parse5 6.0.1 → 7.3.0

package/dist/common/foreign-content.js

@@ -0,0 +1,228 @@
+import { TAG_ID as $, NS, ATTRS, getTagID } from './html.js';
+//MIME types
+const MIME_TYPES = {
+    TEXT_HTML: 'text/html',
+    APPLICATION_XML: 'application/xhtml+xml',
+};
+//Attributes
+const DEFINITION_URL_ATTR = 'definitionurl';
+const ADJUSTED_DEFINITION_URL_ATTR = 'definitionURL';
+const SVG_ATTRS_ADJUSTMENT_MAP = new Map([
+    'attributeName',
+    'attributeType',
+    'baseFrequency',
+    'baseProfile',
+    'calcMode',
+    'clipPathUnits',
+    'diffuseConstant',
+    'edgeMode',
+    'filterUnits',
+    'glyphRef',
+    'gradientTransform',
+    'gradientUnits',
+    'kernelMatrix',
+    'kernelUnitLength',
+    'keyPoints',
+    'keySplines',
+    'keyTimes',
+    'lengthAdjust',
+    'limitingConeAngle',
+    'markerHeight',
+    'markerUnits',
+    'markerWidth',
+    'maskContentUnits',
+    'maskUnits',
+    'numOctaves',
+    'pathLength',
+    'patternContentUnits',
+    'patternTransform',
+    'patternUnits',
+    'pointsAtX',
+    'pointsAtY',
+    'pointsAtZ',
+    'preserveAlpha',
+    'preserveAspectRatio',
+    'primitiveUnits',
+    'refX',
+    'refY',
+    'repeatCount',
+    'repeatDur',
+    'requiredExtensions',
+    'requiredFeatures',
+    'specularConstant',
+    'specularExponent',
+    'spreadMethod',
+    'startOffset',
+    'stdDeviation',
+    'stitchTiles',
+    'surfaceScale',
+    'systemLanguage',
+    'tableValues',
+    'targetX',
+    'targetY',
+    'textLength',
+    'viewBox',
+    'viewTarget',
+    'xChannelSelector',
+    'yChannelSelector',
+    'zoomAndPan',
+].map((attr) => [attr.toLowerCase(), attr]));
+const XML_ATTRS_ADJUSTMENT_MAP = new Map([
+    ['xlink:actuate', { prefix: 'xlink', name: 'actuate', namespace: NS.XLINK }],
+    ['xlink:arcrole', { prefix: 'xlink', name: 'arcrole', namespace: NS.XLINK }],
+    ['xlink:href', { prefix: 'xlink', name: 'href', namespace: NS.XLINK }],
+    ['xlink:role', { prefix: 'xlink', name: 'role', namespace: NS.XLINK }],
+    ['xlink:show', { prefix: 'xlink', name: 'show', namespace: NS.XLINK }],
+    ['xlink:title', { prefix: 'xlink', name: 'title', namespace: NS.XLINK }],
+    ['xlink:type', { prefix: 'xlink', name: 'type', namespace: NS.XLINK }],
+    ['xml:lang', { prefix: 'xml', name: 'lang', namespace: NS.XML }],
+    ['xml:space', { prefix: 'xml', name: 'space', namespace: NS.XML }],
+    ['xmlns', { prefix: '', name: 'xmlns', namespace: NS.XMLNS }],
+    ['xmlns:xlink', { prefix: 'xmlns', name: 'xlink', namespace: NS.XMLNS }],
+]);
+//SVG tag names adjustment map
+export const SVG_TAG_NAMES_ADJUSTMENT_MAP = new Map([
+    'altGlyph',
+    'altGlyphDef',
+    'altGlyphItem',
+    'animateColor',
+    'animateMotion',
+    'animateTransform',
+    'clipPath',
+    'feBlend',
+    'feColorMatrix',
+    'feComponentTransfer',
+    'feComposite',
+    'feConvolveMatrix',
+    'feDiffuseLighting',
+    'feDisplacementMap',
+    'feDistantLight',
+    'feFlood',
+    'feFuncA',
+    'feFuncB',
+    'feFuncG',
+    'feFuncR',
+    'feGaussianBlur',
+    'feImage',
+    'feMerge',
+    'feMergeNode',
+    'feMorphology',
+    'feOffset',
+    'fePointLight',
+    'feSpecularLighting',
+    'feSpotLight',
+    'feTile',
+    'feTurbulence',
+    'foreignObject',
+    'glyphRef',
+    'linearGradient',
+    'radialGradient',
+    'textPath',
+].map((tn) => [tn.toLowerCase(), tn]));
+//Tags that causes exit from foreign content
+const EXITS_FOREIGN_CONTENT = new Set([
+    $.B,
+    $.BIG,
+    $.BLOCKQUOTE,
+    $.BODY,
+    $.BR,
+    $.CENTER,
+    $.CODE,
+    $.DD,
+    $.DIV,
+    $.DL,
+    $.DT,
+    $.EM,
+    $.EMBED,
+    $.H1,
+    $.H2,
+    $.H3,
+    $.H4,
+    $.H5,
+    $.H6,
+    $.HEAD,
+    $.HR,
+    $.I,
+    $.IMG,
+    $.LI,
+    $.LISTING,
+    $.MENU,
+    $.META,
+    $.NOBR,
+    $.OL,
+    $.P,
+    $.PRE,
+    $.RUBY,
+    $.S,
+    $.SMALL,
+    $.SPAN,
+    $.STRONG,
+    $.STRIKE,
+    $.SUB,
+    $.SUP,
+    $.TABLE,
+    $.TT,
+    $.U,
+    $.UL,
+    $.VAR,
+]);
+//Check exit from foreign content
+export function causesExit(startTagToken) {
+    const tn = startTagToken.tagID;
+    const isFontWithAttrs = tn === $.FONT &&
+        startTagToken.attrs.some(({ name }) => name === ATTRS.COLOR || name === ATTRS.SIZE || name === ATTRS.FACE);
+    return isFontWithAttrs || EXITS_FOREIGN_CONTENT.has(tn);
+}
+//Token adjustments
+export function adjustTokenMathMLAttrs(token) {
+    for (let i = 0; i < token.attrs.length; i++) {
+        if (token.attrs[i].name === DEFINITION_URL_ATTR) {
+            token.attrs[i].name = ADJUSTED_DEFINITION_URL_ATTR;
+            break;
+        }
+    }
+}
+export function adjustTokenSVGAttrs(token) {
+    for (let i = 0; i < token.attrs.length; i++) {
+        const adjustedAttrName = SVG_ATTRS_ADJUSTMENT_MAP.get(token.attrs[i].name);
+        if (adjustedAttrName != null) {
+            token.attrs[i].name = adjustedAttrName;
+        }
+    }
+}
+export function adjustTokenXMLAttrs(token) {
+    for (let i = 0; i < token.attrs.length; i++) {
+        const adjustedAttrEntry = XML_ATTRS_ADJUSTMENT_MAP.get(token.attrs[i].name);
+        if (adjustedAttrEntry) {
+            token.attrs[i].prefix = adjustedAttrEntry.prefix;
+            token.attrs[i].name = adjustedAttrEntry.name;
+            token.attrs[i].namespace = adjustedAttrEntry.namespace;
+        }
+    }
+}
+export function adjustTokenSVGTagName(token) {
+    const adjustedTagName = SVG_TAG_NAMES_ADJUSTMENT_MAP.get(token.tagName);
+    if (adjustedTagName != null) {
+        token.tagName = adjustedTagName;
+        token.tagID = getTagID(token.tagName);
+    }
+}
+//Integration points
+function isMathMLTextIntegrationPoint(tn, ns) {
+    return ns === NS.MATHML && (tn === $.MI || tn === $.MO || tn === $.MN || tn === $.MS || tn === $.MTEXT);
+}
+function isHtmlIntegrationPoint(tn, ns, attrs) {
+    if (ns === NS.MATHML && tn === $.ANNOTATION_XML) {
+        for (let i = 0; i < attrs.length; i++) {
+            if (attrs[i].name === ATTRS.ENCODING) {
+                const value = attrs[i].value.toLowerCase();
+                return value === MIME_TYPES.TEXT_HTML || value === MIME_TYPES.APPLICATION_XML;
+            }
+        }
+    }
+    return ns === NS.SVG && (tn === $.FOREIGN_OBJECT || tn === $.DESC || tn === $.TITLE);
+}
+export function isIntegrationPoint(tn, ns, attrs, foreignNS) {
+    return (((!foreignNS || foreignNS === NS.HTML) && isHtmlIntegrationPoint(tn, ns, attrs)) ||
+        ((!foreignNS || foreignNS === NS.MATHML) && isMathMLTextIntegrationPoint(tn, ns)));
+}

package/dist/common/html.d.ts

@@ -0,0 +1,289 @@
+/** All valid namespaces in HTML. */
+export declare enum NS {
+    HTML = "http://www.w3.org/1999/xhtml",
+    MATHML = "http://www.w3.org/1998/Math/MathML",
+    SVG = "http://www.w3.org/2000/svg",
+    XLINK = "http://www.w3.org/1999/xlink",
+    XML = "http://www.w3.org/XML/1998/namespace",
+    XMLNS = "http://www.w3.org/2000/xmlns/"
+}
+export declare enum ATTRS {
+    TYPE = "type",
+    ACTION = "action",
+    ENCODING = "encoding",
+    PROMPT = "prompt",
+    NAME = "name",
+    COLOR = "color",
+    FACE = "face",
+    SIZE = "size"
+}
+/**
+ * The mode of the document.
+ *
+ * @see {@link https://dom.spec.whatwg.org/#concept-document-limited-quirks}
+ */
+export declare enum DOCUMENT_MODE {
+    NO_QUIRKS = "no-quirks",
+    QUIRKS = "quirks",
+    LIMITED_QUIRKS = "limited-quirks"
+}
+export declare enum TAG_NAMES {
+    A = "a",
+    ADDRESS = "address",
+    ANNOTATION_XML = "annotation-xml",
+    APPLET = "applet",
+    AREA = "area",
+    ARTICLE = "article",
+    ASIDE = "aside",
+    B = "b",
+    BASE = "base",
+    BASEFONT = "basefont",
+    BGSOUND = "bgsound",
+    BIG = "big",
+    BLOCKQUOTE = "blockquote",
+    BODY = "body",
+    BR = "br",
+    BUTTON = "button",
+    CAPTION = "caption",
+    CENTER = "center",
+    CODE = "code",
+    COL = "col",
+    COLGROUP = "colgroup",
+    DD = "dd",
+    DESC = "desc",
+    DETAILS = "details",
+    DIALOG = "dialog",
+    DIR = "dir",
+    DIV = "div",
+    DL = "dl",
+    DT = "dt",
+    EM = "em",
+    EMBED = "embed",
+    FIELDSET = "fieldset",
+    FIGCAPTION = "figcaption",
+    FIGURE = "figure",
+    FONT = "font",
+    FOOTER = "footer",
+    FOREIGN_OBJECT = "foreignObject",
+    FORM = "form",
+    FRAME = "frame",
+    FRAMESET = "frameset",
+    H1 = "h1",
+    H2 = "h2",
+    H3 = "h3",
+    H4 = "h4",
+    H5 = "h5",
+    H6 = "h6",
+    HEAD = "head",
+    HEADER = "header",
+    HGROUP = "hgroup",
+    HR = "hr",
+    HTML = "html",
+    I = "i",
+    IMG = "img",
+    IMAGE = "image",
+    INPUT = "input",
+    IFRAME = "iframe",
+    KEYGEN = "keygen",
+    LABEL = "label",
+    LI = "li",
+    LINK = "link",
+    LISTING = "listing",
+    MAIN = "main",
+    MALIGNMARK = "malignmark",
+    MARQUEE = "marquee",
+    MATH = "math",
+    MENU = "menu",
+    META = "meta",
+    MGLYPH = "mglyph",
+    MI = "mi",
+    MO = "mo",
+    MN = "mn",
+    MS = "ms",
+    MTEXT = "mtext",
+    NAV = "nav",
+    NOBR = "nobr",
+    NOFRAMES = "noframes",
+    NOEMBED = "noembed",
+    NOSCRIPT = "noscript",
+    OBJECT = "object",
+    OL = "ol",
+    OPTGROUP = "optgroup",
+    OPTION = "option",
+    P = "p",
+    PARAM = "param",
+    PLAINTEXT = "plaintext",
+    PRE = "pre",
+    RB = "rb",
+    RP = "rp",
+    RT = "rt",
+    RTC = "rtc",
+    RUBY = "ruby",
+    S = "s",
+    SCRIPT = "script",
+    SEARCH = "search",
+    SECTION = "section",
+    SELECT = "select",
+    SOURCE = "source",
+    SMALL = "small",
+    SPAN = "span",
+    STRIKE = "strike",
+    STRONG = "strong",
+    STYLE = "style",
+    SUB = "sub",
+    SUMMARY = "summary",
+    SUP = "sup",
+    TABLE = "table",
+    TBODY = "tbody",
+    TEMPLATE = "template",
+    TEXTAREA = "textarea",
+    TFOOT = "tfoot",
+    TD = "td",
+    TH = "th",
+    THEAD = "thead",
+    TITLE = "title",
+    TR = "tr",
+    TRACK = "track",
+    TT = "tt",
+    U = "u",
+    UL = "ul",
+    SVG = "svg",
+    VAR = "var",
+    WBR = "wbr",
+    XMP = "xmp"
+}
+/**
+ * Tag IDs are numeric IDs for known tag names.
+ *
+ * We use tag IDs to improve the performance of tag name comparisons.
+ */
+export declare enum TAG_ID {
+    UNKNOWN = 0,
+    A = 1,
+    ADDRESS = 2,
+    ANNOTATION_XML = 3,
+    APPLET = 4,
+    AREA = 5,
+    ARTICLE = 6,
+    ASIDE = 7,
+    B = 8,
+    BASE = 9,
+    BASEFONT = 10,
+    BGSOUND = 11,
+    BIG = 12,
+    BLOCKQUOTE = 13,
+    BODY = 14,
+    BR = 15,
+    BUTTON = 16,
+    CAPTION = 17,
+    CENTER = 18,
+    CODE = 19,
+    COL = 20,
+    COLGROUP = 21,
+    DD = 22,
+    DESC = 23,
+    DETAILS = 24,
+    DIALOG = 25,
+    DIR = 26,
+    DIV = 27,
+    DL = 28,
+    DT = 29,
+    EM = 30,
+    EMBED = 31,
+    FIELDSET = 32,
+    FIGCAPTION = 33,
+    FIGURE = 34,
+    FONT = 35,
+    FOOTER = 36,
+    FOREIGN_OBJECT = 37,
+    FORM = 38,
+    FRAME = 39,
+    FRAMESET = 40,
+    H1 = 41,
+    H2 = 42,
+    H3 = 43,
+    H4 = 44,
+    H5 = 45,
+    H6 = 46,
+    HEAD = 47,
+    HEADER = 48,
+    HGROUP = 49,
+    HR = 50,
+    HTML = 51,
+    I = 52,
+    IMG = 53,
+    IMAGE = 54,
+    INPUT = 55,
+    IFRAME = 56,
+    KEYGEN = 57,
+    LABEL = 58,
+    LI = 59,
+    LINK = 60,
+    LISTING = 61,
+    MAIN = 62,
+    MALIGNMARK = 63,
+    MARQUEE = 64,
+    MATH = 65,
+    MENU = 66,
+    META = 67,
+    MGLYPH = 68,
+    MI = 69,
+    MO = 70,
+    MN = 71,
+    MS = 72,
+    MTEXT = 73,
+    NAV = 74,
+    NOBR = 75,
+    NOFRAMES = 76,
+    NOEMBED = 77,
+    NOSCRIPT = 78,
+    OBJECT = 79,
+    OL = 80,
+    OPTGROUP = 81,
+    OPTION = 82,
+    P = 83,
+    PARAM = 84,
+    PLAINTEXT = 85,
+    PRE = 86,
+    RB = 87,
+    RP = 88,
+    RT = 89,
+    RTC = 90,
+    RUBY = 91,
+    S = 92,
+    SCRIPT = 93,
+    SEARCH = 94,
+    SECTION = 95,
+    SELECT = 96,
+    SOURCE = 97,
+    SMALL = 98,
+    SPAN = 99,
+    STRIKE = 100,
+    STRONG = 101,
+    STYLE = 102,
+    SUB = 103,
+    SUMMARY = 104,
+    SUP = 105,
+    TABLE = 106,
+    TBODY = 107,
+    TEMPLATE = 108,
+    TEXTAREA = 109,
+    TFOOT = 110,
+    TD = 111,
+    TH = 112,
+    THEAD = 113,
+    TITLE = 114,
+    TR = 115,
+    TRACK = 116,
+    TT = 117,
+    U = 118,
+    UL = 119,
+    SVG = 120,
+    VAR = 121,
+    WBR = 122,
+    XMP = 123
+}
+export declare function getTagID(tagName: string): TAG_ID;
+export declare const SPECIAL_ELEMENTS: Record<NS, Set<TAG_ID>>;
+export declare const NUMBERED_HEADERS: Set<TAG_ID>;
+export declare function hasUnescapedText(tn: string, scriptingEnabled: boolean): boolean;