@rtif-sdk/formats 3.0.0

package/dist/shared/url.d.ts

@@ -0,0 +1,20 @@
+/**
+ * Shared URL sanitization for every URL-bearing attribute parsed from
+ * external input: HTML `href`/`src` and markdown link destinations.
+ *
+ * This is a protocol *allowlist*, not a blocklist (the v2 lesson): `http:`,
+ * `https:`, `mailto:`, and scheme-less (relative) URLs pass; everything else
+ * (`javascript:`, `data:`, `vbscript:`, `file:`, ...) is rejected.
+ */
+/**
+ * Validate a URL against the protocol allowlist.
+ *
+ * Whitespace and control characters are stripped before scheme detection
+ * because browsers ignore them when resolving URLs — `java\nscript:` is a
+ * real attack vector, so it must be detected as the `javascript` scheme.
+ *
+ * @param raw - The attribute value or link destination as written
+ * @returns the trimmed URL when allowed, or `null` when it must be dropped
+ */
+export declare function sanitizeUrl(raw: string): string | null;
+//# sourceMappingURL=url.d.ts.map

package/dist/shared/url.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"url.d.ts","sourceRoot":"","sources":["../../src/shared/url.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAWH;;;;;;;;;GASG;AACH,wBAAgB,WAAW,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CAMtD"}

package/dist/shared/url.js

@@ -0,0 +1,33 @@
+/**
+ * Shared URL sanitization for every URL-bearing attribute parsed from
+ * external input: HTML `href`/`src` and markdown link destinations.
+ *
+ * This is a protocol *allowlist*, not a blocklist (the v2 lesson): `http:`,
+ * `https:`, `mailto:`, and scheme-less (relative) URLs pass; everything else
+ * (`javascript:`, `data:`, `vbscript:`, `file:`, ...) is rejected.
+ */
+const ALLOWED_SCHEMES = new Set(['http', 'https', 'mailto']);
+/** Matches a URL scheme per RFC 3986: a letter, then letters/digits/`+`/`-`/`.`. */
+const SCHEME_PATTERN = /^([a-z][a-z0-9+.-]*):/;
+/** Whitespace and control characters that browsers ignore inside URLs. */
+// eslint-disable-next-line no-control-regex -- control chars are intentionally matched for URL sanitization
+const IGNORED_CHARS = /[\u0000-\u0020\u00a0]+/g;
+/**
+ * Validate a URL against the protocol allowlist.
+ *
+ * Whitespace and control characters are stripped before scheme detection
+ * because browsers ignore them when resolving URLs — `java\nscript:` is a
+ * real attack vector, so it must be detected as the `javascript` scheme.
+ *
+ * @param raw - The attribute value or link destination as written
+ * @returns the trimmed URL when allowed, or `null` when it must be dropped
+ */
+export function sanitizeUrl(raw) {
+    const url = raw.trim();
+    const probe = url.replace(IGNORED_CHARS, '').toLowerCase();
+    const match = SCHEME_PATTERN.exec(probe);
+    if (match === null)
+        return url; // no scheme — a relative URL
+    return ALLOWED_SCHEMES.has(match[1]) ? url : null;
+}
+//# sourceMappingURL=url.js.map

package/dist/shared/url.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"url.js","sourceRoot":"","sources":["../../src/shared/url.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAEH,MAAM,eAAe,GAAwB,IAAI,GAAG,CAAC,CAAC,MAAM,EAAE,OAAO,EAAE,QAAQ,CAAC,CAAC,CAAC;AAElF,oFAAoF;AACpF,MAAM,cAAc,GAAG,uBAAuB,CAAC;AAE/C,0EAA0E;AAC1E,4GAA4G;AAC5G,MAAM,aAAa,GAAG,yBAAyB,CAAC;AAEhD;;;;;;;;;GASG;AACH,MAAM,UAAU,WAAW,CAAC,GAAW;IACrC,MAAM,GAAG,GAAG,GAAG,CAAC,IAAI,EAAE,CAAC;IACvB,MAAM,KAAK,GAAG,GAAG,CAAC,OAAO,CAAC,aAAa,EAAE,EAAE,CAAC,CAAC,WAAW,EAAE,CAAC;IAC3D,MAAM,KAAK,GAAG,cAAc,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IACzC,IAAI,KAAK,KAAK,IAAI;QAAE,OAAO,GAAG,CAAC,CAAC,6BAA6B;IAC7D,OAAO,eAAe,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,CAAW,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC;AAC9D,CAAC"}

package/package.json

@@ -0,0 +1,28 @@
+{
+  "name": "@rtif-sdk/formats",
+  "version": "3.0.0",
+  "description": "RTIF format codecs: HTML, Markdown, plaintext.",
+  "type": "module",
+  "sideEffects": false,
+  "exports": {
+    "./html": {
+      "types": "./dist/html/index.d.ts",
+      "import": "./dist/html/index.js"
+    },
+    "./markdown": {
+      "types": "./dist/markdown/index.d.ts",
+      "import": "./dist/markdown/index.js"
+    },
+    "./plaintext": {
+      "types": "./dist/plaintext/index.d.ts",
+      "import": "./dist/plaintext/index.js"
+    }
+  },
+  "files": [
+    "dist"
+  ],
+  "dependencies": {
+    "@rtif-sdk/core": "3.0.0"
+  },
+  "license": "MIT"
+}