Reorganize code for more granular imports

This commit is contained in:
2025-02-26 10:34:25 +02:00
parent 8350e106d6
commit ca008b0796
23 changed files with 1549 additions and 1232 deletions

View File

@@ -1,252 +0,0 @@
package common_test
import (
"reflect"
"testing"
"gemini-grc/common"
)
func TestParseURL(t *testing.T) {
t.Parallel()
input := "gemini://caolan.uk/cgi-bin/weather.py/wxfcs/3162"
parsed, err := common.ParseURL(input, "", true)
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)
}
}

View File

@@ -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)
}

13
common/shared.go Normal file
View File

@@ -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"

View File

@@ -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
}

View File

@@ -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
}

View File

@@ -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 := 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 {
return fmt.Sprintf("%s://%s%s", u.Protocol, u.Hostname, u.Path)
if IsGeminiUrl(u.String()) {
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
}
@@ -55,30 +63,43 @@ func (u URL) Value() (driver.Value, error) {
return u.Full, nil
}
func IsGeminiUrl(url string) bool {
return strings.HasPrefix(url, "gemini://")
}
func IsGopherURL(s string) bool {
return strings.HasPrefix(s, "gopher://")
}
func ParseURL(input string, descr string, normalize bool) (*URL, error) {
var u *url.URL
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, 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)
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
@@ -113,7 +134,7 @@ func DeriveAbsoluteURL(currentURL URL, input string) (*URL, error) {
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
@@ -124,7 +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
@@ -135,7 +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":
@@ -144,6 +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()
}
}
@@ -152,7 +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 = "/"
@@ -162,20 +191,25 @@ func NormalizeURL(rawURL string) (*url.URL, error) {
}
}
// Properly escape the path
// First split on '/' to avoid escaping them
// Properly escape the path, but only for unescaped parts
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 = ""
}
@@ -188,7 +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
@@ -202,18 +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
}

420
common/url/url_test.go Normal file
View File

@@ -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)
}
}

320
common/worker.go Normal file
View File

@@ -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
}

View File

@@ -1,19 +1,35 @@
package gemini
package common
import (
"fmt"
"strings"
"gemini-grc/config"
)
type WorkerStatus struct {
id int
status string
ID int
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
statuses[update.id] = update.status
// Update the 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())
}
}