vscode-xml/src/services/XmlFormatter.ts

'use strict';

// Based on pretty-data (https://github.com/vkiryukhin/pretty-data)
export class XmlFormatter {
    constructor(options?: IXmlFormatterOptions) {
        options = options || {};
        
        if (typeof options.preferSpaces === 'undefined') {
            options.preferSpaces = false;
        }
        
        if (typeof options.splitNamespaces === 'undefined') {
            options.splitNamespaces = true;
        }
        
        options.tabSize = options.tabSize || 4;
        options.newLine = options.newLine || '\n';
        
        this.newLine = options.newLine || '\n';
        this.indentPattern = (options.preferSpaces) ? ' '.repeat(options.tabSize) : '\t';
        this.splitNamespaces = options.splitNamespaces;
    }
    
    newLine: string;
    indentPattern: string;
    splitNamespaces: boolean;
    
    format(xml: string): string {
        xml = this.minify(xml, false);
        xml = xml.replace(/</g, '~::~<');
        
        if (this.splitNamespaces) {
            xml = xml
                .replace(/xmlns\:/g, '~::~xmlns:')
                .replace(/xmlns\=/g, '~::~xmlns=');
        }
        
        let parts: string[] = xml.split('~::~');
            console.log(parts);
        let inComment: boolean = false;
        let level: number = 0;
        let output: string = '';

        for (let i = 0; i < parts.length; i++) {
            // <!
            if (parts[i].search(/<!/) > -1) {
                output += this._getIndent(level, parts[i]);
                inComment = true;
                
                // end <!
                if (parts[i].search(/-->/) > -1 || parts[i].search(/\]>/) > -1 || parts[i].search(/!DOCTYPE/) > -1) {
                    inComment = false;
                }
            }
            
            // end <!
            else if (parts[i].search(/-->/) > -1 || parts[i].search(/\]>/) > -1) {
                output += parts[i];
                inComment = false;
            }
            
            // <elm></elm>
            else if (/^<(\w|:)/.test(parts[i - 1]) && /^<\/(\w|:)/.test(parts[i])
                && /^<[\w:\-\.\,\/]+/.exec(parts[i - 1])[0] == /^<\/[\w:\-\.\,]+/.exec(parts[i])[0].replace('/', '')) {

                output += parts[i];
                if (!inComment) level--;
            }
            
            // <elm>
            else if (parts[i].search(/<(\w|:)/) > -1 && parts[i].search(/<\//) == -1 && parts[i].search(/\/>/) == -1) {
                output = (!inComment) ? output += this._getIndent(level++, parts[i]) : output += parts[i];
            }
            
            // <elm>...</elm>
            else if (parts[i].search(/<(\w|:)/) > -1 && parts[i].search(/<\//) > -1) {
                output = (!inComment) ? output += this._getIndent(level, parts[i]) : output += parts[i];
            }
            
            // </elm>
            else if (parts[i].search(/<\//) > -1) {
                output = (!inComment) ? output += this._getIndent(--level, parts[i]) : output += parts[i];
            }
            
            // <elm />
            else if (parts[i].search(/\/>/) > -1 && (!this.splitNamespaces || parts[i].search(/xmlns(:|=)/) == -1)) {
                output = (!inComment) ? output += this._getIndent(level, parts[i]) : output += parts[i];
            }
            
            // xmlns />
            else if (parts[i].search(/\/>/) > -1 && parts[i].search(/xmlns(:|=)/) > -1 && this.splitNamespaces) {
                output = (!inComment) ? output += this._getIndent(level--, parts[i]) : output += parts[i];
            }
            
            // <?xml ... ?>
            else if (parts[i].search(/<\?/) > -1) {
                output += this._getIndent(level, parts[i]);
            }
            
            // xmlns
            else if (this.splitNamespaces && (parts[i].search(/xmlns\:/) > -1 || parts[i].search(/xmlns\=/) > -1)) {
                output += this._getIndent(level, parts[i]);
            }
            
            else {
                output += parts[i];
            }
        }
        
        // remove leading newline
        if (output[0] == this.newLine) {
            output = output.slice(1);
        }
            
        else if (output.substring(0, 1) == this.newLine) {
            output = output.slice(2);
        }
        
        return output;
    }
    
    minify(xml: string, removeComments?: boolean): string {
        if (typeof removeComments === 'undefined') {
            removeComments = false;
        }
        
        xml = this._stripLineBreaks(xml); // all line breaks outside of CDATA elements
        xml = (removeComments) ? xml.replace(/\<![ \r\n\t]*(--([^\-]|[\r\n]|-[^\-])*--[ \r\n\t]*)\>/g, '') : xml;
        xml = xml.replace(/>\s{0,}</g, '><'); // insignificant whitespace between tags
        xml = xml.replace(/"\s+(?=[^\s]+=)/g, '" '); // spaces between attributes
        xml = xml.replace(/"\s+(?=>)/g, '"'); // spaces between the last attribute and tag close (>)
        xml = xml.replace(/"\s+(?=\/>)/g, '" '); // spaces between the last attribute and tag close (/>)
        xml = xml.replace(/[^ <>="]\s+[^ <>="]+=/g, (match: string) => { // spaces between the node name and the first attribute
            return match.replace(/\s+/g, ' ');
        });
        
        return xml;
    }
    
    private _getIndent(level: number, trailingValue?: string): string {
        trailingValue = trailingValue || '';
        
        return `${this.newLine}${this.indentPattern.repeat(level)}${trailingValue}`;
    }
    
    private _stripLineBreaks(xml: string): string {
        let output: string = '';
        let inTag: boolean = false;
        let inTagName: boolean = false;
        let inCdata: boolean = false;
        let inAttribute: boolean = false;
        
        for (let i = 0; i < xml.length; i++) {
            let char: string = xml.charAt(i);
            let prev: string = xml.charAt(i - 1);
            let next: string = xml.charAt(i + 1);
            
            if (char == '!' && (xml.substr(i, 8) == '![CDATA[' || xml.substr(i, 3) == '!--')) {
                inCdata = true;
            }
            
            else if (char == ']' && (xml.substr(i, 3) == ']]>')) {
                inCdata = false;
            }
            
            else if (char == '-' && (xml.substr(i, 3) == '-->')) {
                inCdata = false;
            }
            
            else if (char.search(/[\r\n]/g) > -1 && !inCdata) {
                if (/\r/.test(char) && /\S|\r|\n/.test(prev) && /\S|\r|\n/.test(xml.charAt(i + this.newLine.length))) {
                    output += char;
                }

                else if (/\n/.test(char) && /\S|\r|\n/.test(xml.charAt(i - this.newLine.length)) && /\S|\r|\n/.test(next)) {
                    output += char;
                }

                continue;
            }
            
            output += char;
        }
        
        return output;
    }
}

export interface IXmlFormatterOptions {
    preferSpaces?: boolean;
    tabSize?: number;
    newLine?: string;
    splitNamespaces?: boolean;
}