validator refactor progress

This commit is contained in:
2026-03-03 00:13:37 -05:00
parent e14f53e7d9
commit 3898c43742
81 changed files with 6331 additions and 7934 deletions

12
src/database/enum.rs Normal file
View File

@ -0,0 +1,12 @@
use crate::database::schema::Schema;
use serde::{Deserialize, Serialize};
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
#[serde(default)]
pub struct Enum {
pub name: String,
pub module: String,
pub source: String,
pub values: Vec<String>,
pub schemas: Vec<Schema>,
}

875
src/database/formats.rs Normal file
View File

@ -0,0 +1,875 @@
use std::{
collections::HashMap,
error::Error,
net::{Ipv4Addr, Ipv6Addr},
};
use lazy_static::lazy_static;
use percent_encoding::percent_decode_str;
use serde_json::Value;
use url::Url;
// use crate::ecma; // Assuming ecma is not yet available, stubbing regex for now
/// Defines format for `format` keyword.
#[derive(Clone, Copy)]
pub struct Format {
/// Name of the format
pub name: &'static str,
/// validates given value.
pub func: fn(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>>, // Ensure thread safety if needed
}
lazy_static! {
pub(crate) static ref FORMATS: HashMap<&'static str, Format> = {
let mut m = HashMap::<&'static str, Format>::new();
// Helper to register formats
let mut register = |name, func| m.insert(name, Format { name, func });
// register("regex", validate_regex); // Stubbed
register("ipv4", validate_ipv4);
register("ipv6", validate_ipv6);
register("hostname", validate_hostname);
register("idn-hostname", validate_idn_hostname);
register("email", validate_email);
register("idn-email", validate_idn_email);
register("date", validate_date);
register("time", validate_time);
register("date-time", validate_date_time);
register("duration", validate_duration);
register("period", validate_period);
register("json-pointer", validate_json_pointer);
register("relative-json-pointer", validate_relative_json_pointer);
register("uuid", validate_uuid);
register("uri", validate_uri);
register("iri", validate_iri);
register("uri-reference", validate_uri_reference);
register("iri-reference", validate_iri_reference);
register("uri-template", validate_uri_template);
m
};
}
/*
fn validate_regex(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
// ecma::convert(s).map(|_| ())
Ok(())
}
*/
fn validate_ipv4(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
s.parse::<Ipv4Addr>()?;
Ok(())
}
fn validate_ipv6(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
s.parse::<Ipv6Addr>()?;
Ok(())
}
fn validate_date(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
check_date(s)?;
Ok(())
}
fn matches_char(s: &str, index: usize, ch: char) -> bool {
s.is_char_boundary(index) && s[index..].starts_with(ch)
}
// see https://datatracker.ietf.org/doc/html/rfc3339#section-5.6
fn check_date(s: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
// yyyy-mm-dd
if s.len() != 10 {
Err("must be 10 characters long")?;
}
if !matches_char(s, 4, '-') || !matches_char(s, 7, '-') {
Err("missing hyphen in correct place")?;
}
let mut ymd = s.splitn(3, '-').filter_map(|t| t.parse::<usize>().ok());
let (Some(y), Some(m), Some(d)) = (ymd.next(), ymd.next(), ymd.next()) else {
Err("non-positive year/month/day")?
};
if !matches!(m, 1..=12) {
Err(format!("{m} months in year"))?;
}
if !matches!(d, 1..=31) {
Err(format!("{d} days in month"))?;
}
match m {
2 => {
let mut feb_days = 28;
if y % 4 == 0 && (y % 100 != 0 || y % 400 == 0) {
feb_days += 1; // leap year
};
if d > feb_days {
Err(format!("february has {feb_days} days only"))?;
}
}
4 | 6 | 9 | 11 => {
if d > 30 {
Err("month has 30 days only")?;
}
}
_ => {}
}
Ok(())
}
fn validate_time(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
check_time(s)
}
fn check_time(mut str: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
// min: hh:mm:ssZ
if str.len() < 9 {
Err("less than 9 characters long")?
}
if !matches_char(str, 2, ':') || !matches_char(str, 5, ':') {
Err("missing colon in correct place")?
}
// parse hh:mm:ss
if !str.is_char_boundary(8) {
Err("contains non-ascii char")?
}
let mut hms = (str[..8])
.splitn(3, ':')
.filter_map(|t| t.parse::<usize>().ok());
let (Some(mut h), Some(mut m), Some(s)) = (hms.next(), hms.next(), hms.next()) else {
Err("non-positive hour/min/sec")?
};
if h > 23 || m > 59 || s > 60 {
Err("hour/min/sec out of range")?
}
str = &str[8..];
// parse sec-frac if present
if let Some(rem) = str.strip_prefix('.') {
let n_digits = rem.chars().take_while(char::is_ascii_digit).count();
if n_digits == 0 {
Err("no digits in second fraction")?;
}
str = &rem[n_digits..];
}
if str != "z" && str != "Z" {
// parse time-numoffset
if str.len() != 6 {
Err("offset must be 6 characters long")?;
}
let sign: isize = match str.chars().next() {
Some('+') => -1,
Some('-') => 1,
_ => return Err("offset must begin with plus/minus")?,
};
str = &str[1..];
if !matches_char(str, 2, ':') {
Err("missing colon in offset at correct place")?
}
let mut zhm = str.splitn(2, ':').filter_map(|t| t.parse::<usize>().ok());
let (Some(zh), Some(zm)) = (zhm.next(), zhm.next()) else {
Err("non-positive hour/min in offset")?
};
if zh > 23 || zm > 59 {
Err("hour/min in offset out of range")?
}
// apply timezone
let mut hm = (h * 60 + m) as isize + sign * (zh * 60 + zm) as isize;
if hm < 0 {
hm += 24 * 60;
debug_assert!(hm >= 0);
}
let hm = hm as usize;
(h, m) = (hm / 60, hm % 60);
}
// check leap second
if !(s < 60 || (h == 23 && m == 59)) {
Err("invalid leap second")?
}
Ok(())
}
fn validate_date_time(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
check_date_time(s)
}
fn check_date_time(s: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
// min: yyyy-mm-ddThh:mm:ssZ
if s.len() < 20 {
Err("less than 20 characters long")?;
}
if !s.is_char_boundary(10) || !s[10..].starts_with(['t', 'T']) {
Err("11th character must be t or T")?;
}
if let Err(e) = check_date(&s[..10]) {
Err(format!("invalid date element: {e}"))?;
}
if let Err(e) = check_time(&s[11..]) {
Err(format!("invalid time element: {e}"))?;
}
Ok(())
}
fn validate_duration(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
check_duration(s)?;
Ok(())
}
// see https://datatracker.ietf.org/doc/html/rfc3339#appendix-A
fn check_duration(s: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
// must start with 'P'
let Some(s) = s.strip_prefix('P') else {
Err("must start with P")?
};
if s.is_empty() {
Err("nothing after P")?
}
// dur-week
if let Some(s) = s.strip_suffix('W') {
if s.is_empty() {
Err("no number in week")?
}
if !s.chars().all(|c| c.is_ascii_digit()) {
Err("invalid week")?
}
return Ok(());
}
static UNITS: [&str; 2] = ["YMD", "HMS"];
for (i, s) in s.split('T').enumerate() {
let mut s = s;
if i != 0 && s.is_empty() {
Err("no time elements")?
}
let Some(mut units) = UNITS.get(i).cloned() else {
Err("more than one T")?
};
while !s.is_empty() {
let digit_count = s.chars().take_while(char::is_ascii_digit).count();
if digit_count == 0 {
Err("missing number")?
}
s = &s[digit_count..];
let Some(unit) = s.chars().next() else {
Err("missing unit")?
};
let Some(j) = units.find(unit) else {
if UNITS[i].contains(unit) {
Err(format!("unit {unit} out of order"))?
}
Err(format!("invalid unit {unit}"))?
};
units = &units[j + 1..];
s = &s[1..];
}
}
Ok(())
}
// see https://datatracker.ietf.org/doc/html/rfc3339#appendix-A
fn validate_period(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
let Some(slash) = s.find('/') else {
Err("missing slash")?
};
let (start, end) = (&s[..slash], &s[slash + 1..]);
if start.starts_with('P') {
if let Err(e) = check_duration(start) {
Err(format!("invalid start duration: {e}"))?
}
if let Err(e) = check_date_time(end) {
Err(format!("invalid end date-time: {e}"))?
}
} else {
if let Err(e) = check_date_time(start) {
Err(format!("invalid start date-time: {e}"))?
}
if end.starts_with('P') {
if let Err(e) = check_duration(end) {
Err(format!("invalid end duration: {e}"))?;
}
} else if let Err(e) = check_date_time(end) {
Err(format!("invalid end date-time: {e}"))?;
}
}
Ok(())
}
fn validate_hostname(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
check_hostname(s)?;
Ok(())
}
// see https://en.wikipedia.org/wiki/Hostname#Restrictions_on_valid_host_names
fn check_hostname(s: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
// entire hostname (including the delimiting dots but not a trailing dot) has a maximum of 253 ASCII characters
if s.len() > 253 {
Err("more than 253 characters long")?
}
// Hostnames are composed of series of labels concatenated with dots, as are all domain names
for label in s.split('.') {
// Each label must be from 1 to 63 characters long
if !matches!(label.len(), 1..=63) {
Err("label must be 1 to 63 characters long")?;
}
// labels must not start or end with a hyphen
if label.starts_with('-') {
Err("label starts with hyphen")?;
}
if label.ends_with('-') {
Err("label ends with hyphen")?;
}
// labels may contain only the ASCII letters 'a' through 'z' (in a case-insensitive manner),
// the digits '0' through '9', and the hyphen ('-')
if let Some(ch) = label
.chars()
.find(|c| !matches!(c, 'a'..='z' | 'A'..='Z' | '0'..='9' | '-'))
{
Err(format!("invalid character {ch:?}"))?;
}
// labels must not contain "--" in 3rd and 4th position unless they start with "xn--"
if label.len() >= 4 && &label[2..4] == "--" {
if !label.starts_with("xn--") {
Err("label has -- in 3rd/4th position but does not start with xn--")?;
} else {
let (unicode, errors) = idna::domain_to_unicode(label);
if let Err(_) = errors {
Err("invalid punycode")?;
}
check_unicode_idn_constraints(&unicode).map_err(|e| format!("invalid punycode/IDN: {e}"))?;
}
}
}
Ok(())
}
fn validate_idn_hostname(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
check_idn_hostname(s)?;
Ok(())
}
static DISALLOWED: [char; 10] = [
'\u{0640}', // ARABIC TATWEEL
'\u{07FA}', // NKO LAJANYALAN
'\u{302E}', // HANGUL SINGLE DOT TONE MARK
'\u{302F}', // HANGUL DOUBLE DOT TONE MARK
'\u{3031}', // VERTICAL KANA REPEAT MARK
'\u{3032}', // VERTICAL KANA REPEAT WITH VOICED SOUND MARK
'\u{3033}', // VERTICAL KANA REPEAT MARK UPPER HALF
'\u{3034}', // VERTICAL KANA REPEAT WITH VOICED SOUND MARK UPPER HA
'\u{3035}', // VERTICAL KANA REPEAT MARK LOWER HALF
'\u{303B}', // VERTICAL IDEOGRAPHIC ITERATION MARK
];
fn check_idn_hostname(s: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
let s = idna::domain_to_ascii_strict(s).map_err(|e| format!("idna error: {:?}", e))?;
let (unicode, errors) = idna::domain_to_unicode(&s);
if let Err(e) = errors {
Err(format!("idna decoding error: {:?}", e))?;
}
check_unicode_idn_constraints(&unicode)?;
check_hostname(&s)?;
Ok(())
}
fn check_unicode_idn_constraints(unicode: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
// see https://www.rfc-editor.org/rfc/rfc5892#section-2.6
{
if unicode.contains(DISALLOWED) {
Err("contains disallowed character")?;
}
}
// unicode string must not contain "--" in 3rd and 4th position
// and must not start and end with a '-'
// see https://www.rfc-editor.org/rfc/rfc5891#section-4.2.3.1
{
let count: usize = unicode
.chars()
.skip(2)
.take(2)
.map(|c| if c == '-' { 1 } else { 0 })
.sum();
if count == 2 {
Err("unicode string must not contain '--' in 3rd and 4th position")?;
}
}
// MIDDLE DOT is allowed between 'l' characters only
// see https://www.rfc-editor.org/rfc/rfc5892#appendix-A.3
{
let middle_dot = '\u{00b7}';
let mut s = unicode;
while let Some(i) = s.find(middle_dot) {
let prefix = &s[..i];
let suffix = &s[i + middle_dot.len_utf8()..];
if !prefix.ends_with('l') || !suffix.ends_with('l') {
Err("MIDDLE DOT is allowed between 'l' characters only")?;
}
s = suffix;
}
}
// Greek KERAIA must be followed by Greek character
// see https://www.rfc-editor.org/rfc/rfc5892#appendix-A.4
{
let keralia = '\u{0375}';
let greek = '\u{0370}'..='\u{03FF}';
let mut s = unicode;
while let Some(i) = s.find(keralia) {
let suffix = &s[i + keralia.len_utf8()..];
if !suffix.starts_with(|c| greek.contains(&c)) {
Err("Greek KERAIA must be followed by Greek character")?;
}
s = suffix;
}
}
// Hebrew GERESH must be preceded by Hebrew character
// see https://www.rfc-editor.org/rfc/rfc5892#appendix-A.5
//
// Hebrew GERSHAYIM must be preceded by Hebrew character
// see https://www.rfc-editor.org/rfc/rfc5892#appendix-A.6
{
let geresh = '\u{05F3}';
let gereshayim = '\u{05F4}';
let hebrew = '\u{0590}'..='\u{05FF}';
for ch in [geresh, gereshayim] {
let mut s = unicode;
while let Some(i) = s.find(ch) {
let prefix = &s[..i];
if !prefix.ends_with(|c| hebrew.contains(&c)) {
if i == 0 {
Err("Hebrew GERESH must be preceded by Hebrew character")?;
} else {
Err("Hebrew GERESHYIM must be preceded by Hebrew character")?;
}
}
let suffix = &s[i + ch.len_utf8()..];
s = suffix;
}
}
}
// KATAKANA MIDDLE DOT must be with Hiragana, Katakana, or Han
// see https://www.rfc-editor.org/rfc/rfc5892#appendix-A.7
{
let katakana_middle_dot = '\u{30FB}';
if unicode.contains(katakana_middle_dot) {
let hiragana = '\u{3040}'..='\u{309F}';
let katakana = '\u{30A0}'..='\u{30FF}';
let han = '\u{4E00}'..='\u{9FFF}'; // https://en.wikipedia.org/wiki/CJK_Unified_Ideographs_(Unicode_block): is this range correct??
if unicode.contains(|c| hiragana.contains(&c))
|| unicode.contains(|c| c != katakana_middle_dot && katakana.contains(&c))
|| unicode.contains(|c| han.contains(&c))
{
// ok
} else {
Err("KATAKANA MIDDLE DOT must be with Hiragana, Katakana, or Han")?;
}
}
}
// ARABIC-INDIC DIGITS and Extended Arabic-Indic Digits cannot be mixed
// see https://www.rfc-editor.org/rfc/rfc5892#appendix-A.8
// see https://www.rfc-editor.org/rfc/rfc5892#appendix-A.9
{
let arabic_indic_digits = '\u{0660}'..='\u{0669}';
let extended_arabic_indic_digits = '\u{06F0}'..='\u{06F9}';
if unicode.contains(|c| arabic_indic_digits.contains(&c))
&& unicode.contains(|c| extended_arabic_indic_digits.contains(&c))
{
Err("ARABIC-INDIC DIGITS and Extended Arabic-Indic Digits cannot be mixed")?;
}
}
// ZERO WIDTH JOINER must be preceded by Virama
// see https://www.rfc-editor.org/rfc/rfc5892#appendix-A.2
{
let zero_width_jointer = '\u{200D}';
static VIRAMA: [char; 61] = [
'\u{094D}',
'\u{09CD}',
'\u{0A4D}',
'\u{0ACD}',
'\u{0B4D}',
'\u{0BCD}',
'\u{0C4D}',
'\u{0CCD}',
'\u{0D3B}',
'\u{0D3C}',
'\u{0D4D}',
'\u{0DCA}',
'\u{0E3A}',
'\u{0EBA}',
'\u{0F84}',
'\u{1039}',
'\u{103A}',
'\u{1714}',
'\u{1734}',
'\u{17D2}',
'\u{1A60}',
'\u{1B44}',
'\u{1BAA}',
'\u{1BAB}',
'\u{1BF2}',
'\u{1BF3}',
'\u{2D7F}',
'\u{A806}',
'\u{A82C}',
'\u{A8C4}',
'\u{A953}',
'\u{A9C0}',
'\u{AAF6}',
'\u{ABED}',
'\u{10A3F}',
'\u{11046}',
'\u{1107F}',
'\u{110B9}',
'\u{11133}',
'\u{11134}',
'\u{111C0}',
'\u{11235}',
'\u{112EA}',
'\u{1134D}',
'\u{11442}',
'\u{114C2}',
'\u{115BF}',
'\u{1163F}',
'\u{116B6}',
'\u{1172B}',
'\u{11839}',
'\u{1193D}',
'\u{1193E}',
'\u{119E0}',
'\u{11A34}',
'\u{11A47}',
'\u{11A99}',
'\u{11C3F}',
'\u{11D44}',
'\u{11D45}',
'\u{11D97}',
]; // https://www.compart.com/en/unicode/combining/9
let mut s = unicode;
while let Some(i) = s.find(zero_width_jointer) {
let prefix = &s[..i];
if !prefix.ends_with(VIRAMA) {
Err("ZERO WIDTH JOINER must be preceded by Virama")?;
}
let suffix = &s[i + zero_width_jointer.len_utf8()..];
s = suffix;
}
}
Ok(())
}
fn validate_email(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
check_email(s)?;
Ok(())
}
// see https://en.wikipedia.org/wiki/Email_address
fn check_email(s: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
// entire email address to be no more than 254 characters long
if s.len() > 254 {
Err("more than 254 characters long")?
}
// email address is generally recognized as having two parts joined with an at-sign
let Some(at) = s.rfind('@') else {
Err("missing @")?
};
let (local, domain) = (&s[..at], &s[at + 1..]);
// local part may be up to 64 characters long
if local.len() > 64 {
Err("local part more than 64 characters long")?
}
if local.len() > 1 && local.starts_with('"') && local.ends_with('"') {
// quoted
let local = &local[1..local.len() - 1];
if local.contains(['\\', '"']) {
Err("backslash and quote not allowed within quoted local part")?
}
} else {
// unquoted
if local.starts_with('.') {
Err("starts with dot")?
}
if local.ends_with('.') {
Err("ends with dot")?
}
// consecutive dots not allowed
if local.contains("..") {
Err("consecutive dots")?
}
// check allowd chars
if let Some(ch) = local
.chars()
.find(|c| !(c.is_ascii_alphanumeric() || ".!#$%&'*+-/=?^_`{|}~".contains(*c)))
{
Err(format!("invalid character {ch:?}"))?
}
}
// domain if enclosed in brackets, must match an IP address
if domain.starts_with('[') && domain.ends_with(']') {
let s = &domain[1..domain.len() - 1];
if let Some(s) = s.strip_prefix("IPv6:") {
if let Err(e) = s.parse::<Ipv6Addr>() {
Err(format!("invalid ipv6 address: {e}"))?
}
return Ok(());
}
if let Err(e) = s.parse::<Ipv4Addr>() {
Err(format!("invalid ipv4 address: {e}"))?
}
return Ok(());
}
// domain must match the requirements for a hostname
if let Err(e) = check_hostname(domain) {
Err(format!("invalid domain: {e}"))?
}
Ok(())
}
fn validate_idn_email(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
let Some(at) = s.rfind('@') else {
Err("missing @")?
};
let (local, domain) = (&s[..at], &s[at + 1..]);
let local = idna::domain_to_ascii_strict(local).map_err(|e| format!("idna error: {:?}", e))?;
let domain = idna::domain_to_ascii_strict(domain).map_err(|e| format!("idna error: {:?}", e))?;
if let Err(e) = check_idn_hostname(&domain) {
Err(format!("invalid domain: {e}"))?
}
check_email(&format!("{local}@{domain}"))
}
fn validate_json_pointer(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
check_json_pointer(s)?;
Ok(())
}
// see https://www.rfc-editor.org/rfc/rfc6901#section-3
fn check_json_pointer(s: &str) -> Result<(), Box<dyn Error + Send + Sync>> {
if s.is_empty() {
return Ok(());
}
if !s.starts_with('/') {
Err("not starting with slash")?;
}
for token in s.split('/').skip(1) {
let mut chars = token.chars();
while let Some(ch) = chars.next() {
if ch == '~' {
if !matches!(chars.next(), Some('0' | '1')) {
Err("~ must be followed by 0 or 1")?;
}
} else if !matches!(ch, '\x00'..='\x2E' | '\x30'..='\x7D' | '\x7F'..='\u{10FFFF}') {
Err("contains disallowed character")?;
}
}
}
Ok(())
}
// see https://tools.ietf.org/html/draft-handrews-relative-json-pointer-01#section-3
fn validate_relative_json_pointer(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
// start with non-negative-integer
let num_digits = s.chars().take_while(char::is_ascii_digit).count();
if num_digits == 0 {
Err("must start with non-negative integer")?;
}
if num_digits > 1 && s.starts_with('0') {
Err("starts with zero")?;
}
let s = &s[num_digits..];
// followed by either json-pointer or '#'
if s == "#" {
return Ok(());
}
if let Err(e) = check_json_pointer(s) {
Err(format!("invalid json-pointer element: {e}"))?;
}
Ok(())
}
// see https://datatracker.ietf.org/doc/html/rfc4122#page-4
fn validate_uuid(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
static HEX_GROUPS: [usize; 5] = [8, 4, 4, 4, 12];
let mut i = 0;
for group in s.split('-') {
if i >= HEX_GROUPS.len() {
Err("more than 5 elements")?;
}
if group.len() != HEX_GROUPS[i] {
Err(format!(
"element {} must be {} characters long",
i + 1,
HEX_GROUPS[i]
))?;
}
if let Some(ch) = group.chars().find(|c| !c.is_ascii_hexdigit()) {
Err(format!("non-hex character {ch:?}"))?;
}
i += 1;
}
if i != HEX_GROUPS.len() {
Err("must have 5 elements")?;
}
Ok(())
}
fn validate_uri(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
if fluent_uri::UriRef::parse(s.as_str()).map_err(|e| e.to_string())?.scheme().is_none() {
Err("relative url")?;
};
Ok(())
}
fn validate_iri(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
match Url::parse(s) {
Ok(_) => Ok(()),
Err(url::ParseError::RelativeUrlWithoutBase) => Err("relative url")?,
Err(e) => Err(e)?,
}
}
lazy_static! {
static ref TEMP_URL: Url = Url::parse("http://temp.com").unwrap();
}
fn parse_uri_reference(s: &str) -> Result<Url, Box<dyn Error + Send + Sync>> {
if s.contains('\\') {
Err("contains \\\\")?;
}
Ok(TEMP_URL.join(s)?)
}
fn validate_uri_reference(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
fluent_uri::UriRef::parse(s.as_str()).map_err(|e| e.to_string())?;
Ok(())
}
fn validate_iri_reference(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
parse_uri_reference(s)?;
Ok(())
}
fn validate_uri_template(v: &Value) -> Result<(), Box<dyn Error + Send + Sync>> {
let Value::String(s) = v else {
return Ok(());
};
let url = parse_uri_reference(s)?;
let path = url.path();
// path we got has curly bases percent encoded
let path = percent_decode_str(path).decode_utf8()?;
// ensure curly brackets are not nested and balanced
for part in path.as_ref().split('/') {
let mut want = true;
for got in part
.chars()
.filter(|c| matches!(c, '{' | '}'))
.map(|c| c == '{')
{
if got != want {
Err("nested curly braces")?;
}
want = !want;
}
if !want {
Err("no matching closing brace")?
}
}
Ok(())
}

220
src/database/mod.rs Normal file
View File

@ -0,0 +1,220 @@
pub mod r#enum;
pub mod formats;
pub mod page;
pub mod punc;
pub mod schema;
pub mod r#type;
use crate::database::r#enum::Enum;
use crate::database::punc::Punc;
use crate::database::schema::Schema;
use crate::database::r#type::Type;
use std::collections::HashMap;
pub struct Database {
pub enums: HashMap<String, Enum>,
pub types: HashMap<String, Type>,
pub puncs: HashMap<String, Punc>,
pub schemas: HashMap<String, Schema>,
pub descendants: HashMap<String, Vec<String>>,
}
impl Database {
pub fn new(val: &serde_json::Value) -> Self {
let mut db = Self {
enums: HashMap::new(),
types: HashMap::new(),
puncs: HashMap::new(),
schemas: HashMap::new(),
descendants: HashMap::new(),
};
if let Some(arr) = val.get("enums").and_then(|v| v.as_array()) {
for item in arr {
if let Ok(def) = serde_json::from_value::<Enum>(item.clone()) {
db.enums.insert(def.name.clone(), def);
}
}
}
if let Some(arr) = val.get("types").and_then(|v| v.as_array()) {
for item in arr {
if let Ok(def) = serde_json::from_value::<Type>(item.clone()) {
db.types.insert(def.name.clone(), def);
}
}
}
if let Some(arr) = val.get("puncs").and_then(|v| v.as_array()) {
for item in arr {
if let Ok(def) = serde_json::from_value::<Punc>(item.clone()) {
db.puncs.insert(def.name.clone(), def);
}
}
}
if let Some(arr) = val.get("schemas").and_then(|v| v.as_array()) {
for (i, item) in arr.iter().enumerate() {
if let Ok(mut schema) = serde_json::from_value::<Schema>(item.clone()) {
let id = schema
.obj
.id
.clone()
.unwrap_or_else(|| format!("schema_{}", i));
schema.obj.id = Some(id.clone());
db.schemas.insert(id, schema);
}
}
}
let _ = db.compile();
db
}
/// Organizes the graph of the database, compiling regex, format functions, and pointing schema references.
fn compile(&mut self) -> Result<(), String> {
self.collect_schemas();
// 1. Compile regex and formats sequentially
for schema in self.schemas.values_mut() {
schema.compile();
}
// 2. Compute the Unified Semantic Graph (descendants)
self.collect_descendents();
// 3. For any schema representing a Postgres table, cache its allowed subclasses
self.compile_allowed_types();
// 4. Finally, securely link all string $refs into memory pointers (Arc)
self.compile_pointers();
Ok(())
}
fn collect_schemas(&mut self) {
let mut to_insert = Vec::new();
for (_, type_def) in &self.types {
for schema in &type_def.schemas {
if let Some(id) = &schema.obj.id {
to_insert.push((id.clone(), schema.clone()));
}
}
}
for (_, punc_def) in &self.puncs {
for schema in &punc_def.schemas {
if let Some(id) = &schema.obj.id {
to_insert.push((id.clone(), schema.clone()));
}
}
}
for (_, enum_def) in &self.enums {
for schema in &enum_def.schemas {
if let Some(id) = &schema.obj.id {
to_insert.push((id.clone(), schema.clone()));
}
}
}
for (id, schema) in to_insert {
self.schemas.insert(id, schema);
}
}
fn collect_descendents(&mut self) {
let mut direct_children: HashMap<String, Vec<String>> = HashMap::new();
// First pass: Find all schemas that have a $ref to another schema
let schema_ids: Vec<String> = self.schemas.keys().cloned().collect();
for id in schema_ids {
if let Some(ref_str) = self.schemas.get(&id).and_then(|s| s.obj.ref_string.clone()) {
if self.schemas.contains_key(&ref_str) {
direct_children.entry(ref_str).or_default().push(id.clone());
}
}
}
// Now compute descendants for all schemas
let mut descendants_map: HashMap<String, Vec<String>> = HashMap::new();
for key in self.schemas.keys() {
let mut descendants = Vec::new();
let mut queue = Vec::new();
if let Some(children) = direct_children.get(key) {
queue.extend(children.iter().cloned());
}
let mut visited = std::collections::HashSet::new();
while let Some(child) = queue.pop() {
if visited.insert(child.clone()) {
descendants.push(child.clone());
if let Some(grandchildren) = direct_children.get(&child) {
queue.extend(grandchildren.iter().cloned());
}
}
}
descendants_map.insert(key.clone(), descendants);
}
self.descendants = descendants_map;
}
fn compile_allowed_types(&mut self) {
// 1. Identify which types act as bases (table-backed schemas)
let mut entity_bases = HashMap::new();
for type_def in self.types.values() {
for type_schema in &type_def.schemas {
if let Some(id) = &type_schema.obj.id {
entity_bases.insert(id.clone(), type_def.name.clone());
}
}
}
// 2. Compute compiled_allowed_types for all descendants of entity bases
let mut allowed_types_map: HashMap<String, std::collections::HashSet<String>> = HashMap::new();
for base_id in entity_bases.keys() {
allowed_types_map.insert(
base_id.clone(),
self
.descendants
.get(base_id)
.unwrap_or(&vec![])
.iter()
.cloned()
.collect(),
);
if let Some(descendants) = self.descendants.get(base_id) {
let set: std::collections::HashSet<String> = descendants.iter().cloned().collect();
for desc_id in descendants {
allowed_types_map.insert(desc_id.clone(), set.clone());
}
}
}
// 3. Inject types into the schemas
let schema_ids: Vec<String> = self.schemas.keys().cloned().collect();
for id in schema_ids {
if let Some(set) = allowed_types_map.get(&id) {
if let Some(schema) = self.schemas.get_mut(&id) {
schema.obj.compiled_allowed_types = Some(set.clone());
}
}
}
}
fn compile_pointers(&mut self) {
let schema_ids: Vec<String> = self.schemas.keys().cloned().collect();
for id in schema_ids {
let mut compiled_ref = None;
if let Some(schema) = self.schemas.get(&id) {
if let Some(ref_str) = &schema.obj.ref_string {
if let Some(target) = self.schemas.get(ref_str) {
compiled_ref = Some(std::sync::Arc::new(target.clone()));
}
}
}
if let Some(schema) = self.schemas.get_mut(&id) {
schema.obj.compiled_ref = compiled_ref;
}
}
}
}

35
src/database/page.rs Normal file
View File

@ -0,0 +1,35 @@
use indexmap::IndexMap;
use serde::{Deserialize, Serialize};
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
#[serde(default)]
pub struct Page {
#[serde(skip_serializing_if = "Option::is_none")]
pub path: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub title: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub sidebar: Option<Sidebar>,
#[serde(skip_serializing_if = "Option::is_none")]
pub actions: Option<IndexMap<String, Action>>,
}
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
#[serde(default)]
pub struct Sidebar {
#[serde(skip_serializing_if = "Option::is_none")]
pub category: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub priority: Option<i32>,
}
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
#[serde(default)]
pub struct Action {
#[serde(skip_serializing_if = "Option::is_none")]
pub punc: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub navigate: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub present: Option<String>,
}

20
src/database/punc.rs Normal file
View File

@ -0,0 +1,20 @@
use crate::database::page::Page;
use crate::database::schema::Schema;
use serde::{Deserialize, Serialize};
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
#[serde(default)]
pub struct Punc {
pub id: String,
pub r#type: String,
pub name: String,
pub module: String,
pub source: String,
pub description: Option<String>,
pub public: bool,
pub form: bool,
pub get: Option<String>,
pub page: Option<Page>,
#[serde(default)]
pub schemas: Vec<Schema>,
}

15
src/database/relation.rs Normal file
View File

@ -0,0 +1,15 @@
use serde::{Deserialize, Serialize};
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
#[serde(default)]
pub struct Relation {
pub id: String,
pub constraint_name: String,
pub source_type: String,
#[serde(default)]
pub source_columns: Vec<String>,
pub destination_type: String,
#[serde(default)]
pub destination_columns: Vec<String>,
pub prefix: Option<String>,
}

356
src/database/schema.rs Normal file
View File

@ -0,0 +1,356 @@
use serde::{Deserialize, Serialize};
use serde_json::Value;
use std::collections::BTreeMap;
use std::sync::Arc;
// Schema mirrors the Go Punc Generator's schema struct for consistency.
// It is an order-preserving representation of a JSON Schema.
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
pub struct SchemaObject {
// Core Schema Keywords
#[serde(rename = "$id")]
pub id: Option<String>,
#[serde(rename = "$ref")]
pub ref_string: Option<String>,
/*
Note: The `Ref` field in the Go struct is a pointer populated by the linker.
In Rust, we might handle this differently (e.g., separate lookup or Rc/Arc),
so we omit the direct recursive `Ref` field for now and rely on `ref_string`.
*/
pub description: Option<String>,
pub title: Option<String>,
#[serde(default)] // Allow missing type
#[serde(rename = "type")]
pub type_: Option<SchemaTypeOrArray>, // Handles string or array of strings
// Object Keywords
pub properties: Option<BTreeMap<String, Arc<Schema>>>,
#[serde(rename = "patternProperties")]
pub pattern_properties: Option<BTreeMap<String, Arc<Schema>>>,
#[serde(rename = "additionalProperties")]
pub additional_properties: Option<Arc<Schema>>,
#[serde(rename = "$family")]
pub family: Option<String>,
pub required: Option<Vec<String>>,
// dependencies can be schema dependencies or property dependencies
pub dependencies: Option<BTreeMap<String, Dependency>>,
// Array Keywords
#[serde(rename = "items")]
pub items: Option<Arc<Schema>>,
#[serde(rename = "prefixItems")]
pub prefix_items: Option<Vec<Arc<Schema>>>,
// String Validation
#[serde(rename = "minLength")]
pub min_length: Option<f64>,
#[serde(rename = "maxLength")]
pub max_length: Option<f64>,
pub pattern: Option<String>,
// Array Validation
#[serde(rename = "minItems")]
pub min_items: Option<f64>,
#[serde(rename = "maxItems")]
pub max_items: Option<f64>,
#[serde(rename = "uniqueItems")]
pub unique_items: Option<bool>,
#[serde(rename = "contains")]
pub contains: Option<Arc<Schema>>,
#[serde(rename = "minContains")]
pub min_contains: Option<f64>,
#[serde(rename = "maxContains")]
pub max_contains: Option<f64>,
// Object Validation
#[serde(rename = "minProperties")]
pub min_properties: Option<f64>,
#[serde(rename = "maxProperties")]
pub max_properties: Option<f64>,
#[serde(rename = "propertyNames")]
pub property_names: Option<Arc<Schema>>,
// Numeric Validation
pub format: Option<String>,
#[serde(rename = "enum")]
pub enum_: Option<Vec<Value>>, // `enum` is a reserved keyword in Rust
#[serde(
default,
rename = "const",
deserialize_with = "crate::validator::util::deserialize_some"
)]
pub const_: Option<Value>,
// Numeric Validation
#[serde(rename = "multipleOf")]
pub multiple_of: Option<f64>,
pub minimum: Option<f64>,
pub maximum: Option<f64>,
#[serde(rename = "exclusiveMinimum")]
pub exclusive_minimum: Option<f64>,
#[serde(rename = "exclusiveMaximum")]
pub exclusive_maximum: Option<f64>,
// Combining Keywords
#[serde(rename = "allOf")]
pub all_of: Option<Vec<Arc<Schema>>>,
#[serde(rename = "anyOf")]
pub any_of: Option<Vec<Arc<Schema>>>,
#[serde(rename = "oneOf")]
pub one_of: Option<Vec<Arc<Schema>>>,
#[serde(rename = "not")]
pub not: Option<Arc<Schema>>,
#[serde(rename = "if")]
pub if_: Option<Arc<Schema>>,
#[serde(rename = "then")]
pub then_: Option<Arc<Schema>>,
#[serde(rename = "else")]
pub else_: Option<Arc<Schema>>,
// Custom Vocabularies
pub form: Option<Vec<String>>,
pub display: Option<Vec<String>>,
#[serde(rename = "enumNames")]
pub enum_names: Option<Vec<String>>,
pub control: Option<String>,
pub actions: Option<BTreeMap<String, Action>>,
pub computer: Option<String>,
#[serde(default)]
pub extensible: Option<bool>,
// Compiled Fields (Hidden from JSON/Serde)
#[serde(skip)]
pub compiled_ref: Option<Arc<Schema>>,
#[serde(skip)]
pub compiled_allowed_types: Option<std::collections::HashSet<String>>,
#[serde(skip)]
pub compiled_format: Option<CompiledFormat>,
#[serde(skip)]
pub compiled_pattern: Option<CompiledRegex>,
#[serde(skip)]
pub compiled_pattern_properties: Option<Vec<(CompiledRegex, Arc<Schema>)>>,
}
pub enum ResolvedRef<'a> {
Local(&'a Schema),
Global(&'a Schema, &'a Schema),
}
/// Represents a compiled format validator
#[derive(Clone)]
pub enum CompiledFormat {
Func(fn(&serde_json::Value) -> Result<(), Box<dyn std::error::Error + Send + Sync>>),
Regex(regex::Regex),
}
impl std::fmt::Debug for CompiledFormat {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
match self {
CompiledFormat::Func(_) => write!(f, "CompiledFormat::Func(...)"),
CompiledFormat::Regex(r) => write!(f, "CompiledFormat::Regex({:?})", r),
}
}
}
/// A wrapper for compiled regex patterns
#[derive(Debug, Clone)]
pub struct CompiledRegex(pub regex::Regex);
#[derive(Debug, Clone, Serialize)]
pub struct Schema {
#[serde(flatten)]
pub obj: SchemaObject,
#[serde(skip)]
pub always_fail: bool,
}
impl Default for Schema {
fn default() -> Self {
Schema {
obj: SchemaObject::default(),
always_fail: false,
}
}
}
impl std::ops::Deref for Schema {
type Target = SchemaObject;
fn deref(&self) -> &Self::Target {
&self.obj
}
}
impl std::ops::DerefMut for Schema {
fn deref_mut(&mut self) -> &mut Self::Target {
&mut self.obj
}
}
impl Schema {
pub fn resolve_ref(&self, _ref_string: &str) -> Option<&Arc<Schema>> {
// This is vestigial for now. References are global pointers. We will remove this shortly.
None
}
pub fn compile(&mut self) {
if let Some(format_str) = &self.obj.format {
if let Some(fmt) = crate::database::formats::FORMATS.get(format_str.as_str()) {
self.obj.compiled_format = Some(crate::database::schema::CompiledFormat::Func(fmt.func));
}
}
if let Some(pattern_str) = &self.obj.pattern {
if let Ok(re) = regex::Regex::new(pattern_str) {
self.obj.compiled_pattern = Some(crate::database::schema::CompiledRegex(re));
}
}
if let Some(pattern_props) = &self.obj.pattern_properties {
let mut compiled = Vec::new();
for (k, v) in pattern_props {
if let Ok(re) = regex::Regex::new(k) {
compiled.push((crate::database::schema::CompiledRegex(re), v.clone()));
}
}
if !compiled.is_empty() {
self.obj.compiled_pattern_properties = Some(compiled);
}
}
// Crawl children recursively to compile their internals
if let Some(props) = &mut self.obj.properties {
for (_, v) in props {
// Safe deep mutation workaround without unsafe Arc unwrap
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
}
if let Some(arr) = &mut self.obj.prefix_items {
for v in arr.iter_mut() {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
}
if let Some(arr) = &mut self.obj.all_of {
for v in arr.iter_mut() {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
}
if let Some(arr) = &mut self.obj.any_of {
for v in arr.iter_mut() {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
}
if let Some(arr) = &mut self.obj.one_of {
for v in arr.iter_mut() {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
}
if let Some(v) = &mut self.obj.additional_properties {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
if let Some(v) = &mut self.obj.items {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
if let Some(v) = &mut self.obj.contains {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
if let Some(v) = &mut self.obj.property_names {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
if let Some(v) = &mut self.obj.not {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
if let Some(v) = &mut self.obj.if_ {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
if let Some(v) = &mut self.obj.then_ {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
if let Some(v) = &mut self.obj.else_ {
let mut inner = (**v).clone();
inner.compile();
*v = Arc::new(inner);
}
}
}
impl<'de> Deserialize<'de> for Schema {
fn deserialize<D>(deserializer: D) -> Result<Self, D::Error>
where
D: serde::Deserializer<'de>,
{
let v: Value = Deserialize::deserialize(deserializer)?;
if let Some(b) = v.as_bool() {
let mut obj = SchemaObject::default();
if b {
obj.extensible = Some(true);
}
return Ok(Schema {
obj,
always_fail: !b,
});
}
let obj: SchemaObject = serde_json::from_value(v.clone()).map_err(serde::de::Error::custom)?;
Ok(Schema {
obj,
always_fail: false,
})
}
}
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(untagged)]
pub enum SchemaTypeOrArray {
Single(String),
Multiple(Vec<String>),
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Action {
pub navigate: Option<String>,
pub punc: Option<String>,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(untagged)]
pub enum Dependency {
Props(Vec<String>),
Schema(Arc<Schema>),
}

35
src/database/type.rs Normal file
View File

@ -0,0 +1,35 @@
use crate::database::schema::Schema;
use serde::{Deserialize, Serialize};
use serde_json::Value;
#[derive(Debug, Clone, Serialize, Deserialize, Default)]
#[serde(default)]
pub struct Type {
pub id: String,
pub r#type: String,
pub name: String,
pub module: String,
pub source: String,
#[serde(default)]
pub historical: bool,
#[serde(default)]
pub sensitive: bool,
#[serde(default)]
pub ownable: bool,
pub longevity: Option<i32>,
#[serde(default)]
pub hierarchy: Vec<String>,
pub relationship: Option<bool>,
#[serde(default)]
pub fields: Vec<String>,
pub grouped_fields: Option<Value>,
#[serde(default)]
pub lookup_fields: Vec<String>,
#[serde(default)]
pub null_fields: Vec<String>,
#[serde(default)]
pub default_fields: Vec<String>,
pub field_types: Option<Value>,
#[serde(default)]
pub schemas: Vec<Schema>,
}