readability_js 0.0.1

data/lib/readability_js/node/node_modules/css-tree/lib/tokenizer/utils.js

@@ -0,0 +1,254 @@
+import {
+    isDigit,
+    isHexDigit,
+    isUppercaseLetter,
+    isName,
+    isWhiteSpace,
+    isValidEscape
+} from './char-code-definitions.js';
+
+function getCharCode(source, offset) {
+    return offset < source.length ? source.charCodeAt(offset) : 0;
+}
+
+export function getNewlineLength(source, offset, code) {
+    if (code === 13 /* \r */ && getCharCode(source, offset + 1) === 10 /* \n */) {
+        return 2;
+    }
+
+    return 1;
+}
+
+export function cmpChar(testStr, offset, referenceCode) {
+    let code = testStr.charCodeAt(offset);
+
+    // code.toLowerCase() for A..Z
+    if (isUppercaseLetter(code)) {
+        code = code | 32;
+    }
+
+    return code === referenceCode;
+}
+
+export function cmpStr(testStr, start, end, referenceStr) {
+    if (end - start !== referenceStr.length) {
+        return false;
+    }
+
+    if (start < 0 || end > testStr.length) {
+        return false;
+    }
+
+    for (let i = start; i < end; i++) {
+        const referenceCode = referenceStr.charCodeAt(i - start);
+        let testCode = testStr.charCodeAt(i);
+
+        // testCode.toLowerCase() for A..Z
+        if (isUppercaseLetter(testCode)) {
+            testCode = testCode | 32;
+        }
+
+        if (testCode !== referenceCode) {
+            return false;
+        }
+    }
+
+    return true;
+}
+
+export function findWhiteSpaceStart(source, offset) {
+    for (; offset >= 0; offset--) {
+        if (!isWhiteSpace(source.charCodeAt(offset))) {
+            break;
+        }
+    }
+
+    return offset + 1;
+}
+
+export function findWhiteSpaceEnd(source, offset) {
+    for (; offset < source.length; offset++) {
+        if (!isWhiteSpace(source.charCodeAt(offset))) {
+            break;
+        }
+    }
+
+    return offset;
+}
+
+export function findDecimalNumberEnd(source, offset) {
+    for (; offset < source.length; offset++) {
+        if (!isDigit(source.charCodeAt(offset))) {
+            break;
+        }
+    }
+
+    return offset;
+}
+
+// § 4.3.7. Consume an escaped code point
+export function consumeEscaped(source, offset) {
+    // It assumes that the U+005C REVERSE SOLIDUS (\) has already been consumed and
+    // that the next input code point has already been verified to be part of a valid escape.
+    offset += 2;
+
+    // hex digit
+    if (isHexDigit(getCharCode(source, offset - 1))) {
+        // Consume as many hex digits as possible, but no more than 5.
+        // Note that this means 1-6 hex digits have been consumed in total.
+        for (const maxOffset = Math.min(source.length, offset + 5); offset < maxOffset; offset++) {
+            if (!isHexDigit(getCharCode(source, offset))) {
+                break;
+            }
+        }
+
+        // If the next input code point is whitespace, consume it as well.
+        const code = getCharCode(source, offset);
+        if (isWhiteSpace(code)) {
+            offset += getNewlineLength(source, offset, code);
+        }
+    }
+
+    return offset;
+}
+
+// §4.3.11. Consume a name
+// Note: This algorithm does not do the verification of the first few code points that are necessary
+// to ensure the returned code points would constitute an <ident-token>. If that is the intended use,
+// ensure that the stream starts with an identifier before calling this algorithm.
+export function consumeName(source, offset) {
+    // Let result initially be an empty string.
+    // Repeatedly consume the next input code point from the stream:
+    for (; offset < source.length; offset++) {
+        const code = source.charCodeAt(offset);
+
+        // name code point
+        if (isName(code)) {
+            // Append the code point to result.
+            continue;
+        }
+
+        // the stream starts with a valid escape
+        if (isValidEscape(code, getCharCode(source, offset + 1))) {
+            // Consume an escaped code point. Append the returned code point to result.
+            offset = consumeEscaped(source, offset) - 1;
+            continue;
+        }
+
+        // anything else
+        // Reconsume the current input code point. Return result.
+        break;
+    }
+
+    return offset;
+}
+
+// §4.3.12. Consume a number
+export function consumeNumber(source, offset) {
+    let code = source.charCodeAt(offset);
+
+    // 2. If the next input code point is U+002B PLUS SIGN (+) or U+002D HYPHEN-MINUS (-),
+    // consume it and append it to repr.
+    if (code === 0x002B || code === 0x002D) {
+        code = source.charCodeAt(offset += 1);
+    }
+
+    // 3. While the next input code point is a digit, consume it and append it to repr.
+    if (isDigit(code)) {
+        offset = findDecimalNumberEnd(source, offset + 1);
+        code = source.charCodeAt(offset);
+    }
+
+    // 4. If the next 2 input code points are U+002E FULL STOP (.) followed by a digit, then:
+    if (code === 0x002E && isDigit(source.charCodeAt(offset + 1))) {
+        // 4.1 Consume them.
+        // 4.2 Append them to repr.
+        offset += 2;
+
+        // 4.3 Set type to "number".
+        // TODO
+
+        // 4.4 While the next input code point is a digit, consume it and append it to repr.
+
+        offset = findDecimalNumberEnd(source, offset);
+    }
+
+    // 5. If the next 2 or 3 input code points are U+0045 LATIN CAPITAL LETTER E (E)
+    // or U+0065 LATIN SMALL LETTER E (e), ... , followed by a digit, then:
+    if (cmpChar(source, offset, 101 /* e */)) {
+        let sign = 0;
+        code = source.charCodeAt(offset + 1);
+
+        // ... optionally followed by U+002D HYPHEN-MINUS (-) or U+002B PLUS SIGN (+) ...
+        if (code === 0x002D || code === 0x002B) {
+            sign = 1;
+            code = source.charCodeAt(offset + 2);
+        }
+
+        // ... followed by a digit
+        if (isDigit(code)) {
+            // 5.1 Consume them.
+            // 5.2 Append them to repr.
+
+            // 5.3 Set type to "number".
+            // TODO
+
+            // 5.4 While the next input code point is a digit, consume it and append it to repr.
+            offset = findDecimalNumberEnd(source, offset + 1 + sign + 1);
+        }
+    }
+
+    return offset;
+}
+
+// § 4.3.14. Consume the remnants of a bad url
+// ... its sole use is to consume enough of the input stream to reach a recovery point
+// where normal tokenizing can resume.
+export function consumeBadUrlRemnants(source, offset) {
+    // Repeatedly consume the next input code point from the stream:
+    for (; offset < source.length; offset++) {
+        const code = source.charCodeAt(offset);
+
+        // U+0029 RIGHT PARENTHESIS ())
+        // EOF
+        if (code === 0x0029) {
+            // Return.
+            offset++;
+            break;
+        }
+
+        if (isValidEscape(code, getCharCode(source, offset + 1))) {
+            // Consume an escaped code point.
+            // Note: This allows an escaped right parenthesis ("\)") to be encountered
+            // without ending the <bad-url-token>. This is otherwise identical to
+            // the "anything else" clause.
+            offset = consumeEscaped(source, offset);
+        }
+    }
+
+    return offset;
+}
+
+// § 4.3.7. Consume an escaped code point
+// Note: This algorithm assumes that escaped is valid without leading U+005C REVERSE SOLIDUS (\)
+export function decodeEscaped(escaped) {
+    // Single char escaped that's not a hex digit
+    if (escaped.length === 1 && !isHexDigit(escaped.charCodeAt(0))) {
+        return escaped[0];
+    }
+
+    // Interpret the hex digits as a hexadecimal number.
+    let code = parseInt(escaped, 16);
+
+    if (
+        (code === 0) ||                       // If this number is zero,
+        (code >= 0xD800 && code <= 0xDFFF) || // or is for a surrogate,
+        (code > 0x10FFFF)                     // or is greater than the maximum allowed code point
+    ) {
+        // ... return U+FFFD REPLACEMENT CHARACTER
+        code = 0xFFFD;
+    }
+
+    // Otherwise, return the code point with that value.
+    return String.fromCodePoint(code);
+}

data/lib/readability_js/node/node_modules/css-tree/lib/utils/List.js

@@ -0,0 +1,469 @@
+//
+//                              list
+//                            ┌──────┐
+//             ┌──────────────┼─head │
+//             │              │ tail─┼──────────────┐
+//             │              └──────┘              │
+//             ▼                                    ▼
+//            item        item        item        item
+//          ┌──────┐    ┌──────┐    ┌──────┐    ┌──────┐
+//  null ◀──┼─prev │◀───┼─prev │◀───┼─prev │◀───┼─prev │
+//          │ next─┼───▶│ next─┼───▶│ next─┼───▶│ next─┼──▶ null
+//          ├──────┤    ├──────┤    ├──────┤    ├──────┤
+//          │ data │    │ data │    │ data │    │ data │
+//          └──────┘    └──────┘    └──────┘    └──────┘
+//
+
+let releasedCursors = null;
+
+export class List {
+    static createItem(data) {
+        return {
+            prev: null,
+            next: null,
+            data
+        };
+    }
+
+    constructor() {
+        this.head = null;
+        this.tail = null;
+        this.cursor = null;
+    }
+    createItem(data) {
+        return List.createItem(data);
+    }
+
+    // cursor helpers
+    allocateCursor(prev, next) {
+        let cursor;
+
+        if (releasedCursors !== null) {
+            cursor = releasedCursors;
+            releasedCursors = releasedCursors.cursor;
+            cursor.prev = prev;
+            cursor.next = next;
+            cursor.cursor = this.cursor;
+        } else {
+            cursor = {
+                prev,
+                next,
+                cursor: this.cursor
+            };
+        }
+
+        this.cursor = cursor;
+
+        return cursor;
+    }
+    releaseCursor() {
+        const { cursor } = this;
+
+        this.cursor = cursor.cursor;
+        cursor.prev = null;
+        cursor.next = null;
+        cursor.cursor = releasedCursors;
+        releasedCursors = cursor;
+    }
+    updateCursors(prevOld, prevNew, nextOld, nextNew) {
+        let { cursor } = this;
+
+        while (cursor !== null) {
+            if (cursor.prev === prevOld) {
+                cursor.prev = prevNew;
+            }
+
+            if (cursor.next === nextOld) {
+                cursor.next = nextNew;
+            }
+
+            cursor = cursor.cursor;
+        }
+    }
+    *[Symbol.iterator]() {
+        for (let cursor = this.head; cursor !== null; cursor = cursor.next) {
+            yield cursor.data;
+        }
+    }
+
+    // getters
+    get size() {
+        let size = 0;
+
+        for (let cursor = this.head; cursor !== null; cursor = cursor.next) {
+            size++;
+        }
+
+        return size;
+    }
+    get isEmpty() {
+        return this.head === null;
+    }
+    get first() {
+        return this.head && this.head.data;
+    }
+    get last() {
+        return this.tail && this.tail.data;
+    }
+
+    // convertors
+    fromArray(array) {
+        let cursor = null;
+        this.head = null;
+
+        for (let data of array) {
+            const item = List.createItem(data);
+
+            if (cursor !== null) {
+                cursor.next = item;
+            } else {
+                this.head = item;
+            }
+
+            item.prev = cursor;
+            cursor = item;
+        }
+
+        this.tail = cursor;
+        return this;
+    }
+    toArray() {
+        return [...this];
+    }
+    toJSON() {
+        return [...this];
+    }
+
+    // array-like methods
+    forEach(fn, thisArg = this) {
+        // push cursor
+        const cursor = this.allocateCursor(null, this.head);
+
+        while (cursor.next !== null) {
+            const item = cursor.next;
+            cursor.next = item.next;
+            fn.call(thisArg, item.data, item, this);
+        }
+
+        // pop cursor
+        this.releaseCursor();
+    }
+    forEachRight(fn, thisArg = this) {
+        // push cursor
+        const cursor = this.allocateCursor(this.tail, null);
+
+        while (cursor.prev !== null) {
+            const item = cursor.prev;
+            cursor.prev = item.prev;
+            fn.call(thisArg, item.data, item, this);
+        }
+
+        // pop cursor
+        this.releaseCursor();
+    }
+    reduce(fn, initialValue, thisArg = this) {
+        // push cursor
+        let cursor = this.allocateCursor(null, this.head);
+        let acc = initialValue;
+        let item;
+
+        while (cursor.next !== null) {
+            item = cursor.next;
+            cursor.next = item.next;
+
+            acc = fn.call(thisArg, acc, item.data, item, this);
+        }
+
+        // pop cursor
+        this.releaseCursor();
+
+        return acc;
+    }
+    reduceRight(fn, initialValue, thisArg = this) {
+        // push cursor
+        let cursor = this.allocateCursor(this.tail, null);
+        let acc = initialValue;
+        let item;
+
+        while (cursor.prev !== null) {
+            item = cursor.prev;
+            cursor.prev = item.prev;
+
+            acc = fn.call(thisArg, acc, item.data, item, this);
+        }
+
+        // pop cursor
+        this.releaseCursor();
+
+        return acc;
+    }
+    some(fn, thisArg = this) {
+        for (let cursor = this.head; cursor !== null; cursor = cursor.next) {
+            if (fn.call(thisArg, cursor.data, cursor, this)) {
+                return true;
+            }
+        }
+
+        return false;
+    }
+    map(fn, thisArg = this) {
+        const result = new List();
+
+        for (let cursor = this.head; cursor !== null; cursor = cursor.next) {
+            result.appendData(fn.call(thisArg, cursor.data, cursor, this));
+        }
+
+        return result;
+    }
+    filter(fn, thisArg = this) {
+        const result = new List();
+
+        for (let cursor = this.head; cursor !== null; cursor = cursor.next) {
+            if (fn.call(thisArg, cursor.data, cursor, this)) {
+                result.appendData(cursor.data);
+            }
+        }
+
+        return result;
+    }
+
+    nextUntil(start, fn, thisArg = this) {
+        if (start === null) {
+            return;
+        }
+
+        // push cursor
+        const cursor = this.allocateCursor(null, start);
+
+        while (cursor.next !== null) {
+            const item = cursor.next;
+            cursor.next = item.next;
+            if (fn.call(thisArg, item.data, item, this)) {
+                break;
+            }
+        }
+
+        // pop cursor
+        this.releaseCursor();
+    }
+    prevUntil(start, fn, thisArg = this) {
+        if (start === null) {
+            return;
+        }
+
+        // push cursor
+        const cursor = this.allocateCursor(start, null);
+
+        while (cursor.prev !== null) {
+            const item = cursor.prev;
+            cursor.prev = item.prev;
+            if (fn.call(thisArg, item.data, item, this)) {
+                break;
+            }
+        }
+
+        // pop cursor
+        this.releaseCursor();
+    }
+
+    // mutation
+    clear() {
+        this.head = null;
+        this.tail = null;
+    }
+    copy() {
+        const result = new List();
+
+        for (let data of this) {
+            result.appendData(data);
+        }
+
+        return result;
+    }
+    prepend(item) {
+        //      head
+        //    ^
+        // item
+        this.updateCursors(null, item, this.head, item);
+
+        // insert to the beginning of the list
+        if (this.head !== null) {
+            // new item <- first item
+            this.head.prev = item;
+            // new item -> first item
+            item.next = this.head;
+        } else {
+            // if list has no head, then it also has no tail
+            // in this case tail points to the new item
+            this.tail = item;
+        }
+
+        // head always points to new item
+        this.head = item;
+        return this;
+    }
+    prependData(data) {
+        return this.prepend(List.createItem(data));
+    }
+    append(item) {
+        return this.insert(item);
+    }
+    appendData(data) {
+        return this.insert(List.createItem(data));
+    }
+    insert(item, before = null) {
+        if (before !== null) {
+            // prev   before
+            //      ^
+            //     item
+            this.updateCursors(before.prev, item, before, item);
+
+            if (before.prev === null) {
+                // insert to the beginning of list
+                if (this.head !== before) {
+                    throw new Error('before doesn\'t belong to list');
+                }
+                // since head points to before therefore list doesn't empty
+                // no need to check tail
+                this.head = item;
+                before.prev = item;
+                item.next = before;
+                this.updateCursors(null, item);
+            } else {
+                // insert between two items
+                before.prev.next = item;
+                item.prev = before.prev;
+                before.prev = item;
+                item.next = before;
+            }
+        } else {
+            // tail
+            //      ^
+            //      item
+            this.updateCursors(this.tail, item, null, item);
+
+            // insert to the ending of the list
+            if (this.tail !== null) {
+                // last item -> new item
+                this.tail.next = item;
+                // last item <- new item
+                item.prev = this.tail;
+            } else {
+                // if list has no tail, then it also has no head
+                // in this case head points to new item
+                this.head = item;
+            }
+
+            // tail always points to new item
+            this.tail = item;
+        }
+
+        return this;
+    }
+    insertData(data, before) {
+        return this.insert(List.createItem(data), before);
+    }
+    remove(item) {
+        //      item
+        //       ^
+        // prev     next
+        this.updateCursors(item, item.prev, item, item.next);
+
+        if (item.prev !== null) {
+            item.prev.next = item.next;
+        } else {
+            if (this.head !== item) {
+                throw new Error('item doesn\'t belong to list');
+            }
+
+            this.head = item.next;
+        }
+
+        if (item.next !== null) {
+            item.next.prev = item.prev;
+        } else {
+            if (this.tail !== item) {
+                throw new Error('item doesn\'t belong to list');
+            }
+
+            this.tail = item.prev;
+        }
+
+        item.prev = null;
+        item.next = null;
+
+        return item;
+    }
+    push(data) {
+        this.insert(List.createItem(data));
+    }
+    pop() {
+        return this.tail !== null ? this.remove(this.tail) : null;
+    }
+    unshift(data) {
+        this.prepend(List.createItem(data));
+    }
+    shift() {
+        return this.head !== null ? this.remove(this.head) : null;
+    }
+    prependList(list) {
+        return this.insertList(list, this.head);
+    }
+    appendList(list) {
+        return this.insertList(list);
+    }
+    insertList(list, before) {
+        // ignore empty lists
+        if (list.head === null) {
+            return this;
+        }
+
+        if (before !== undefined && before !== null) {
+            this.updateCursors(before.prev, list.tail, before, list.head);
+
+            // insert in the middle of dist list
+            if (before.prev !== null) {
+                // before.prev <-> list.head
+                before.prev.next = list.head;
+                list.head.prev = before.prev;
+            } else {
+                this.head = list.head;
+            }
+
+            before.prev = list.tail;
+            list.tail.next = before;
+        } else {
+            this.updateCursors(this.tail, list.tail, null, list.head);
+
+            // insert to end of the list
+            if (this.tail !== null) {
+                // if destination list has a tail, then it also has a head,
+                // but head doesn't change
+                // dest tail -> source head
+                this.tail.next = list.head;
+                // dest tail <- source head
+                list.head.prev = this.tail;
+            } else {
+                // if list has no a tail, then it also has no a head
+                // in this case points head to new item
+                this.head = list.head;
+            }
+
+            // tail always start point to new item
+            this.tail = list.tail;
+        }
+
+        list.head = null;
+        list.tail = null;
+        return this;
+    }
+    replace(oldItem, newItemOrList) {
+        if ('head' in newItemOrList) {
+            this.insertList(newItemOrList, oldItem);
+        } else {
+            this.insert(newItemOrList, oldItem);
+        }
+
+        this.remove(oldItem);
+    }
+}