nil {
+	err = tlsConn.SetReadDeadline(time.Now().Add(timeoutDuration))
-		return nil, fmt.Errorf("%w: %w", common.ErrNetworkTLS, err)
+	if err != nil {
+		return nil, errors2.NewHostError(err)
+	}
+	err = tlsConn.SetWriteDeadline(time.Now().Add(timeoutDuration))
+	if err != nil {
+		return nil, errors2.NewHostError(err)
+	}
+	err = tlsConn.Handshake()
+	if err != nil {
+		return nil, errors2.NewHostError(err)
 	}
 
 	// We read `buf`-sized chunks and add data to `data`.
@@ -91,10 +133,10 @@ func ConnectAndGetData(url string) ([]byte, error) {
 	// Fix for stupid server bug:
 	// Some servers return 'Header: 53 No proxying to other hosts or ports!'
 	// when the port is 1965 and is still specified explicitly in the URL.
-	_url, _ := common.ParseURL(url, "")
+	url2, _ := _url.ParseURL(url, "", true)
-	_, err = tlsConn.Write([]byte(fmt.Sprintf("%s\r\n", _url.StringNoDefaultPort())))
+	_, err = tlsConn.Write([]byte(fmt.Sprintf("%s\r\n", url2.StringNoDefaultPort())))
 	if err != nil {
-		return nil, fmt.Errorf("%w: %w", common.ErrNetworkCannotWrite, err)
+		return nil, errors2.NewHostError(err)
 	}
 	// Read response bytes in len(buf) byte chunks
 	for {
@@ -103,90 +145,50 @@ func ConnectAndGetData(url string) ([]byte, error) {
 			data = append(data, buf[:n]...)
 		}
 		if len(data) > config.CONFIG.MaxResponseSize {
-			return nil, fmt.Errorf("%w: %v", common.ErrNetworkResponseSizeExceededMax, config.CONFIG.MaxResponseSize)
+			return nil, errors2.NewHostError(err)
 		}
 		if err != nil {
 			if errors.Is(err, io.EOF) {
 				break
 			}
-			return nil, fmt.Errorf("%w: %w", common.ErrNetwork, err)
+			return nil, errors2.NewHostError(err)
 		}
 	}
 	return data, nil
 }
 
-// Visit given URL, using the Gemini protocol.
+func processData(s snapshot.Snapshot, data []byte) (*snapshot.Snapshot, error) {
-// Mutates given Snapshot with the data.
-// In case of error, we store the error string
-// inside snapshot and return the error.
-func Visit(s *common.Snapshot) (err error) {
-	// Don't forget to also store error
-	// response code (if we have one)
-	// and header
-	defer func() {
-		if err != nil {
-			s.Error = null.StringFrom(err.Error())
-			if errors.As(err, new(*common.GeminiError)) {
-				s.Header = null.StringFrom(err.(*common.GeminiError).Header)
-				s.ResponseCode = null.IntFrom(int64(err.(*common.GeminiError).Code))
-			}
-		}
-	}()
-	s.Timestamp = null.TimeFrom(time.Now())
-	data, err := ConnectAndGetData(s.URL.String())
-	if err != nil {
-		return err
-	}
-	pageData, err := processData(data)
-	if err != nil {
-		return err
-	}
-	s.Header = null.StringFrom(pageData.ResponseHeader)
-	s.ResponseCode = null.IntFrom(int64(pageData.ResponseCode))
-	s.MimeType = null.StringFrom(pageData.MimeType)
-	s.Lang = null.StringFrom(pageData.Lang)
-	if pageData.GemText != "" {
-		s.GemText = null.StringFrom(pageData.GemText)
-	}
-	if pageData.Data != nil {
-		s.Data = null.ValueFrom(pageData.Data)
-	}
-	return nil
-}
-
-// processData returne results from
-// parsing Gemini header data:
-// Code, mime type and lang (optional)
-// Returns error if header was invalid
-func processData(data []byte) (*PageData, error) {
 	header, body, err := getHeadersAndData(data)
 	if err != nil {
 		return nil, err
 	}
 	code, mimeType, lang := getMimeTypeAndLang(header)
-	logging.LogDebug("Header: %s", strings.TrimSpace(header))
+
-	if code != 20 {
+	if code != 0 {
-		return nil, common.NewErrGeminiStatusCode(code, header)
+		s.ResponseCode = null.IntFrom(int64(code))
+	}
+	if header != "" {
+		s.Header = null.StringFrom(header)
+	}
+	if mimeType != "" {
+		s.MimeType = null.StringFrom(mimeType)
+	}
+	if lang != "" {
+		s.Lang = null.StringFrom(lang)
 	}
 
-	pageData := PageData{
-		ResponseCode:   code,
-		ResponseHeader: header,
-		MimeType:       mimeType,
-		Lang:           lang,
-	}
 	// If we've got a Gemini document, populate
 	// `GemText` field, otherwise raw data goes to `Data`.
 	if mimeType == "text/gemini" {
 		validBody, err := BytesToValidUTF8(body)
 		if err != nil {
-			return nil, fmt.Errorf("%w: %w", common.ErrUTF8Parse, err)
+			return nil, errors.NewError(err)
 		}
-		pageData.GemText = validBody
+		s.GemText = null.StringFrom(validBody)
 	} else {
-		pageData.Data = body
+		s.Data = null.ValueFrom(body)
 	}
-	return &pageData, nil
+	return &s, nil
 }
 
 // Checks for a Gemini header, which is
@@ -196,29 +198,42 @@ func processData(data []byte) (*PageData, error) {
 func getHeadersAndData(data []byte) (string, []byte, error) {
 	firstLineEnds := slices.Index(data, '\n')
 	if firstLineEnds == -1 {
-		return "", nil, common.ErrGeminiResponseHeader
+		return "", nil, errors2.NewHostError(fmt.Errorf("error parsing header"))
 	}
 	firstLine := string(data[:firstLineEnds])
 	rest := data[firstLineEnds+1:]
-	return firstLine, rest, nil
+	return strings.TrimSpace(firstLine), rest, nil
 }
 
-// Parses code, mime type and language
+// getMimeTypeAndLang Parses code, mime type and language
-// from a Gemini header.
+// given a Gemini header.
-// Examples:
-// `20 text/gemini lang=en` (code, mimetype, lang)
-// `20 text/gemini` (code, mimetype)
-// `31 gemini://redirected.to/other/site` (code)
 func getMimeTypeAndLang(headers string) (int, string, string) {
-	// Regex that parses code, mimetype & optional charset/lang parameters
+	// First try to match the full format: "<code> <mimetype> [charset=<value>] [lang=<value>]"
-	re := regexp.MustCompile(`^(\d+)\s+([a-zA-Z0-9/\-+]+)(?:[;\s]+(?:(?:charset|lang)=([a-zA-Z0-9-]+)))?\s*$`)
+	// The regex looks for:
+	// - A number (\d+)
+	// - Followed by whitespace and a mimetype ([a-zA-Z0-9/\-+]+)
+	// - Optionally followed by charset and/or lang parameters in any order
+	// - Only capturing the lang value, ignoring charset
+	re := regexp.MustCompile(`^(\d+)\s+([a-zA-Z0-9/\-+]+)(?:(?:[\s;]+(?:charset=[^;\s]+|lang=([a-zA-Z0-9-]+)))*)\s*$`)
 	matches := re.FindStringSubmatch(headers)
 	if matches == nil || len(matches) <= 1 {
-		// Try to get code at least
+		// If full format doesn't match, try to match redirect format: "<code> <URL>"
-		re := regexp.MustCompile(`^(\d+)\s+`)
+		// This handles cases like "31 gemini://example.com"
+		re := regexp.MustCompile(`^(\d+)\s+(.+)$`)
 		matches := re.FindStringSubmatch(headers)
 		if matches == nil || len(matches) <= 1 {
-			return 0, "", ""
+			// If redirect format doesn't match, try to match just a status code
+			// This handles cases like "99"
+			re := regexp.MustCompile(`^(\d+)\s*$`)
+			matches := re.FindStringSubmatch(headers)
+			if matches == nil || len(matches) <= 1 {
+				return 0, "", ""
+			}
+			code, err := strconv.Atoi(matches[1])
+			if err != nil {
+				return 0, "", ""
+			}
+			return code, "", ""
 		}
 		code, err := strconv.Atoi(matches[1])
 		if err != nil {
@@ -231,6 +246,10 @@ func getMimeTypeAndLang(headers string) (int, string, string) {
 		return 0, "", ""
 	}
 	mimeType := matches[2]
-	param := matches[3] // This will capture either charset or lang value
+	lang := matches[3] // Will be empty string if no lang parameter was found
-	return code, mimeType, param
+	return code, mimeType, lang
+}
+
+func isGeminiCapsule(s *snapshot.Snapshot) bool {
+	return !s.Error.Valid && s.MimeType.Valid && s.MimeType.String == "text/gemini"
 }

gemini/network_test.go

@@ -1,78 +1,366 @@
 package gemini
 
 import (
+	"slices"
+	"strings"
 	"testing"
+
+	"gemini-grc/common/snapshot"
 )
 
-// Test for input: `20 text/gemini`
+func TestGetHeadersAndData(t *testing.T) {
-func TestGetMimeTypeAndLang1(t *testing.T) {
 	t.Parallel()
-	code, mimeType, lang := getMimeTypeAndLang("20 text/gemini")
+	tests := []struct {
-	if code != 20 || mimeType != "text/gemini" || lang != "" {
+		input       []byte
-		t.Errorf("Expected (20, 'text/gemini', ''), got (%d, '%s', '%s')", code, mimeType, lang)
+		header      string
+		body        []byte
+		expectError bool
+	}{
+		{[]byte("20 text/gemini\r\nThis is the body"), "20 text/gemini", []byte("This is the body"), false},
+		{[]byte("20 text/gemini\nThis is the body"), "20 text/gemini", []byte("This is the body"), false},
+		{[]byte("53 No proxying!\r\n"), "53 No proxying!", []byte(""), false},
+		{[]byte("No header"), "", nil, true},
+	}
+
+	for _, test := range tests {
+		header, body, err := getHeadersAndData(test.input)
+
+		if test.expectError && err == nil {
+			t.Errorf("Expected error, got nil for input: %s", test.input)
+		}
+
+		if !test.expectError && err != nil {
+			t.Errorf("Unexpected error for input '%s': %v", test.input, err)
+		}
+
+		if header != test.header {
+			t.Errorf("Expected header '%s', got '%s' for input: %s", test.header, header, test.input)
+		}
+
+		if !slices.Equal(body, test.body) {
+			t.Errorf("Expected body '%s', got '%s' for input: %s", test.body, string(body), test.input)
+		}
 	}
 }
 
-func TestGetMimeTypeAndLang11(t *testing.T) {
+func TestGetMimeTypeAndLang(t *testing.T) {
 	t.Parallel()
-	code, mimeType, lang := getMimeTypeAndLang("20 text/gemini\n")
+	tests := []struct {
-	if code != 20 || mimeType != "text/gemini" || lang != "" {
+		header   string
-		t.Errorf("Expected (20, 'text/gemini', ''), got (%d, '%s', '%s')", code, mimeType, lang)
+		code     int
+		mimeType string
+		lang     string
+	}{
+		{"20 text/gemini lang=en", 20, "text/gemini", "en"},
+		{"20 text/gemini", 20, "text/gemini", ""},
+		{"31 gemini://redirected.to/other/site", 31, "", ""},
+		{"20 text/plain;charset=utf-8", 20, "text/plain", ""},
+		{"20 text/plain;lang=el-GR", 20, "text/plain", "el-GR"},
+		{"20 text/gemini;lang=en-US;charset=utf-8", 20, "text/gemini", "en-US"}, // charset should be ignored
+		{"Invalid header", 0, "", ""},
+		{"99", 99, "", ""},
+	}
+
+	for _, test := range tests {
+		code, mimeType, lang := getMimeTypeAndLang(test.header)
+
+		if code != test.code {
+			t.Errorf("Expected code %d, got %d for header: %s", test.code, code, test.header)
+		}
+
+		if mimeType != test.mimeType {
+			t.Errorf("Expected mimeType '%s', got '%s' for header: %s", test.mimeType, mimeType, test.header)
+		}
+
+		if lang != test.lang {
+			t.Errorf("Expected lang '%s', got '%s' for header: %s", test.lang, lang, test.header)
+		}
 	}
 }
 
-func TestGetMimeTypeAndLang12(t *testing.T) {
+func TestProcessData(t *testing.T) {
 	t.Parallel()
-	code, mimeType, lang := getMimeTypeAndLang("20 text/plain; charset=utf-8")
+	tests := []struct {
-	if code != 20 || mimeType != "text/plain" || lang != "utf-8" {
+		name          string
-		t.Errorf("Expected (20, 'text/plain', ''), got (%d, '%s', '%s')", code, mimeType, lang)
+		inputData     []byte
+		expectedCode  int
+		expectedMime  string
+		expectedLang  string
+		expectedData  []byte
+		expectedError bool
+	}{
+		{
+			name:          "Gemini document",
+			inputData:     []byte("20 text/gemini\r\n# Hello\nWorld"),
+			expectedCode:  20,
+			expectedMime:  "text/gemini",
+			expectedLang:  "",
+			expectedData:  []byte("# Hello\nWorld"),
+			expectedError: false,
+		},
+		{
+			name:          "Gemini document with language",
+			inputData:     []byte("20 text/gemini lang=en\r\n# Hello\nWorld"),
+			expectedCode:  20,
+			expectedMime:  "text/gemini",
+			expectedLang:  "en",
+			expectedData:  []byte("# Hello\nWorld"),
+			expectedError: false,
+		},
+		{
+			name:          "Non-Gemini document",
+			inputData:     []byte("20 text/html\r\n<h1>Hello</h1>"),
+			expectedCode:  20,
+			expectedMime:  "text/html",
+			expectedLang:  "",
+			expectedData:  []byte("<h1>Hello</h1>"),
+			expectedError: false,
+		},
+		{
+			name:          "Error header",
+			inputData:     []byte("53 No proxying!\r\n"),
+			expectedCode:  53,
+			expectedMime:  "",
+			expectedLang:  "",
+			expectedData:  []byte(""),
+			expectedError: false,
+		},
+		{
+			name:          "Invalid header",
+			inputData:     []byte("Invalid header"),
+			expectedError: true,
+		},
+	}
+
+	for _, test := range tests {
+		t.Run(test.name, func(t *testing.T) {
+			s := snapshot.Snapshot{}
+			result, err := processData(s, test.inputData)
+
+			if test.expectedError && err == nil {
+				t.Errorf("Expected error, got nil")
+				return
+			}
+
+			if !test.expectedError && err != nil {
+				t.Errorf("Unexpected error: %v", err)
+				return
+			}
+
+			if test.expectedError {
+				return
+			}
+
+			if int(result.ResponseCode.ValueOrZero()) != test.expectedCode {
+				t.Errorf("Expected code %d, got %d", test.expectedCode, int(result.ResponseCode.ValueOrZero()))
+			}
+
+			if result.MimeType.ValueOrZero() != test.expectedMime {
+				t.Errorf("Expected mimeType '%s', got '%s'", test.expectedMime, result.MimeType.ValueOrZero())
+			}
+
+			if result.Lang.ValueOrZero() != test.expectedLang {
+				t.Errorf("Expected lang '%s', got '%s'", test.expectedLang, result.Lang.ValueOrZero())
+			}
+
+			if test.expectedMime == "text/gemini" {
+				if !strings.Contains(result.GemText.String, string(test.expectedData)) {
+					t.Errorf("Expected GemText '%s', got '%s'", test.expectedData, result.GemText.String)
+				}
+			} else {
+				if !slices.Equal(result.Data.ValueOrZero(), test.expectedData) {
+					t.Errorf("Expected data '%s', got '%s'", test.expectedData, result.Data.ValueOrZero())
+				}
+			}
+		})
 	}
 }
 
-func TestGetMimeTypeAndLang13(t *testing.T) {
+//// Mock Gemini server for testing ConnectAndGetData
+//func mockGeminiServer(response string, delay time.Duration, closeConnection bool) net.Listener {
+//	listener, err := net.Listen("tcp", "127.0.0.1:0") // Bind to a random available port
+//	if err != nil {
+//		panic(fmt.Sprintf("Failed to create mock server: %v", err))
+//	}
+//
+//	go func() {
+//		conn, err := listener.Accept()
+//		if err != nil {
+//			if !closeConnection { // Don't panic if we closed the connection on purpose
+//				panic(fmt.Sprintf("Failed to accept connection: %v", err))
+//			}
+//			return
+//		}
+//		defer conn.Close()
+//
+//		time.Sleep(delay) // Simulate network latency
+//
+//		_, err = conn.Write([]byte(response))
+//		if err != nil && !closeConnection {
+//			panic(fmt.Sprintf("Failed to write response: %v", err))
+//		}
+//	}()
+//
+//	return listener
+//}
+
+// func TestConnectAndGetData(t *testing.T) {
+// 	config.CONFIG = config.ConfigStruct{
+// 		ResponseTimeout: 5,
+// 		MaxResponseSize: 1024 * 1024,
+// 	}
+// 	tests := []struct {
+// 		name            string
+// 		serverResponse  string
+// 		serverDelay     time.Duration
+// 		expectedData    []byte
+// 		expectedError   bool
+// 		closeConnection bool
+// 	}{
+// 		{
+// 			name:           "Successful response",
+// 			serverResponse: "20 text/gemini\r\n# Hello",
+// 			expectedData:   []byte("20 text/gemini\r\n# Hello"),
+// 			expectedError:  false,
+// 		},
+// 		{
+// 			name:           "Server error",
+// 			serverResponse: "50 Server error\r\n",
+// 			expectedData:   []byte("50 Server error\r\n"),
+// 			expectedError:  false,
+// 		},
+// 		{
+// 			name:          "Timeout",
+// 			serverDelay:   6 * time.Second, // Longer than the timeout
+// 			expectedError: true,
+// 		},
+// 		{
+// 			name:            "Server closes connection",
+// 			closeConnection: true,
+// 			expectedError:   true,
+// 		},
+// 	}
+
+// 	for _, test := range tests {
+// 		t.Run(test.name, func(t *testing.T) {
+// 			listener := mockGeminiServer(test.serverResponse, test.serverDelay, test.closeConnection)
+// 			defer func() {
+// 				test.closeConnection = true // Prevent panic in mock server
+// 				listener.Close()
+// 			}()
+// 			addr := listener.Addr().String()
+// 			data, err := ConnectAndGetData(fmt.Sprintf("gemini://%s/", addr))
+
+// 			if test.expectedError && err == nil {
+// 				t.Errorf("Expected error, got nil")
+// 			}
+
+// 			if !test.expectedError && err != nil {
+// 				t.Errorf("Unexpected error: %v", err)
+// 			}
+
+// 			if !slices.Equal(data, test.expectedData) {
+// 				t.Errorf("Expected data '%s', got '%s'", test.expectedData, data)
+// 			}
+// 		})
+// 	}
+// }
+
+// func TestVisit(t *testing.T) {
+// 	config.CONFIG = config.ConfigStruct{
+// 		ResponseTimeout: 5,
+// 		MaxResponseSize: 1024 * 1024,
+// 	}
+// 	tests := []struct {
+// 		name           string
+// 		serverResponse string
+// 		expectedCode   int
+// 		expectedMime   string
+// 		expectedError  bool
+// 		expectedLinks  []string
+// 	}{
+// 		{
+// 			name:           "Successful response",
+// 			serverResponse: "20 text/gemini\r\n# Hello\n=> /link1 Link 1\n=> /link2 Link 2",
+// 			expectedCode:   20,
+// 			expectedMime:   "text/gemini",
+// 			expectedError:  false,
+// 			expectedLinks:  []string{"gemini://127.0.0.1:1965/link1", "gemini://127.0.0.1:1965/link2"},
+// 		},
+// 		{
+// 			name:           "Server error",
+// 			serverResponse: "50 Server error\r\n",
+// 			expectedCode:   50,
+// 			expectedMime:   "Server error",
+// 			expectedError:  false,
+// 			expectedLinks:  []string{},
+// 		},
+// 	}
+
+// 	for _, test := range tests {
+// 		t.Run(test.name, func(t *testing.T) {
+// 			listener := mockGeminiServer(test.serverResponse, 0, false)
+// 			defer listener.Close()
+// 			addr := listener.Addr().String()
+// 			snapshot, err := Visit(fmt.Sprintf("gemini://%s/", addr))
+
+// 			if test.expectedError && err == nil {
+// 				t.Errorf("Expected error, got nil")
+// 			}
+
+// 			if !test.expectedError && err != nil {
+// 				t.Errorf("Unexpected error: %v", err)
+// 			}
+
+// 			if snapshot.ResponseCode.ValueOrZero() != int64(test.expectedCode) {
+// 				t.Errorf("Expected code %d, got %d", test.expectedCode, snapshot.ResponseCode.ValueOrZero())
+// 			}
+
+// 			if snapshot.MimeType.ValueOrZero() != test.expectedMime {
+// 				t.Errorf("Expected mimeType '%s', got '%s'", test.expectedMime, snapshot.MimeType.ValueOrZero())
+// 			}
+
+// 			if test.expectedLinks != nil {
+// 				links, _ := snapshot.Links.Value()
+
+// 				if len(links) != len(test.expectedLinks) {
+// 					t.Errorf("Expected %d links, got %d", len(test.expectedLinks), len(links))
+// 				}
+// 				for i, link := range links {
+// 					if link != test.expectedLinks[i] {
+// 						t.Errorf("Expected link '%s', got '%s'", test.expectedLinks[i], link)
+// 					}
+// 				}
+// 			}
+// 		})
+// 	}
+// }
+
+func TestVisit_InvalidURL(t *testing.T) {
 	t.Parallel()
-	code, mimeType, lang := getMimeTypeAndLang("20 text/gemini; charset=utf-8")
+	_, err := Visit("invalid-url")
-	if code != 20 || mimeType != "text/gemini" || lang != "utf-8" {
+	if err == nil {
-		t.Errorf("Expected (20, 'text/plain', ''), got (%d, '%s', '%s')", code, mimeType, lang)
+		t.Errorf("Expected error for invalid URL, got nil")
 	}
 }
 
-func TestGetTypeAndLang2(t *testing.T) {
+//func TestVisit_GeminiError(t *testing.T) {
-	t.Parallel()
+//	listener := mockGeminiServer("51 Not Found\r\n", 0, false)
-	code, mimeType, lang := getMimeTypeAndLang("20 text/gemini charset=en")
+//	defer listener.Close()
-	if code != 20 || mimeType != "text/gemini" || lang != "en" {
+//	addr := listener.Addr().String()
-		t.Errorf("Expected (20, 'text/gemini', 'en'), got (%d, '%s', '%s')", code, mimeType, lang)
+//
-	}
+//	s, err := Visit(fmt.Sprintf("gemini://%s/", addr))
-}
+//	if err != nil {
-
+//		t.Errorf("Unexpected error: %v", err)
-func TestGetTypeAndLang21(t *testing.T) {
+//	}
-	t.Parallel()
+//
-	code, mimeType, lang := getMimeTypeAndLang("20 text/gemini lang=en")
+//	expectedError := "51 Not Found"
-	if code != 20 || mimeType != "text/gemini" || lang != "en" {
+//	if s.Error.ValueOrZero() != expectedError {
-		t.Errorf("Expected (20, 'text/gemini', 'en'), got (%d, '%s', '%s')", code, mimeType, lang)
+//		t.Errorf("Expected error in snapshot: %v, got %v", expectedError, s.Error)
-	}
+//	}
-}
+//
-
+//	expectedCode := 51
-func TestGetMimeTypeAndLang3(t *testing.T) {
+//	if s.ResponseCode.ValueOrZero() != int64(expectedCode) {
-	t.Parallel()
+//		t.Errorf("Expected code %d, got %d", expectedCode, s.ResponseCode.ValueOrZero())
-	code, mimeType, lang := getMimeTypeAndLang("31 gemini://redirect.to/page")
+//	}
-	if code != 31 || mimeType != "" || lang != "" {
+//}
-		t.Errorf("Expected (20, '', ''), got (%d, '%s', '%s')", code, mimeType, lang)
-	}
-}
-
-func TestGetMimeTypeAndLang4(t *testing.T) {
-	t.Parallel()
-	code, mimeType, lang := getMimeTypeAndLang("aaafdasdasd")
-	if code != 0 || mimeType != "" || lang != "" {
-		t.Errorf("Expected (0, '', ''), got (%d, '%s', '%s')", code, mimeType, lang)
-	}
-}
-
-func TestGetMimeTypeAndLang5(t *testing.T) {
-	t.Parallel()
-	code, mimeType, lang := getMimeTypeAndLang("")
-	if code != 0 || mimeType != "" || lang != "" {
-		t.Errorf("Expected (0, '', ''), got (%d, '%s', '%s')", code, mimeType, lang)
-	}
-}

gemini/processing.go

@@ -26,7 +26,7 @@ func BytesToValidUTF8(input []byte) (string, error) {
 	if len(input) > maxSize {
 		return "", fmt.Errorf("%w: %d bytes (max %d)", ErrInputTooLarge, len(input), maxSize)
 	}
-	// Remove NULL byte 0x00 (ReplaceAll accepts slices)
+	// remove NULL byte 0x00 (ReplaceAll accepts slices)
 	inputNoNull := bytes.ReplaceAll(input, []byte{byte(0)}, []byte{})
 	if utf8.Valid(inputNoNull) {
 		return string(inputNoNull), nil

gemini/robotmatch.go

@@ -2,10 +2,11 @@ package gemini
 
 import (
 	"fmt"
-	"gemini-grc/common"
 	"strings"
 	"sync"
 
+	"gemini-grc/common/snapshot"
+	geminiUrl "gemini-grc/common/url"
 	"gemini-grc/logging"
 )
 
@@ -16,7 +17,7 @@ import (
 // list is stored for caching.
 var RobotsCache sync.Map //nolint:gochecknoglobals
 
-func populateBlacklist(key string) (entries []string) {
+func populateRobotsCache(key string) (entries []string, _err error) {
 	// We either store an empty list when
 	// no rules, or a list of disallowed URLs.
 	// This applies even if we have an error
@@ -27,53 +28,60 @@ func populateBlacklist(key string) (entries []string) {
 	url := fmt.Sprintf("gemini://%s/robots.txt", key)
 	robotsContent, err := ConnectAndGetData(url)
 	if err != nil {
-		logging.LogDebug("robots.txt error %s", err)
+		return []string{}, err
-		return []string{}
 	}
-	robotsData, err := processData(robotsContent)
+	s, err := snapshot.SnapshotFromURL(url, true)
+	if err != nil {
+		return []string{}, nil
+	}
+	s, err = processData(*s, robotsContent)
 	if err != nil {
 		logging.LogDebug("robots.txt error %s", err)
-		return []string{}
+		return []string{}, nil
 	}
-	if robotsData.ResponseCode != 20 {
+	if s.ResponseCode.ValueOrZero() != 20 {
-		logging.LogDebug("robots.txt error code %d, ignoring", robotsData.ResponseCode)
+		logging.LogDebug("robots.txt error code %d, ignoring", s.ResponseCode.ValueOrZero())
-		return []string{}
+		return []string{}, nil
 	}
 	// Some return text/plain, others text/gemini.
 	// According to spec, the first is correct,
 	// however let's be lenient
 	var data string
 	switch {
-	case robotsData.MimeType == "text/plain":
+	case s.MimeType.ValueOrZero() == "text/plain":
-		data = string(robotsData.Data)
+		data = string(s.Data.ValueOrZero())
-	case robotsData.MimeType == "text/gemini":
+	case s.MimeType.ValueOrZero() == "text/gemini":
-		data = robotsData.GemText
+		data = s.GemText.ValueOrZero()
 	default:
-		return []string{}
+		return []string{}, nil
 	}
 	entries = ParseRobotsTxt(data, key)
-	return entries
+	return entries, nil
 }
 
 // RobotMatch checks if the snapshot URL matches
 // a robots.txt allow rule.
-func RobotMatch(u string) bool {
+func RobotMatch(u string) (bool, error) {
-	url, err := common.ParseURL(u, "")
+	url, err := geminiUrl.ParseURL(u, "", true)
 	if err != nil {
-		return false
+		return false, err
 	}
 	key := strings.ToLower(fmt.Sprintf("%s:%d", url.Hostname, url.Port))
-	logging.LogDebug("Checking robots.txt cache for %s", key)
 	var disallowedURLs []string
 	cacheEntries, ok := RobotsCache.Load(key)
 	if !ok {
 		// First time check, populate robot cache
-		disallowedURLs = populateBlacklist(key)
+		disallowedURLs, err := populateRobotsCache(key)
-		logging.LogDebug("Added to robots.txt cache: %v => %v", key, disallowedURLs)
+		if err != nil {
+			return false, err
+		}
+		if len(disallowedURLs) > 0 {
+			logging.LogDebug("Added to robots.txt cache: %v => %v", key, disallowedURLs)
+		}
 	} else {
 		disallowedURLs, _ = cacheEntries.([]string)
 	}
-	return isURLblocked(disallowedURLs, url.Full)
+	return isURLblocked(disallowedURLs, url.Full), nil
 }
 
 func isURLblocked(disallowedURLs []string, input string) bool {

gemini/worker.go

@@ -1,344 +0,0 @@
-package gemini
-
-import (
-	"errors"
-	"fmt"
-	"gemini-grc/common"
-	_db "gemini-grc/db"
-	"strings"
-	"time"
-
-	"gemini-grc/logging"
-	"gemini-grc/util"
-	"github.com/guregu/null/v5"
-	"github.com/jmoiron/sqlx"
-)
-
-func SpawnWorkers(numOfWorkers int, db *sqlx.DB) {
-	logging.LogInfo("Spawning %d workers", numOfWorkers)
-	statusChan = make(chan WorkerStatus, numOfWorkers)
-	go PrintWorkerStatus(numOfWorkers, statusChan)
-
-	for i := range numOfWorkers {
-		go func(i int) {
-			// Jitter to avoid starting everything at the same time
-			time.Sleep(time.Duration(util.SecureRandomInt(10)) * time.Second)
-			for {
-				RunWorkerWithTx(i, db, nil)
-			}
-		}(i)
-	}
-}
-
-func RunWorkerWithTx(workerID int, db *sqlx.DB, url *string) {
-	statusChan <- WorkerStatus{
-		id:     workerID,
-		status: "Starting up",
-	}
-	defer func() {
-		statusChan <- WorkerStatus{
-			id:     workerID,
-			status: "Done",
-		}
-	}()
-	tx, err := db.Beginx()
-	if err != nil {
-		panic(fmt.Sprintf("Failed to begin transaction: %v", err))
-	}
-	runWorker(workerID, tx, url)
-	logging.LogDebug("[%d] Committing transaction", workerID)
-	err = tx.Commit()
-	// On deadlock errors, rollback and return, otherwise panic.
-	if err != nil {
-		logging.LogError("[%d] Failed to commit transaction: %w", workerID, err)
-		if _db.IsDeadlockError(err) {
-			logging.LogError("[%d] Deadlock detected. Rolling back", workerID)
-			time.Sleep(time.Duration(10) * time.Second)
-			err := tx.Rollback()
-			if err != nil {
-				panic(fmt.Sprintf("[%d] Failed to roll back transaction: %v", workerID, err))
-			}
-			return
-		}
-		panic(fmt.Sprintf("[%d] Failed to commit transaction: %v", workerID, err))
-	}
-	logging.LogDebug("[%d] Worker done!", workerID)
-}
-
-func runWorker(workerID int, tx *sqlx.Tx, url *string) {
-	var urls []string
-	var err error
-
-	// If not given a specific URL,
-	// get some random ones to visit from db.
-	if url == nil {
-		statusChan <- WorkerStatus{
-			id:     workerID,
-			status: "Getting URLs",
-		}
-		urls, err = _db.GetURLsToVisit(tx)
-		if err != nil {
-			logging.LogError("[%d] GeminiError retrieving snapshot: %w", workerID, err)
-			panic("This should never happen")
-		} else if len(urls) == 0 {
-			logging.LogInfo("[%d] No URLs to visit.", workerID)
-			time.Sleep(1 * time.Minute)
-			return
-		}
-	} else {
-		geminiURL, err := common.ParseURL(*url, "")
-		if err != nil {
-			logging.LogError("Invalid URL given: %s", *url)
-			return
-		}
-		urls = []string{geminiURL.String()}
-	}
-
-	// Start visiting URLs.
-	total := len(urls)
-	for i, u := range urls {
-		logging.LogDebug("[%d] Starting %d/%d %s", workerID, i+1, total, u)
-		// We differentiate between errors:
-		// Unexpected errors are the ones returned from the following function.
-		// If an error is unexpected (which should never happen) we panic.
-		// Expected errors are stored as strings within the snapshot.
-		err := workOnUrl(workerID, tx, u)
-		if err != nil {
-			logging.LogError("[%d] Unexpected GeminiError %w while visiting %s", workerID, err, u)
-			util.PrintStackAndPanic(err)
-		}
-		logging.LogDebug("[%d] Done %d/%d.", workerID, i+1, total)
-	}
-}
-
-// workOnUrl visits a URL and stores the result.
-// unexpected errors are returned.
-// expected errors are stored within the snapshot.
-func workOnUrl(workerID int, tx *sqlx.Tx, url string) (err error) {
-	if url == "" {
-		return fmt.Errorf("nil URL given")
-	}
-
-	if IsBlacklisted(url) {
-		logging.LogDebug("[%d] URL matches Blacklist, ignoring %s", workerID, url)
-		return nil
-	}
-
-	s := common.SnapshotFromURL(url)
-
-	// If URL matches a robots.txt disallow line,
-	// add it as an error so next time it won't be
-	// crawled.
-	if RobotMatch(url) {
-		s.Error = null.StringFrom(common.ErrGeminiRobotsDisallowed.Error())
-		err = _db.OverwriteSnapshot(workerID, tx, s)
-		if err != nil {
-			return fmt.Errorf("[%d] %w", workerID, err)
-		}
-		return nil
-	}
-
-	// Resolve IP address via DNS
-	statusChan <- WorkerStatus{
-		id:     workerID,
-		status: fmt.Sprintf("Resolving %s", url),
-	}
-	IPs, err := getHostIPAddresses(s.Host)
-	if err != nil {
-		s.Error = null.StringFrom(err.Error())
-		err = _db.OverwriteSnapshot(workerID, tx, s)
-		if err != nil {
-			return fmt.Errorf("[%d] %w", workerID, err)
-		}
-		return nil
-	}
-
-	for {
-		count := 1
-		if isAnotherWorkerVisitingHost(workerID, IPs) {
-			logging.LogDebug("[%d] Another worker is visiting this host, waiting", workerID)
-			statusChan <- WorkerStatus{
-				id:     workerID,
-				status: fmt.Sprintf("Waiting to grab lock for host %s", s.Host),
-			}
-			time.Sleep(2 * time.Second) // Avoid flood-retrying
-			count++
-			if count == 3 {
-				return
-			}
-		} else {
-			break
-		}
-	}
-
-	statusChan <- WorkerStatus{
-		id:     workerID,
-		status: fmt.Sprintf("Adding to pool %s", url),
-	}
-	AddIPsToPool(IPs)
-	// After finishing, remove the host IPs from
-	// the connections pool, with a small delay
-	// to avoid potentially hitting the same IP quickly.
-	defer func() {
-		go func() {
-			time.Sleep(1 * time.Second)
-			statusChan <- WorkerStatus{
-				id:     workerID,
-				status: fmt.Sprintf("Removing from pool %s", url),
-			}
-			RemoveIPsFromPool(IPs)
-		}()
-	}()
-
-	statusChan <- WorkerStatus{
-		id:     workerID,
-		status: fmt.Sprintf("Visiting %s", url),
-	}
-
-	err = Visit(s)
-	if err != nil {
-		if !common.IsKnownError(err) {
-			logging.LogError("[%d] Unknown error visiting %s: %w", workerID, url, err)
-			return err
-		}
-		s.Error = null.StringFrom(err.Error())
-		// Check if error is redirection, and handle it
-		if errors.As(err, new(*common.GeminiError)) &&
-		    err.(*common.GeminiError).Msg == "redirect" {
-			err = handleRedirection(workerID, tx, s)
-			if err != nil {
-				if common.IsKnownError(err) {
-					s.Error = null.StringFrom(err.Error())
-				} else {
-					return err
-				}
-			}
-		}
-	}
-	// If this is a gemini page, parse possible links inside
-	if !s.Error.Valid && s.MimeType.Valid && s.MimeType.String == "text/gemini" {
-		links := GetPageLinks(s.URL, s.GemText.String)
-		if len(links) > 0 {
-			logging.LogDebug("[%d] Found %d links", workerID, len(links))
-			s.Links = null.ValueFrom(links)
-			err = storeLinks(tx, s)
-			if err != nil {
-				return err
-			}
-		}
-	} else {
-		logging.LogDebug("[%d] Not text/gemini, so not looking for page links", workerID)
-	}
-
-	err = _db.OverwriteSnapshot(workerID, tx, s)
-	logging.LogInfo("[%3d] %2d %s", workerID, s.ResponseCode.ValueOrZero(), s.URL.String())
-	if err != nil {
-		return err
-	}
-
-	return nil
-}
-
-func isAnotherWorkerVisitingHost(workerID int, IPs []string) bool {
-	IPPool.Lock.RLock()
-	defer func() {
-		IPPool.Lock.RUnlock()
-	}()
-	logging.LogDebug("[%d] Checking pool for IPs", workerID)
-	for _, ip := range IPs {
-		_, ok := IPPool.IPs[ip]
-		if ok {
-			return true
-		}
-	}
-	return false
-}
-
-func storeLinks(tx *sqlx.Tx, s *common.Snapshot) error {
-	if s.Links.Valid {
-		var batchSnapshots []*common.Snapshot
-		for _, link := range s.Links.ValueOrZero() {
-			if shouldPersistURL(&link) {
-				newSnapshot := &common.Snapshot{
-					URL:       link,
-					Host:      link.Hostname,
-					Timestamp: null.TimeFrom(time.Now()),
-				}
-				batchSnapshots = append(batchSnapshots, newSnapshot)
-			}
-		}
-
-		if len(batchSnapshots) > 0 {
-			err := _db.SaveLinksToDBinBatches(tx, batchSnapshots)
-			if err != nil {
-				return err
-			}
-		}
-	}
-	return nil
-}
-
-// shouldPersistURL returns true if we
-// should save the URL in the _db.
-// Only gemini:// urls are saved.
-func shouldPersistURL(u *common.URL) bool {
-	return strings.HasPrefix(u.String(), "gemini://")
-}
-
-func haveWeVisitedURL(tx *sqlx.Tx, u *common.URL) (bool, error) {
-	var result bool
-	err := tx.Select(&result, `SELECT TRUE FROM urls WHERE url=$1`, u.String())
-	if err != nil {
-		return false, fmt.Errorf("%w: %w", common.ErrDatabase, err)
-	}
-	if result {
-		return result, nil
-	}
-	err = tx.Select(&result, `SELECT TRUE FROM snapshots WHERE snapshot.url=$1`, u.String())
-	if err != nil {
-		return false, fmt.Errorf("%w: %w", common.ErrDatabase, err)
-	}
-	return result, nil
-}
-
-// handleRedirection saves redirect URL as new snapshot
-func handleRedirection(workerID int, tx *sqlx.Tx, s *common.Snapshot) error {
-	newURL, err := extractRedirectTarget(s.URL, s.Error.ValueOrZero())
-	if err != nil {
-		if errors.Is(err, common.ErrGeminiRedirect) {
-			logging.LogDebug("[%d] %s", workerID, err)
-		}
-		return err
-	}
-	logging.LogDebug("[%d] Page redirects to %s", workerID, newURL)
-	// Insert fresh snapshot with new URL
-	if shouldPersistURL(newURL) {
-		snapshot := &common.Snapshot{
-			// UID:       uid.UID(),
-			URL:       *newURL,
-			Host:      newURL.Hostname,
-			Timestamp: null.TimeFrom(time.Now()),
-		}
-		logging.LogDebug("[%d] Saving redirection URL %s", workerID, snapshot.URL.String())
-		err = _db.SaveSnapshotIfNew(tx, snapshot)
-		if err != nil {
-			return err
-		}
-	}
-	return nil
-}
-
-func GetSnapshotFromURL(tx *sqlx.Tx, url string) ([]common.Snapshot, error) {
-	query := `
-	SELECT *
-	FROM snapshots
-	WHERE url=$1
-	LIMIT 1
-	`
-	var snapshots []common.Snapshot
-	err := tx.Select(&snapshots, query, url)
-	if err != nil {
-		return nil, err
-	}
-	return snapshots, nil
-}

go.mod

@@ -4,13 +4,13 @@ go 1.23.1
 
 require (
 	github.com/guregu/null/v5 v5.0.0
-	github.com/jackc/pgx/v5 v5.7.1
+	github.com/jackc/pgx/v5 v5.7.2
 	github.com/jmoiron/sqlx v1.4.0
 	github.com/lib/pq v1.10.9
 	github.com/matoous/go-nanoid/v2 v2.1.0
 	github.com/rs/zerolog v1.33.0
 	github.com/stretchr/testify v1.9.0
-	golang.org/x/text v0.19.0
+	golang.org/x/text v0.21.0
 )
 
 require (
@@ -19,12 +19,12 @@ require (
 	github.com/jackc/pgservicefile v0.0.0-20240606120523-5a60cdf6a761 // indirect
 	github.com/jackc/puddle/v2 v2.2.2 // indirect
 	github.com/kr/text v0.2.0 // indirect
-	github.com/mattn/go-colorable v0.1.13 // indirect
+	github.com/mattn/go-colorable v0.1.14 // indirect
 	github.com/mattn/go-isatty v0.0.20 // indirect
 	github.com/pmezard/go-difflib v1.0.0 // indirect
 	github.com/rogpeppe/go-internal v1.13.1 // indirect
-	golang.org/x/crypto v0.27.0 // indirect
+	golang.org/x/crypto v0.32.0 // indirect
-	golang.org/x/sync v0.8.0 // indirect
+	golang.org/x/sync v0.10.0 // indirect
-	golang.org/x/sys v0.25.0 // indirect
+	golang.org/x/sys v0.29.0 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect
 )

go.sum

@@ -14,8 +14,8 @@ github.com/jackc/pgpassfile v1.0.0 h1:/6Hmqy13Ss2zCq62VdNG8tM1wchn8zjSGOBJ6icpsI
 github.com/jackc/pgpassfile v1.0.0/go.mod h1:CEx0iS5ambNFdcRtxPj5JhEz+xB6uRky5eyVu/W2HEg=
 github.com/jackc/pgservicefile v0.0.0-20240606120523-5a60cdf6a761 h1:iCEnooe7UlwOQYpKFhBabPMi4aNAfoODPEFNiAnClxo=
 github.com/jackc/pgservicefile v0.0.0-20240606120523-5a60cdf6a761/go.mod h1:5TJZWKEWniPve33vlWYSoGYefn3gLQRzjfDlhSJ9ZKM=
-github.com/jackc/pgx/v5 v5.7.1 h1:x7SYsPBYDkHDksogeSmZZ5xzThcTgRz++I5E+ePFUcs=
+github.com/jackc/pgx/v5 v5.7.2 h1:mLoDLV6sonKlvjIEsV56SkWNCnuNv531l94GaIzO+XI=
-github.com/jackc/pgx/v5 v5.7.1/go.mod h1:e7O26IywZZ+naJtWWos6i6fvWK+29etgITqrqHLfoZA=
+github.com/jackc/pgx/v5 v5.7.2/go.mod h1:ncY89UGWxg82EykZUwSpUKEfccBGGYq1xjrOpsbsfGQ=
 github.com/jackc/puddle/v2 v2.2.2 h1:PR8nw+E/1w0GLuRFSmiioY6UooMp6KJv0/61nB7icHo=
 github.com/jackc/puddle/v2 v2.2.2/go.mod h1:vriiEXHvEE654aYKXXjOvZM39qJ0q+azkZFrfEOc3H4=
 github.com/jmoiron/sqlx v1.4.0 h1:1PLqN7S1UYp5t4SrVVnt4nUVNemrDAtxlulVe+Qgm3o=
@@ -28,8 +28,9 @@ github.com/lib/pq v1.10.9 h1:YXG7RB+JIjhP29X+OtkiDnYaXQwpS4JEWq7dtCCRUEw=
 github.com/lib/pq v1.10.9/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
 github.com/matoous/go-nanoid/v2 v2.1.0 h1:P64+dmq21hhWdtvZfEAofnvJULaRR1Yib0+PnU669bE=
 github.com/matoous/go-nanoid/v2 v2.1.0/go.mod h1:KlbGNQ+FhrUNIHUxZdL63t7tl4LaPkZNpUULS8H4uVM=
-github.com/mattn/go-colorable v0.1.13 h1:fFA4WZxdEF4tXPZVKMLwD8oUnCTTo08duU7wxecdEvA=
 github.com/mattn/go-colorable v0.1.13/go.mod h1:7S9/ev0klgBDR4GtXTXX8a3vIGJpMovkB8vQcUbaXHg=
+github.com/mattn/go-colorable v0.1.14 h1:9A9LHSqF/7dyVVX6g0U9cwm9pG3kP9gSzcuIPHPsaIE=
+github.com/mattn/go-colorable v0.1.14/go.mod h1:6LmQG8QLFO4G5z1gPvYEzlUgJ2wF+stgPZH1UqBm1s8=
 github.com/mattn/go-isatty v0.0.16/go.mod h1:kYGgaQfpe5nmfYZH+SKPsOc2e4SrIfOl2e/yFXSvRLM=
 github.com/mattn/go-isatty v0.0.19/go.mod h1:W+V8PltTTMOvKvAeJH7IuucS94S2C6jfK/D7dTCTo3Y=
 github.com/mattn/go-isatty v0.0.20 h1:xfD0iDuEKnDkl03q4limB+vH+GxLEtL/jb4xVJSWWEY=
@@ -49,17 +50,17 @@ github.com/stretchr/testify v1.3.0/go.mod h1:M5WIy9Dh21IEIfnGCwXGc5bZfKNJtfHm1UV
 github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
 github.com/stretchr/testify v1.9.0 h1:HtqpIVDClZ4nwg75+f6Lvsy/wHu+3BoSGCbBAcpTsTg=
 github.com/stretchr/testify v1.9.0/go.mod h1:r2ic/lqez/lEtzL7wO/rwa5dbSLXVDPFyf8C91i36aY=
-golang.org/x/crypto v0.27.0 h1:GXm2NjJrPaiv/h1tb2UH8QfgC/hOf/+z0p6PT8o1w7A=
+golang.org/x/crypto v0.32.0 h1:euUpcYgM8WcP71gNpTqQCn6rC2t6ULUPiOzfWaXVVfc=
-golang.org/x/crypto v0.27.0/go.mod h1:1Xngt8kV6Dvbssa53Ziq6Eqn0HqbZi5Z6R0ZpwQzt70=
+golang.org/x/crypto v0.32.0/go.mod h1:ZnnJkOaASj8g0AjIduWNlq2NRxL0PlBrbKVyZ6V/Ugc=
-golang.org/x/sync v0.8.0 h1:3NFvSEYkUoMifnESzZl15y791HH1qU2xm6eCJU5ZPXQ=
+golang.org/x/sync v0.10.0 h1:3NQrjDixjgGwUOCaF8w2+VYHv0Ve/vGYSbdkTa98gmQ=
-golang.org/x/sync v0.8.0/go.mod h1:Czt+wKu1gCyEFDUtn0jG5QVvpJ6rzVqr5aXyt9drQfk=
+golang.org/x/sync v0.10.0/go.mod h1:Czt+wKu1gCyEFDUtn0jG5QVvpJ6rzVqr5aXyt9drQfk=
 golang.org/x/sys v0.0.0-20220811171246-fbc7d0a398ab/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.6.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.12.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/sys v0.25.0 h1:r+8e+loiHxRqhXVl6ML1nO3l1+oFoWbnlu2Ehimmi34=
+golang.org/x/sys v0.29.0 h1:TPYlXGxvx1MGTn2GiZDhnjPA9wZzZeGKHHmKhHYvgaU=
-golang.org/x/sys v0.25.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
+golang.org/x/sys v0.29.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
-golang.org/x/text v0.19.0 h1:kTxAhCbGbxhK0IwgSKiMO5awPoDQ0RpfiVYBfK860YM=
+golang.org/x/text v0.21.0 h1:zyQAAkrwaneQ066sspRyJaG9VNi/YJ1NfzcGB3hZ/qo=
-golang.org/x/text v0.19.0/go.mod h1:BuEKDfySbSR4drPmRPG/7iBdf8hvFMuRexcpahXilzY=
+golang.org/x/text v0.21.0/go.mod h1:4IBbMaMmOPCJ8SecivzSH54+73PCFmPWxNTLm+vZkEQ=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
 gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c h1:Hei/4ADfdWqJk1ZMxUNpqntNwaWcugrBjAiHlqqRiVk=
 gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c/go.mod h1:JHkPIbrfpd72SG/EVd6muEfDQjcINNoR0C8j2r3qZ4Q=

gopher/errors.go

@@ -0,0 +1,32 @@
+package gopher
+
+import (
+	"gemini-grc/errors"
+)
+
+// GopherError is an error encountered while
+// visiting a Gopher host, and is only for
+// Gopher errors (item type indicator 3).
+type GopherError struct {
+	Err error
+}
+
+func (e *GopherError) Error() string {
+	return e.Err.Error()
+}
+
+func (e *GopherError) Unwrap() error {
+	return e.Err
+}
+
+func NewGopherError(err error) error {
+	return &GopherError{Err: err}
+}
+
+func IsGopherError(err error) bool {
+	if err == nil {
+		return false
+	}
+	var asError *GopherError
+	return errors.As(err, &asError)
+}

gopher/network.go

@@ -0,0 +1,283 @@
+package gopher
+
+import (
+	"fmt"
+	"io"
+	"net"
+	stdurl "net/url"
+	"regexp"
+	"strings"
+	"time"
+	"unicode/utf8"
+
+	errors2 "gemini-grc/common/errors"
+	"gemini-grc/common/linkList"
+	"gemini-grc/common/snapshot"
+	_url "gemini-grc/common/url"
+	"gemini-grc/config"
+	"gemini-grc/errors"
+	"gemini-grc/logging"
+	"github.com/guregu/null/v5"
+)
+
+// References:
+// RFC 1436 https://www.rfc-editor.org/rfc/rfc1436.html
+
+// The default port for Gopher is 70.
+// Originally Gopher used ASCII or
+// ISO-8859-1, now others use UTF-8.
+// In any case, just converting to UTF-8
+// will work. If not, we bail.
+
+// Here's the complete list of Gopher item type indicators (prefixes):
+//
+// `0` - Plain Text File
+// `1` - Directory/Menu
+// `2` - CSO Phone Book Server
+// `3` - Error Message
+// `4` - BinHexed Macintosh File
+// `5` - DOS Binary Archive
+// `6` - UNIX uuencoded File
+// `7` - Index/Search Server
+// `8` - Telnet Session
+// `9` - Binary File
+// `+` - Mirror/Redundant Server
+// `g` - GIF Image
+// `I` - Image File (non-GIF)
+// `T` - TN3270 Session
+// `i` - Informational Message (menu line)
+// `h` - HTML File
+// `s` - Sound/Music File
+// `d` - Document File
+// `w` - WHOIS Service
+// `;` - Document File with Alternative View
+// `<` - Video File
+// `M` - MIME File (mail message or similar)
+// `:` - Bitmap Image
+// `c` - Calendar File
+// `p` - PostScript File
+
+// The most commonly used ones are `0` (text), `1` (directory), `i` (info), and `3` (error).
+// The original Gopher protocol only specified types 0-9, `+`, `g`, `I`, and `T`.
+// The others were added by various implementations and extensions over time.
+
+// Error methodology:
+// HostError for DNS/network errors
+// GopherError for network/gopher errors
+// NewError for other errors
+// NewFatalError for other fatal errors
+
+func Visit(url string) (*snapshot.Snapshot, error) {
+	s, err := snapshot.SnapshotFromURL(url, false)
+	if err != nil {
+		return nil, err
+	}
+
+	data, err := connectAndGetData(url)
+	if err != nil {
+		logging.LogDebug("Error: %s", err.Error())
+		if IsGopherError(err) || errors2.IsHostError(err) {
+			s.Error = null.StringFrom(err.Error())
+			return s, nil
+		}
+		return nil, err
+	}
+
+	isValidUTF8 := utf8.ValidString(string(data))
+	if isValidUTF8 {
+		s.GemText = null.StringFrom(removeNullChars(string(data)))
+	} else {
+		s.Data = null.ValueFrom(data)
+	}
+
+	if !isValidUTF8 {
+		return s, nil
+	}
+
+	responseError := checkForError(string(data))
+	if responseError != nil {
+		s.Error = null.StringFrom(responseError.Error())
+		return s, nil
+	}
+
+	links := getGopherPageLinks(string(data))
+	linkURLs := linkList.LinkList(make([]_url.URL, len(links)))
+	for i, link := range links {
+		linkURL, err := _url.ParseURL(link, "", true)
+		if err == nil {
+			linkURLs[i] = *linkURL
+		}
+	}
+	if len(links) != 0 {
+		s.Links = null.ValueFrom(linkURLs)
+	}
+
+	return s, nil
+}
+
+func connectAndGetData(url string) ([]byte, error) {
+	parsedURL, err := stdurl.Parse(url)
+	if err != nil {
+		return nil, errors.NewError(err)
+	}
+
+	hostname := parsedURL.Hostname()
+	port := parsedURL.Port()
+	if port == "" {
+		port = "70"
+	}
+	host := fmt.Sprintf("%s:%s", hostname, port)
+	timeoutDuration := time.Duration(config.CONFIG.ResponseTimeout) * time.Second
+	// Establish the underlying TCP connection.
+	dialer := &net.Dialer{
+		Timeout: timeoutDuration,
+	}
+	logging.LogDebug("Dialing %s", host)
+	conn, err := dialer.Dial("tcp", host)
+	if err != nil {
+		return nil, errors2.NewHostError(err)
+	}
+	// Make sure we always close the connection.
+	defer func() {
+		_ = conn.Close()
+	}()
+
+	// Set read and write timeouts on the TCP connection.
+	err = conn.SetReadDeadline(time.Now().Add(timeoutDuration))
+	if err != nil {
+		return nil, errors2.NewHostError(err)
+	}
+	err = conn.SetWriteDeadline(time.Now().Add(timeoutDuration))
+	if err != nil {
+		return nil, errors2.NewHostError(err)
+	}
+
+	// We read `buf`-sized chunks and add data to `data`.
+	buf := make([]byte, 4096)
+	var data []byte
+
+	// Send Gopher request to trigger server response.
+	payload := constructPayloadFromPath(parsedURL.Path)
+	_, err = conn.Write([]byte(fmt.Sprintf("%s\r\n", payload)))
+	if err != nil {
+		return nil, errors2.NewHostError(err)
+	}
+	// Read response bytes in len(buf) byte chunks
+	for {
+		n, err := conn.Read(buf)
+		if n > 0 {
+			data = append(data, buf[:n]...)
+		}
+		if err != nil {
+			if errors.Is(err, io.EOF) {
+				break
+			}
+			return nil, errors2.NewHostError(err)
+		}
+		if len(data) > config.CONFIG.MaxResponseSize {
+			return nil, errors2.NewHostError(fmt.Errorf("response exceeded max"))
+		}
+	}
+	logging.LogDebug("Got %d bytes", len(data))
+	return data, nil
+}
+
+func constructPayloadFromPath(urlpath string) string {
+	// remove Gopher item type in URL from payload, if one.
+	re := regexp.MustCompile(`^/[\w]/.*`)
+	payloadWithoutItemtype := urlpath
+	if re.Match([]byte(urlpath)) {
+		payloadWithoutItemtype = strings.Join(strings.Split(urlpath, "/")[2:], "/")
+	}
+	if !strings.HasPrefix(payloadWithoutItemtype, "/") {
+		payloadWithoutItemtype = fmt.Sprintf("/%s", payloadWithoutItemtype)
+	}
+	return payloadWithoutItemtype
+}
+
+func checkForError(utfData string) error {
+	lines := strings.Split(strings.TrimSpace(utfData), "\n")
+	var firstLine string
+	if len(lines) > 0 {
+		firstLine = lines[0]
+	} else {
+		return nil
+	}
+	if strings.HasPrefix(firstLine, "3") {
+		split := strings.Split(firstLine, "\t")
+		return NewGopherError(fmt.Errorf("gopher error: %s", strings.TrimSpace(split[0])))
+	}
+	return nil
+}
+
+func getGopherPageLinks(content string) []string {
+	var links []string
+
+	lines := strings.Split(strings.TrimSpace(content), "\n")
+
+	for _, line := range lines {
+		if line == "" || line == "." {
+			continue
+		}
+
+		if len(line) < 1 {
+			continue
+		}
+
+		itemType := line[0]
+		if itemType == 'i' {
+			continue
+		}
+
+		parts := strings.SplitN(line[1:], "\t", 4)
+		if len(parts) < 3 {
+			continue
+		}
+
+		selector := strings.TrimSpace(parts[1])
+		host := strings.TrimSpace(parts[2])
+
+		if host == "" {
+			continue
+		}
+
+		// Handle HTML links first
+		if itemType == 'h' && strings.HasPrefix(selector, "URL:") {
+			if url := strings.TrimSpace(selector[4:]); url != "" {
+				links = append(links, url)
+			}
+			continue
+		}
+
+		// For gopher links, build URL carefully
+		var url strings.Builder
+
+		// Protocol and host:port
+		url.WriteString("gopher://")
+		url.WriteString(host)
+		url.WriteString(":")
+		if len(parts) > 3 && strings.TrimSpace(parts[3]) != "" {
+			url.WriteString(strings.TrimSpace(parts[3]))
+		} else {
+			url.WriteString("70")
+		}
+
+		// Path: always /type + selector
+		url.WriteString("/")
+		url.WriteString(string(itemType))
+		if strings.HasPrefix(selector, "/") {
+			url.WriteString(selector)
+		} else {
+			url.WriteString("/" + selector)
+		}
+
+		links = append(links, url.String())
+	}
+
+	return links
+}
+
+func removeNullChars(input string) string {
+	// Replace all null characters with an empty string
+	return strings.ReplaceAll(input, "\u0000", "")
+}

gopher/network_test.go

@@ -0,0 +1,298 @@
+package gopher
+
+import (
+	"net"
+	"testing"
+
+	"gemini-grc/common/errors"
+	"gemini-grc/config"
+	"github.com/stretchr/testify/assert"
+)
+
+func TestConstructPayloadFromPath(t *testing.T) {
+	tests := []struct {
+		name     string
+		input    string
+		expected string
+	}{
+		{
+			name:     "Path with Gopher item type",
+			input:    "/1/path/to/resource",
+			expected: "/path/to/resource",
+		},
+		{
+			name:     "Path with different item type",
+			input:    "/0/another/path",
+			expected: "/another/path",
+		},
+		{
+			name:     "Path without item type but with leading slash",
+			input:    "/simple/path",
+			expected: "/simple/path",
+		},
+		{
+			name:     "Path without item type and without leading slash",
+			input:    "no/leading/slash",
+			expected: "/no/leading/slash",
+		},
+		{
+			name:     "Empty path",
+			input:    "",
+			expected: "/",
+		},
+		{
+			name:     "Single character item type",
+			input:    "/h/homepage",
+			expected: "/homepage",
+		},
+		{
+			name:     "Single slash",
+			input:    "/",
+			expected: "/",
+		},
+		{
+			name:     "Item type-looking path",
+			input:    "/1",
+			expected: "/1",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result := constructPayloadFromPath(tt.input)
+			if result != tt.expected {
+				t.Errorf("constructPayloadFromPath(%q) = %q, want %q",
+					tt.input, result, tt.expected)
+			}
+		})
+	}
+}
+
+func TestParseLinks(t *testing.T) {
+	tests := []struct {
+		name       string
+		currentURL string
+		input      string
+		want       int // number of expected links
+		wantErr    bool
+	}{
+		{
+			name:       "Empty input",
+			currentURL: "gopher://example.com:70",
+			input:      "",
+			want:       0,
+			wantErr:    false,
+		},
+		{
+			name:       "Single directory link",
+			currentURL: "gopher://example.com:70",
+			input:      "1About Us\t/about\texample.com",
+			want:       1,
+			wantErr:    false,
+		},
+		{
+			name:       "Single text file link",
+			currentURL: "gopher://example.com:70",
+			input:      "0README\t/readme.txt\texample.com",
+			want:       1,
+			wantErr:    false,
+		},
+		{
+			name:       "Multiple links of different types",
+			currentURL: "gopher://example.com:70",
+			input:      "1About Us\t/about\texample.com\n0README\t/readme.txt\texample.com\n1Contact\t/contact\texample.com",
+			want:       3,
+			wantErr:    false,
+		},
+		{
+			name:       "Ignore non-linkable types",
+			currentURL: "gopher://example.com:70",
+			input:      "iInfo line\t/info\texample.com\n1Directory\t/dir\texample.com\n0Text\t/text.txt\texample.com",
+			want:       2,
+			wantErr:    false,
+		},
+		{
+			name:       "Malformed lines",
+			currentURL: "gopher://example.com:70",
+			input:      "1Incomplete line\n0No tabs\n1Missing parts\t",
+			want:       0,
+			wantErr:    false,
+		},
+		{
+			name:       "Mixed valid and invalid lines",
+			currentURL: "gopher://example.com:70",
+			input:      "1Valid link\t/valid\texample.com\n1Incomplete\t\n0Text file\t/text.txt\texample.com\n1Another valid\t/another\texample.com",
+			want:       3,
+			wantErr:    false,
+		},
+		{
+			name:       "Absolute URLs",
+			currentURL: "gopher://example.com:70",
+			input:      "1External link\tgopher://external.com/path\texternal.com\n0Document\tgopher://other.com/doc.txt\tother.com",
+			want:       2,
+			wantErr:    false,
+		},
+		{
+			name:       "With whitespace",
+			currentURL: "gopher://example.com:70",
+			input:      "  1Padded line  \t/padded\texample.com\n0Text file  \t/doc.txt\texample.com",
+			want:       2,
+			wantErr:    false,
+		},
+		{
+			name:       "Special characters in paths",
+			currentURL: "gopher://example.com:70",
+			input:      "1Special chars\t/path with spaces\texample.com\n0Doc\t/über/päth.txt\texample.com",
+			want:       2,
+			wantErr:    false,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got := getGopherPageLinks(tt.input)
+			assert.Equal(t, tt.want, len(got), "expected %d links, got %d", tt.want, len(got))
+		})
+	}
+}
+
+func TestCheckForError(t *testing.T) {
+	tests := []struct {
+		name        string
+		input       string
+		wantError   bool
+		errorPrefix string
+	}{
+		{
+			name:        "No error",
+			input:       "1Directory\t/dir\texample.com\n0Text\t/text.txt\texample.com",
+			wantError:   false,
+			errorPrefix: "",
+		},
+		{
+			name:        "Simple error message",
+			input:       "3Error: File not found\t\texample.com",
+			wantError:   true,
+			errorPrefix: "gopher error: 3Error: File not found",
+		},
+		{
+			name:        "Error with multiple tabs",
+			input:       "3File not found\t/error\texample.com\t70",
+			wantError:   true,
+			errorPrefix: "gopher error: 3File not found",
+		},
+		{
+			name: "Error among valid entries",
+			input: `1Welcome\t/welcome\texample.com
+3Access denied\t\texample.com
+0README\t/readme.txt\texample.com`,
+			wantError:   false,
+			errorPrefix: "",
+		},
+		{
+			name:        "Error with no tabs",
+			input:       "3Server is down for maintenance",
+			wantError:   true,
+			errorPrefix: "gopher error: 3Server is down for maintenance",
+		},
+		{
+			name: "Multiple errors (should return first)",
+			input: `3First error\t\texample.com
+3Second error\t\texample.com`,
+			wantError:   true,
+			errorPrefix: "gopher error: 3First error",
+		},
+		{
+			name:        "Error with whitespace",
+			input:       "  3  Error with spaces  \t\texample.com",
+			wantError:   true,
+			errorPrefix: "gopher error: 3  Error with spaces",
+		},
+		{
+			name:        "Empty input",
+			input:       "",
+			wantError:   false,
+			errorPrefix: "",
+		},
+		{
+			name:        "Just newlines",
+			input:       "\n\n\n",
+			wantError:   false,
+			errorPrefix: "",
+		},
+		{
+			name: "Error after empty lines",
+			input: `
+
+3Error after empty lines\t\texample.com`,
+			wantError:   true,
+			errorPrefix: "gopher error: 3Error after empty lines",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			err := checkForError(tt.input)
+
+			if !tt.wantError {
+				assert.NoError(t, err)
+				return
+			}
+
+			assert.Error(t, err)
+			assert.Contains(t, err.Error(), tt.errorPrefix)
+		})
+	}
+}
+
+func TestConnectAndGetDataTimeout(t *testing.T) {
+	// Start a test server that doesn't respond
+	listener, err := net.Listen("tcp", "localhost:0")
+	if err != nil {
+		t.Fatalf("Failed to start listener: %v", err)
+	}
+	defer listener.Close()
+
+	// Accept the connection but don't respond
+	go func() {
+		conn, err := listener.Accept()
+		if err != nil {
+			t.Logf("Failed to accept connection: %v", err)
+			return
+		}
+		defer conn.Close()
+
+		// Keep the connection open without sending any data to simulate a timeout
+		select {}
+	}()
+
+	// Construct the URL of our test server
+	address := listener.Addr().String()
+	testURL := "gopher://" + address + "/testpath"
+
+	// Save original config values
+	originalTimeout := config.CONFIG.ResponseTimeout
+	originalMaxSize := config.CONFIG.MaxResponseSize
+
+	// Set test config values
+	config.CONFIG.ResponseTimeout = 1    // Set a very short timeout for this test
+	config.CONFIG.MaxResponseSize = 1024 // Just for consistency, we won't reach this
+
+	// Test the function
+	_, err = connectAndGetData(testURL)
+
+	// Reset config values
+	config.CONFIG.ResponseTimeout = originalTimeout
+	config.CONFIG.MaxResponseSize = originalMaxSize
+
+	// Check if the error is due to timeout
+	if err == nil {
+		t.Error("Expected an error due to timeout, but got no error")
+	} else if !errors.IsHostError(err) {
+		t.Errorf("Expected a HostError, but got: %v", err)
+	} else {
+		// Here you might want to check if the specific error message contains 'timeout'
+		// However, since we don't have the exact error string, we're checking the type
+		t.Logf("Successfully timed out: %v", err)
+	}
+}

hostPool/hostPool.go

@@ -11,44 +11,39 @@ var hostPool = HostPool{hostnames: make(map[string]struct{})} //nolint:gocheckno
 
 type HostPool struct {
 	hostnames map[string]struct{}
-	Lock      sync.RWMutex
+	lock      sync.RWMutex
 }
 
-func (p *HostPool) Add(key string) {
+//func (p *HostPool) add(key string) {
-	p.Lock.Lock()
+//	p.lock.Lock()
-	defer p.Lock.Unlock()
+//	defer p.lock.Unlock()
-	p.hostnames[key] = struct{}{}
+//	p.hostnames[key] = struct{}{}
-}
+//}
+//
+//func (p *HostPool) has(key string) bool {
+//	p.lock.RLock()
+//	defer p.lock.RUnlock()
+//	_, ok := p.hostnames[key]
+//	return ok
+//}
 
-func (p *HostPool) Get(key string) bool {
+func RemoveHostFromPool(key string) {
-	p.Lock.RLock()
+	hostPool.lock.Lock()
-	defer p.Lock.RUnlock()
+	defer hostPool.lock.Unlock()
-	_, ok := p.hostnames[key]
+	delete(hostPool.hostnames, key)
-	return ok
-}
-
-func (p *HostPool) Delete(key string) {
-	p.Lock.Lock()
-	defer p.Lock.Unlock()
-	delete(p.hostnames, key)
 }
 
 func AddHostToHostPool(key string) {
 	for {
-		// Sleep until the host doesn't exist in pool,
+		hostPool.lock.Lock()
-		// then add it.
+		_, exists := hostPool.hostnames[key]
-		if hostPool.Get(key) {
+		if !exists {
-			time.Sleep(1 * time.Second) // Avoid flood-retrying
+			hostPool.hostnames[key] = struct{}{}
-			logging.LogInfo("Waiting to add %s to pool...", key)
+			hostPool.lock.Unlock()
-		} else {
-			hostPool.Add(key)
 			return
 		}
-	}
+		hostPool.lock.Unlock()
-}
+		time.Sleep(1 * time.Second)
-
+		logging.LogInfo("Waiting to add %s to pool...", key)
-func RemoveHostFromHostPool(key string) {
-	if hostPool.Get(key) {
-		hostPool.Delete(key)
 	}
 }

main.go

@@ -1,13 +1,16 @@
 package main
 
 import (
-	main2 "gemini-grc/db"
+	"fmt"
 	"os"
 	"os/signal"
 	"syscall"
 
+	"gemini-grc/common"
+	"gemini-grc/common/blackList"
 	"gemini-grc/config"
-	"gemini-grc/gemini"
+	"gemini-grc/db"
+	"gemini-grc/errors"
 	"gemini-grc/logging"
 	"github.com/jmoiron/sqlx"
 	"github.com/rs/zerolog"
@@ -19,39 +22,61 @@ func main() {
 	zerolog.TimeFieldFormat = zerolog.TimeFormatUnix
 	zerolog.SetGlobalLevel(config.CONFIG.LogLevel)
 	zlog.Logger = zlog.Output(zerolog.ConsoleWriter{Out: os.Stderr, TimeFormat: "[2006-01-02 15:04:05]"})
-	if err := runApp(); err != nil {
+	err := runApp()
-		logging.LogError("Application error: %w", err)
+	if err != nil {
+		var asErr *errors.Error
+		if errors.As(err, &asErr) {
+			logging.LogError("Unexpected error: %v", err)
+			_, _ = fmt.Fprintf(os.Stderr, "Unexpected error: %v", err)
+		} else {
+			logging.LogError("Unexpected error: %v", err)
+		}
 		os.Exit(1)
 	}
 }
 
-func runApp() error {
+func runApp() (err error) {
-	logging.LogInfo("Starting up. Press Ctrl+C to exit")
+	logging.LogInfo("gemcrawl %s starting up. Press Ctrl+C to exit", common.VERSION)
 	signals := make(chan os.Signal, 1)
 	signal.Notify(signals, syscall.SIGINT, syscall.SIGTERM)
 
-	db := main2.ConnectToDB()
+	_db, err := db.ConnectToDB()
+	if err != nil {
+		return err
+	}
 
 	defer func(db *sqlx.DB) {
-		err := db.Close()
+		_ = db.Close()
-		if err != nil {
+	}(_db)
-			// TODO properly log & hangle error
-			panic(err)
-		}
-	}(db)
 
-	gemini.LoadBlacklist()
+	err = blackList.LoadBlacklist()
+	if err != nil {
+		return err
+	}
+
+	common.StatusChan = make(chan common.WorkerStatus, config.CONFIG.NumOfWorkers)
+	common.ErrorsChan = make(chan error, config.CONFIG.NumOfWorkers)
 
 	// If there's an argument, visit this
 	// URL only and don't spawn other workers
 	if len(os.Args) > 1 {
 		url := os.Args[1]
-		go gemini.RunWorkerWithTx(0, db, &url)
+		err = common.CrawlOneURL(_db, &url)
-	} else {
+		return err
-		go gemini.SpawnWorkers(config.CONFIG.NumOfWorkers, db)
 	}
 
-	<-signals
+	go common.SpawnWorkers(config.CONFIG.NumOfWorkers, _db)
-	logging.LogWarn("Received SIGINT or SIGTERM signal, exiting")
+
-	return nil
+	for {
+		select {
+		case <-signals:
+			logging.LogWarn("Received SIGINT or SIGTERM signal, exiting")
+			return nil
+		case err := <-common.ErrorsChan:
+			if errors.IsFatal(err) {
+				return err
+			}
+			logging.LogError("%s", fmt.Sprintf("%v", err))
+		}
+	}
 }

util/util.go

@@ -5,11 +5,12 @@ import (
 	"encoding/json"
 	"fmt"
 	"math/big"
+	"regexp"
 	"runtime/debug"
 )
 
 func PrintStackAndPanic(err error) {
-	fmt.Printf("Error %s Stack trace:\n%s", err, debug.Stack())
+	fmt.Printf("PANIC Error %s Stack trace:\n%s", err, debug.Stack())
 	panic("PANIC")
 }
 
@@ -34,3 +35,10 @@ func PrettyJson(data string) string {
 	marshalled, _ := json.MarshalIndent(data, "", "  ")
 	return fmt.Sprintf("%s\n", marshalled)
 }
+
+// GetLinesMatchingRegex returns all lines that match given regex
+func GetLinesMatchingRegex(input string, pattern string) []string {
+	re := regexp.MustCompile(pattern)
+	matches := re.FindAllString(input, -1)
+	return matches
+}