fix: accept any d-tag identifier; percent-encode in URLs

NIP-01 places no restriction on d tag characters and NIP-34 only
recommends kebab-case without mandating it. Rejecting identifiers with
whitespace or other URL-unsafe characters was therefore overly strict.

The correct approach (per NIP-34 PR #2312 and GRASP-01) is to store
identifiers verbatim on disk and percent-encode them when constructing
URLs. The previous commit already handled the incoming direction
(percent-decoding URL paths before filesystem lookup); this commit
handles the outgoing direction and removes the validation restriction.

Changes:
- validate_identifier: drop whitespace rejection; only reject chars
  that are unsafe as filesystem directory names (/, \, null, . / ..)
- git/mod.rs: add percent_encode() alongside percent_decode()
- landing.rs: percent-encode identifier in nostr:// clone URL and
  gitworkshop link (also fixes a pre-existing bug where the clone URL
  displayed literal '{npub}' / '{identifier}' instead of the values)
This commit is contained in:
DanConwayDev
2026-04-10 16:42:35 +00:00
parent 2d74b9ca69
commit dfd20a39a7
4 changed files with 81 additions and 39 deletions
+4
View File
@@ -7,6 +7,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
## [Unreleased]
### Fixed
- Repository identifiers containing characters that require percent-encoding in URLs (e.g. spaces, emoji) are now accepted and served correctly. NIP-01 places no restriction on `d` tag values and NIP-34 only recommends kebab-case without mandating it, so rejecting non-kebab identifiers was overly strict. Identifiers are stored verbatim on disk and percent-encoded when used in URLs, per the `nostr://` clone URL spec formalised in [NIP-34 PR #2312](https://github.com/nostr-protocol/nips/pull/2312) and the GRASP-01 HTTP path spec. The landing page clone URL now also correctly percent-encodes the identifier.
### Changed
- Remove arbitrary default max connections limit; when `NGIT_MAX_CONNECTIONS` is unset the relay imposes no connection cap, deferring to OS fd limits and infrastructure controls
+49 -2
View File
@@ -451,6 +451,29 @@ pub fn get_repository_head(repo_path: &Path) -> Option<String> {
}
}
/// Percent-encode a string for use as a URL path segment (RFC 3986 §2.1).
///
/// Encodes all bytes that are not unreserved characters (`A-Z a-z 0-9 - _ . ~`).
/// This is suitable for encoding a repository identifier in a `nostr://` URL or
/// an HTTP path component such as `/<npub>/<encoded-identifier>.git`.
pub fn percent_encode(s: &str) -> String {
let mut out = String::with_capacity(s.len());
for byte in s.bytes() {
match byte {
// RFC 3986 unreserved characters — never encoded
b'A'..=b'Z' | b'a'..=b'z' | b'0'..=b'9' | b'-' | b'_' | b'.' | b'~' => {
out.push(byte as char);
}
_ => {
out.push('%');
out.push(char::from_digit((byte >> 4) as u32, 16).unwrap().to_ascii_uppercase());
out.push(char::from_digit((byte & 0xf) as u32, 16).unwrap().to_ascii_uppercase());
}
}
}
out
}
/// Decode percent-encoded characters in a URL path component.
///
/// Handles `%XX` sequences (e.g. `%20` → space). Invalid sequences are left as-is.
@@ -481,8 +504,8 @@ pub fn percent_decode(s: &str) -> String {
///
/// The identifier component is percent-decoded so that URLs like
/// `/npub1.../my%20repo.git/info/refs` resolve to the filesystem path
/// `my repo.git` (though such identifiers should be rejected at announcement
/// validation time — see `validate_announcement`).
/// `my repo.git`. Per NIP-34 and GRASP-01, identifiers MUST be percent-encoded
/// in URLs; they are stored verbatim on disk.
///
/// Returns (npub, identifier, subpath) where subpath is the part after .git/
/// and identifier has been percent-decoded.
@@ -671,6 +694,30 @@ mod tests {
assert_eq!(percent_decode("foo%zz"), "foo%zz");
}
#[test]
fn test_percent_encode_basic() {
assert_eq!(percent_encode("my-repo"), "my-repo");
assert_eq!(percent_encode("my_repo"), "my_repo");
assert_eq!(percent_encode("repo123"), "repo123");
assert_eq!(percent_encode("hello world"), "hello%20world");
assert_eq!(percent_encode("kuboslopp by Shakespeare"), "kuboslopp%20by%20Shakespeare");
}
#[test]
fn test_percent_encode_special_chars() {
assert_eq!(percent_encode("a/b"), "a%2Fb");
assert_eq!(percent_encode("a\\b"), "a%5Cb");
assert_eq!(percent_encode("a b\tc"), "a%20b%09c");
}
#[test]
fn test_percent_encode_decode_roundtrip() {
let identifiers = ["my-repo", "my repo", "kuboslopp by Shakespeare", "a/b", "foo\0bar"];
for id in &identifiers {
assert_eq!(percent_decode(&percent_encode(id)), *id);
}
}
#[test]
fn test_commit_exists_nonexistent() {
let (_temp_dir, repo_path) = create_test_repo();
+4 -2
View File
@@ -2,6 +2,7 @@
///
/// Generates HTML landing page for the Nostr relay.
use crate::config::Config;
use crate::git::percent_encode;
use crate::http::nip11::RelayInformationDocument;
use std::collections::HashMap;
@@ -847,7 +848,7 @@ pub fn get_repo_html(config: &Config, npub: &str, identifier: &str) -> String {
<div class="card">
<div class="clone-box">
<div class="clone-line"><span class="cmd">curl -Ls https://ngit.dev/install.sh | bash</span></div>
<div class="clone-line"><span class="cmd">git clone</span> <span class="url" id="nostr-clone-url">nostr://{{npub}}/<span id="relayref"></span>/{{identifier}}</span></div>
<div class="clone-line"><span class="cmd">git clone</span> <span class="url" id="nostr-clone-url">nostr://{npub}/<span id="relayref"></span>/{encoded_identifier}</span></div>
</div>
</div>
</div>
@@ -867,7 +868,7 @@ pub fn get_repo_html(config: &Config, npub: &str, identifier: &str) -> String {
// Construct gitworkshop link: gitworkshop.dev/npub/relayref/identifier
const gitworkshopLink = document.getElementById('gitworkshop-link');
gitworkshopLink.setAttribute('href', 'https://gitworkshop.dev/{npub}/' + relayref + '/{identifier}');
gitworkshopLink.setAttribute('href', 'https://gitworkshop.dev/{npub}/' + relayref + '/{encoded_identifier}');
// Set footer domain
var footerDomain = document.getElementById('footer-domain');
@@ -882,6 +883,7 @@ pub fn get_repo_html(config: &Config, npub: &str, identifier: &str) -> String {
relay_name = config.relay_name(),
npub = npub,
identifier = identifier,
encoded_identifier = percent_encode(identifier),
version = get_version(),
theme_toggle = get_theme_toggle_html(),
theme_script = get_theme_script(),
+24 -35
View File
@@ -361,19 +361,20 @@ impl RepositoryState {
}
}
/// Validate that a repository identifier is safe for use as a filesystem path component
/// and as a URL path segment without percent-encoding.
/// Validate that a repository identifier is safe for use as a filesystem path component.
///
/// Rejects identifiers that:
/// - Are empty
/// - Contain path separators (`/`, `\`)
/// - Contain null bytes
/// - Contain whitespace (spaces, tabs, newlines, etc.) — these require percent-encoding
/// in URLs and cause a mismatch between the stored path and the URL-decoded request
/// - Are `.` or `..` (directory traversal)
/// NIP-34 places no restriction on `d` tag characters beyond NIP-01 (any string).
/// Identifiers are stored on disk verbatim and percent-encoded when used in URLs
/// (per NIP-34 `nostr://` spec and GRASP-01 HTTP path spec). This function only
/// rejects identifiers that cannot safely be stored as a filesystem directory name:
///
/// NIP-34 recommends kebab-case identifiers; this function enforces the minimum
/// safety constraints needed for correct filesystem and HTTP serving behaviour.
/// - Empty string
/// - Path separators (`/`, `\`) — would escape the repository directory
/// - Null bytes — rejected by most filesystems
/// - `.` or `..` — reserved path components (directory traversal)
///
/// Whitespace and other characters that require percent-encoding in URLs are
/// explicitly allowed here; callers are responsible for encoding them in URLs.
pub fn validate_identifier(identifier: &str) -> Result<(), String> {
if identifier.is_empty() {
return Err("identifier must not be empty".to_string());
@@ -397,12 +398,6 @@ pub fn validate_identifier(identifier: &str) -> Result<(), String> {
identifier
));
}
if ch.is_whitespace() {
return Err(format!(
"identifier '{}' contains whitespace — use hyphens instead (e.g. 'my-repo')",
identifier
));
}
}
Ok(())
}
@@ -1573,6 +1568,11 @@ mod tests {
assert!(validate_identifier("my_repo").is_ok());
assert!(validate_identifier("repo123").is_ok());
assert!(validate_identifier("kuboslopp").is_ok());
// Whitespace is valid — identifiers are stored verbatim on disk and
// percent-encoded when used in URLs (NIP-34 / GRASP-01)
assert!(validate_identifier("kuboslopp by Shakespeare").is_ok());
assert!(validate_identifier("my\trepo").is_ok());
assert!(validate_identifier("my repo").is_ok());
}
#[test]
@@ -1593,19 +1593,12 @@ mod tests {
}
#[test]
fn test_validate_identifier_rejects_whitespace() {
assert!(validate_identifier("kuboslopp by Shakespeare").is_err());
assert!(validate_identifier("my\trepo").is_err());
assert!(validate_identifier("my\nrepo").is_err());
}
#[test]
fn test_validate_announcement_rejects_identifier_with_spaces() {
fn test_validate_announcement_accepts_identifier_with_spaces() {
use crate::config::Config;
use crate::nostr::policy::AnnouncementResult;
let keys = create_test_keys();
// Identifier contains spaces — should be rejected regardless of clone/relay tags
// Identifier contains spaces — valid per NIP-34; must be percent-encoded in URLs
let event = create_announcement_event(
&keys,
"kuboslopp by Shakespeare",
@@ -1618,14 +1611,10 @@ mod tests {
..Config::for_testing()
};
let result = validate_announcement(&event, &config);
if let AnnouncementResult::Reject(reason) = result {
assert!(
reason.contains("whitespace") || reason.contains("identifier"),
"unexpected rejection reason: {}",
reason
);
} else {
panic!("Expected Reject for identifier with spaces, got {:?}", result);
}
assert!(
matches!(result, AnnouncementResult::Accept),
"Expected Accept for identifier with spaces, got {:?}",
result
);
}
}