diff --git a/.gitignore b/.gitignore
index 9154f4c..998dabb 100644
--- a/.gitignore
+++ b/.gitignore
@@ -24,3 +24,5 @@
hs_err_pid*
replay_pid*
+#idea
+*.idea
diff --git a/README.md b/README.md
index 158dffd..8d22385 100644
--- a/README.md
+++ b/README.md
@@ -1,3 +1,3 @@
# ocr-rell-service
-"OCR-Rell-Service" онлайн-сервис по распознаванию текста,
\ No newline at end of file
+"OCR-Rell-Service" онлайн-сервис по распознаванию текста
\ No newline at end of file
diff --git a/pom.xml b/pom.xml
new file mode 100644
index 0000000..02b52ff
--- /dev/null
+++ b/pom.xml
@@ -0,0 +1,123 @@
+
+
Element | Missed Instructions | Cov. | Missed Branches | Cov. | Missed | Cxty | Missed | Lines | Missed | Methods | Missed | Classes |
Total | 5 of 8 | 37 % | 0 of 0 | n/a | 1 | 2 | 2 | 3 | 1 | 2 | 0 | 1 |
ru.rell | 37 % | n/a | 1 | 2 | 2 | 3 | 1 | 2 | 0 | 1 |
+ * + * For a fairly comprehensive set of languages see the + * README + * file that came with this source. At a minimum, the lexer should work on a + * number of languages including C and friends, Java, Python, Bash, SQL, HTML, + * XML, CSS, Javascript, and Makefiles. It works passably on Ruby, PHP and Awk + * and a subset of Perl, but, because of commenting conventions, doesn't work on + * Smalltalk, Lisp-like, or CAML-like languages without an explicit lang class. + *
+ * Usage:
} and {@code } tags in your source with
+ * {@code class=prettyprint.}
+ * You can also use the (html deprecated) {@code } tag, but the pretty
+ * printer needs to do more substantial DOM manipulations to support that, so
+ * some css styles may not be preserved.
+ *
} or {@code } element to specify the
+ * language, as in {@code }. Any class that
+ * starts with "lang-" followed by a file extension, specifies the file type.
+ * See the "lang-*.js" files in this directory for code that implements
+ * per-language file handlers.
+ *
+ * Change log:
+ * cbeust, 2006/08/22
+ *
+ * Java annotations (start with "@") are now captured as literals ("lit")
+ *
+ * @requires console
+ */
+
+// JSLint declarations
+/*global console, document, navigator, setTimeout, window */
+
+/**
+ * Split {@code prettyPrint} into multiple timeouts so as not to interfere with
+ * UI events.
+ * If set to {@code false}, {@code prettyPrint()} is synchronous.
+ */
+window['PR_SHOULD_USE_CONTINUATION'] = true;
+
+/** the number of characters between tab columns */
+window['PR_TAB_WIDTH'] = 8;
+
+/** Walks the DOM returning a properly escaped version of innerHTML.
+ * @param {Node} node
+ * @param {Array.} out output buffer that receives chunks of HTML.
+ */
+window['PR_normalizedHtml']
+
+/** Contains functions for creating and registering new language handlers.
+ * @type {Object}
+ */
+ = window['PR']
+
+/** Pretty print a chunk of code.
+ *
+ * @param {string} sourceCodeHtml code as html
+ * @return {string} code as html, but prettier
+ */
+ = window['prettyPrintOne']
+/** Find all the {@code } and {@code } tags in the DOM with
+ * {@code class=prettyprint} and prettify them.
+ * @param {Function?} opt_whenDone if specified, called when the last entry
+ * has been finished.
+ */
+ = window['prettyPrint'] = void 0;
+
+/** browser detection. @extern @returns false if not IE, otherwise the major version. */
+window['_pr_isIE6'] = function () {
+ var ieVersion = navigator && navigator.userAgent &&
+ navigator.userAgent.match(/\bMSIE ([678])\./);
+ ieVersion = ieVersion ? +ieVersion[1] : false;
+ window['_pr_isIE6'] = function () { return ieVersion; };
+ return ieVersion;
+};
+
+
+(function () {
+ // Keyword lists for various languages.
+ var FLOW_CONTROL_KEYWORDS =
+ "break continue do else for if return while ";
+ var C_KEYWORDS = FLOW_CONTROL_KEYWORDS + "auto case char const default " +
+ "double enum extern float goto int long register short signed sizeof " +
+ "static struct switch typedef union unsigned void volatile ";
+ var COMMON_KEYWORDS = C_KEYWORDS + "catch class delete false import " +
+ "new operator private protected public this throw true try typeof ";
+ var CPP_KEYWORDS = COMMON_KEYWORDS + "alignof align_union asm axiom bool " +
+ "concept concept_map const_cast constexpr decltype " +
+ "dynamic_cast explicit export friend inline late_check " +
+ "mutable namespace nullptr reinterpret_cast static_assert static_cast " +
+ "template typeid typename using virtual wchar_t where ";
+ var JAVA_KEYWORDS = COMMON_KEYWORDS +
+ "abstract boolean byte extends final finally implements import " +
+ "instanceof null native package strictfp super synchronized throws " +
+ "transient ";
+ var CSHARP_KEYWORDS = JAVA_KEYWORDS +
+ "as base by checked decimal delegate descending event " +
+ "fixed foreach from group implicit in interface internal into is lock " +
+ "object out override orderby params partial readonly ref sbyte sealed " +
+ "stackalloc string select uint ulong unchecked unsafe ushort var ";
+ var JSCRIPT_KEYWORDS = COMMON_KEYWORDS +
+ "debugger eval export function get null set undefined var with " +
+ "Infinity NaN ";
+ var PERL_KEYWORDS = "caller delete die do dump elsif eval exit foreach for " +
+ "goto if import last local my next no our print package redo require " +
+ "sub undef unless until use wantarray while BEGIN END ";
+ var PYTHON_KEYWORDS = FLOW_CONTROL_KEYWORDS + "and as assert class def del " +
+ "elif except exec finally from global import in is lambda " +
+ "nonlocal not or pass print raise try with yield " +
+ "False True None ";
+ var RUBY_KEYWORDS = FLOW_CONTROL_KEYWORDS + "alias and begin case class def" +
+ " defined elsif end ensure false in module next nil not or redo rescue " +
+ "retry self super then true undef unless until when yield BEGIN END ";
+ var SH_KEYWORDS = FLOW_CONTROL_KEYWORDS + "case done elif esac eval fi " +
+ "function in local set then until ";
+ var ALL_KEYWORDS = (
+ CPP_KEYWORDS + CSHARP_KEYWORDS + JSCRIPT_KEYWORDS + PERL_KEYWORDS +
+ PYTHON_KEYWORDS + RUBY_KEYWORDS + SH_KEYWORDS);
+
+ // token style names. correspond to css classes
+ /** token style for a string literal */
+ var PR_STRING = 'str';
+ /** token style for a keyword */
+ var PR_KEYWORD = 'kwd';
+ /** token style for a comment */
+ var PR_COMMENT = 'com';
+ /** token style for a type */
+ var PR_TYPE = 'typ';
+ /** token style for a literal value. e.g. 1, null, true. */
+ var PR_LITERAL = 'lit';
+ /** token style for a punctuation string. */
+ var PR_PUNCTUATION = 'pun';
+ /** token style for a punctuation string. */
+ var PR_PLAIN = 'pln';
+
+ /** token style for an sgml tag. */
+ var PR_TAG = 'tag';
+ /** token style for a markup declaration such as a DOCTYPE. */
+ var PR_DECLARATION = 'dec';
+ /** token style for embedded source. */
+ var PR_SOURCE = 'src';
+ /** token style for an sgml attribute name. */
+ var PR_ATTRIB_NAME = 'atn';
+ /** token style for an sgml attribute value. */
+ var PR_ATTRIB_VALUE = 'atv';
+
+ /**
+ * A class that indicates a section of markup that is not code, e.g. to allow
+ * embedding of line numbers within code listings.
+ */
+ var PR_NOCODE = 'nocode';
+
+ /** A set of tokens that can precede a regular expression literal in
+ * javascript.
+ * http://www.mozilla.org/js/language/js20/rationale/syntax.html has the full
+ * list, but I've removed ones that might be problematic when seen in
+ * languages that don't support regular expression literals.
+ *
+ * Specifically, I've removed any keywords that can't precede a regexp
+ * literal in a syntactically legal javascript program, and I've removed the
+ * "in" keyword since it's not a keyword in many languages, and might be used
+ * as a count of inches.
+ *
+ *
The link a above does not accurately describe EcmaScript rules since
+ * it fails to distinguish between (a=++/b/i) and (a++/b/i) but it works
+ * very well in practice.
+ *
+ * @private
+ */
+ var REGEXP_PRECEDER_PATTERN = function () {
+ var preceders = [
+ "!", "!=", "!==", "#", "%", "%=", "&", "&&", "&&=",
+ "&=", "(", "*", "*=", /* "+", */ "+=", ",", /* "-", */ "-=",
+ "->", /*".", "..", "...", handled below */ "/", "/=", ":", "::", ";",
+ "<", "<<", "<<=", "<=", "=", "==", "===", ">",
+ ">=", ">>", ">>=", ">>>", ">>>=", "?", "@", "[",
+ "^", "^=", "^^", "^^=", "{", "|", "|=", "||",
+ "||=", "~" /* handles =~ and !~ */,
+ "break", "case", "continue", "delete",
+ "do", "else", "finally", "instanceof",
+ "return", "throw", "try", "typeof"
+ ];
+ var pattern = '(?:^^|[+-]';
+ for (var i = 0; i < preceders.length; ++i) {
+ pattern += '|' + preceders[i].replace(/([^=<>:&a-z])/g, '\\$1');
+ }
+ pattern += ')\\s*'; // matches at end, and matches empty string
+ return pattern;
+ // CAVEAT: this does not properly handle the case where a regular
+ // expression immediately follows another since a regular expression may
+ // have flags for case-sensitivity and the like. Having regexp tokens
+ // adjacent is not valid in any language I'm aware of, so I'm punting.
+ // TODO: maybe style special characters inside a regexp as punctuation.
+ }();
+
+ // Define regexps here so that the interpreter doesn't have to create an
+ // object each time the function containing them is called.
+ // The language spec requires a new object created even if you don't access
+ // the $1 members.
+ var pr_amp = /&/g;
+ var pr_lt = //g;
+ var pr_quot = /\"/g;
+ /** like textToHtml but escapes double quotes to be attribute safe. */
+ function attribToHtml(str) {
+ return str.replace(pr_amp, '&')
+ .replace(pr_lt, '<')
+ .replace(pr_gt, '>')
+ .replace(pr_quot, '"');
+ }
+
+ /** escapest html special characters to html. */
+ function textToHtml(str) {
+ return str.replace(pr_amp, '&')
+ .replace(pr_lt, '<')
+ .replace(pr_gt, '>');
+ }
+
+
+ var pr_ltEnt = /</g;
+ var pr_gtEnt = />/g;
+ var pr_aposEnt = /'/g;
+ var pr_quotEnt = /"/g;
+ var pr_ampEnt = /&/g;
+ var pr_nbspEnt = / /g;
+ /** unescapes html to plain text. */
+ function htmlToText(html) {
+ var pos = html.indexOf('&');
+ if (pos < 0) { return html; }
+ // Handle numeric entities specially. We can't use functional substitution
+ // since that doesn't work in older versions of Safari.
+ // These should be rare since most browsers convert them to normal chars.
+ for (--pos; (pos = html.indexOf('', pos + 1)) >= 0;) {
+ var end = html.indexOf(';', pos);
+ if (end >= 0) {
+ var num = html.substring(pos + 3, end);
+ var radix = 10;
+ if (num && num.charAt(0) === 'x') {
+ num = num.substring(1);
+ radix = 16;
+ }
+ var codePoint = parseInt(num, radix);
+ if (!isNaN(codePoint)) {
+ html = (html.substring(0, pos) + String.fromCharCode(codePoint) +
+ html.substring(end + 1));
+ }
+ }
+ }
+
+ return html.replace(pr_ltEnt, '<')
+ .replace(pr_gtEnt, '>')
+ .replace(pr_aposEnt, "'")
+ .replace(pr_quotEnt, '"')
+ .replace(pr_nbspEnt, ' ')
+ .replace(pr_ampEnt, '&');
+ }
+
+ /** is the given node's innerHTML normally unescaped? */
+ function isRawContent(node) {
+ return 'XMP' === node.tagName;
+ }
+
+ var newlineRe = /[\r\n]/g;
+ /**
+ * Are newlines and adjacent spaces significant in the given node's innerHTML?
+ */
+ function isPreformatted(node, content) {
+ // PRE means preformatted, and is a very common case, so don't create
+ // unnecessary computed style objects.
+ if ('PRE' === node.tagName) { return true; }
+ if (!newlineRe.test(content)) { return true; } // Don't care
+ var whitespace = '';
+ // For disconnected nodes, IE has no currentStyle.
+ if (node.currentStyle) {
+ whitespace = node.currentStyle.whiteSpace;
+ } else if (window.getComputedStyle) {
+ // Firefox makes a best guess if node is disconnected whereas Safari
+ // returns the empty string.
+ whitespace = window.getComputedStyle(node, null).whiteSpace;
+ }
+ return !whitespace || whitespace === 'pre';
+ }
+
+ function normalizedHtml(node, out, opt_sortAttrs) {
+ switch (node.nodeType) {
+ case 1: // an element
+ var name = node.tagName.toLowerCase();
+
+ out.push('<', name);
+ var attrs = node.attributes;
+ var n = attrs.length;
+ if (n) {
+ if (opt_sortAttrs) {
+ var sortedAttrs = [];
+ for (var i = n; --i >= 0;) { sortedAttrs[i] = attrs[i]; }
+ sortedAttrs.sort(function (a, b) {
+ return (a.name < b.name) ? -1 : a.name === b.name ? 0 : 1;
+ });
+ attrs = sortedAttrs;
+ }
+ for (var i = 0; i < n; ++i) {
+ var attr = attrs[i];
+ if (!attr.specified) { continue; }
+ out.push(' ', attr.name.toLowerCase(),
+ '="', attribToHtml(attr.value), '"');
+ }
+ }
+ out.push('>');
+ for (var child = node.firstChild; child; child = child.nextSibling) {
+ normalizedHtml(child, out, opt_sortAttrs);
+ }
+ if (node.firstChild || !/^(?:br|link|img)$/.test(name)) {
+ out.push('<\/', name, '>');
+ }
+ break;
+ case 3: case 4: // text
+ out.push(textToHtml(node.nodeValue));
+ break;
+ }
+ }
+
+ /**
+ * Given a group of {@link RegExp}s, returns a {@code RegExp} that globally
+ * matches the union o the sets o strings matched d by the input RegExp.
+ * Since it matches globally, if the input strings have a start-of-input
+ * anchor (/^.../), it is ignored for the purposes of unioning.
+ * @param {Array.} regexs non multiline, non-global regexs.
+ * @return {RegExp} a global regex.
+ */
+ function combinePrefixPatterns(regexs) {
+ var capturedGroupIndex = 0;
+
+ var needToFoldCase = false;
+ var ignoreCase = false;
+ for (var i = 0, n = regexs.length; i < n; ++i) {
+ var regex = regexs[i];
+ if (regex.ignoreCase) {
+ ignoreCase = true;
+ } else if (/[a-z]/i.test(regex.source.replace(
+ /\\u[0-9a-f]{4}|\\x[0-9a-f]{2}|\\[^ux]/gi, ''))) {
+ needToFoldCase = true;
+ ignoreCase = false;
+ break;
+ }
+ }
+
+ function decodeEscape(charsetPart) {
+ if (charsetPart.charAt(0) !== '\\') { return charsetPart.charCodeAt(0); }
+ switch (charsetPart.charAt(1)) {
+ case 'b': return 8;
+ case 't': return 9;
+ case 'n': return 0xa;
+ case 'v': return 0xb;
+ case 'f': return 0xc;
+ case 'r': return 0xd;
+ case 'u': case 'x':
+ return parseInt(charsetPart.substring(2), 16)
+ || charsetPart.charCodeAt(1);
+ case '0': case '1': case '2': case '3': case '4':
+ case '5': case '6': case '7':
+ return parseInt(charsetPart.substring(1), 8);
+ default: return charsetPart.charCodeAt(1);
+ }
+ }
+
+ function encodeEscape(charCode) {
+ if (charCode < 0x20) {
+ return (charCode < 0x10 ? '\\x0' : '\\x') + charCode.toString(16);
+ }
+ var ch = String.fromCharCode(charCode);
+ if (ch === '\\' || ch === '-' || ch === '[' || ch === ']') {
+ ch = '\\' + ch;
+ }
+ return ch;
+ }
+
+ function caseFoldCharset(charSet) {
+ var charsetParts = charSet.substring(1, charSet.length - 1).match(
+ new RegExp(
+ '\\\\u[0-9A-Fa-f]{4}'
+ + '|\\\\x[0-9A-Fa-f]{2}'
+ + '|\\\\[0-3][0-7]{0,2}'
+ + '|\\\\[0-7]{1,2}'
+ + '|\\\\[\\s\\S]'
+ + '|-'
+ + '|[^-\\\\]',
+ 'g'));
+ var groups = [];
+ var ranges = [];
+ var inverse = charsetParts[0] === '^';
+ for (var i = inverse ? 1 : 0, n = charsetParts.length; i < n; ++i) {
+ var p = charsetParts[i];
+ switch (p) {
+ case '\\B': case '\\b':
+ case '\\D': case '\\d':
+ case '\\S': case '\\s':
+ case '\\W': case '\\w':
+ groups.push(p);
+ continue;
+ }
+ var start = decodeEscape(p);
+ var end;
+ if (i + 2 < n && '-' === charsetParts[i + 1]) {
+ end = decodeEscape(charsetParts[i + 2]);
+ i += 2;
+ } else {
+ end = start;
+ }
+ ranges.push([start, end]);
+ // If the range might intersect letters, then expand it.
+ if (!(end < 65 || start > 122)) {
+ if (!(end < 65 || start > 90)) {
+ ranges.push([Math.max(65, start) | 32, Math.min(end, 90) | 32]);
+ }
+ if (!(end < 97 || start > 122)) {
+ ranges.push([Math.max(97, start) & ~32, Math.min(end, 122) & ~32]);
+ }
+ }
+ }
+
+ // [[1, 10], [3, 4], [8, 12], [14, 14], [16, 16], [17, 17]]
+ // -> [[1, 12], [14, 14], [16, 17]]
+ ranges.sort(function (a, b) { return (a[0] - b[0]) || (b[1] - a[1]); });
+ var consolidatedRanges = [];
+ var lastRange = [NaN, NaN];
+ for (var i = 0; i < ranges.length; ++i) {
+ var range = ranges[i];
+ if (range[0] <= lastRange[1] + 1) {
+ lastRange[1] = Math.max(lastRange[1], range[1]);
+ } else {
+ consolidatedRanges.push(lastRange = range);
+ }
+ }
+
+ var out = ['['];
+ if (inverse) { out.push('^'); }
+ out.push.apply(out, groups);
+ for (var i = 0; i < consolidatedRanges.length; ++i) {
+ var range = consolidatedRanges[i];
+ out.push(encodeEscape(range[0]));
+ if (range[1] > range[0]) {
+ if (range[1] + 1 > range[0]) { out.push('-'); }
+ out.push(encodeEscape(range[1]));
+ }
+ }
+ out.push(']');
+ return out.join('');
+ }
+
+ function allowAnywhereFoldCaseAndRenumberGroups(regex) {
+ // Split into character sets, escape sequences, punctuation strings
+ // like ('(', '(?:', ')', '^'), and runs of characters that do not
+ // include any of the above.
+ var parts = regex.source.match(
+ new RegExp(
+ '(?:'
+ + '\\[(?:[^\\x5C\\x5D]|\\\\[\\s\\S])*\\]' // a character set
+ + '|\\\\u[A-Fa-f0-9]{4}' // a unicode escape
+ + '|\\\\x[A-Fa-f0-9]{2}' // a hex escape
+ + '|\\\\[0-9]+' // a back-reference or octal escape
+ + '|\\\\[^ux0-9]' // other escape sequence
+ + '|\\(\\?[:!=]' // start of a non-capturing group
+ + '|[\\(\\)\\^]' // start/emd of a group, or line start
+ + '|[^\\x5B\\x5C\\(\\)\\^]+' // run of other characters
+ + ')',
+ 'g'));
+ var n = parts.length;
+
+ // Maps captured group numbers to the number they will occupy in
+ // the output or to -1 if that has not been determined, or to
+ // undefined if they need not be capturing in the output.
+ var capturedGroups = [];
+
+ // Walk over and identify back references to build the capturedGroups
+ // mapping.
+ for (var i = 0, groupIndex = 0; i < n; ++i) {
+ var p = parts[i];
+ if (p === '(') {
+ // groups are 1-indexed, so max group index is count of '('
+ ++groupIndex;
+ } else if ('\\' === p.charAt(0)) {
+ var decimalValue = +p.substring(1);
+ if (decimalValue && decimalValue <= groupIndex) {
+ capturedGroups[decimalValue] = -1;
+ }
+ }
+ }
+
+ // Renumber groups and reduce capturing groups to non-capturing groups
+ // where possible.
+ for (var i = 1; i < capturedGroups.length; ++i) {
+ if (-1 === capturedGroups[i]) {
+ capturedGroups[i] = ++capturedGroupIndex;
+ }
+ }
+ for (var i = 0, groupIndex = 0; i < n; ++i) {
+ var p = parts[i];
+ if (p === '(') {
+ ++groupIndex;
+ if (capturedGroups[groupIndex] === undefined) {
+ parts[i] = '(?:';
+ }
+ } else if ('\\' === p.charAt(0)) {
+ var decimalValue = +p.substring(1);
+ if (decimalValue && decimalValue <= groupIndex) {
+ parts[i] = '\\' + capturedGroups[groupIndex];
+ }
+ }
+ }
+
+ // Remove any prefix anchors so that the output will match anywhere.
+ // ^^ really does mean an anchored match though.
+ for (var i = 0, groupIndex = 0; i < n; ++i) {
+ if ('^' === parts[i] && '^' !== parts[i + 1]) { parts[i] = ''; }
+ }
+
+ // Expand letters to groupts to handle mixing of case-sensitive and
+ // case-insensitive patterns if necessary.
+ if (regex.ignoreCase && needToFoldCase) {
+ for (var i = 0; i < n; ++i) {
+ var p = parts[i];
+ var ch0 = p.charAt(0);
+ if (p.length >= 2 && ch0 === '[') {
+ parts[i] = caseFoldCharset(p);
+ } else if (ch0 !== '\\') {
+ // TODO: handle letters in numeric escapes.
+ parts[i] = p.replace(
+ /[a-zA-Z]/g,
+ function (ch) {
+ var cc = ch.charCodeAt(0);
+ return '[' + String.fromCharCode(cc & ~32, cc | 32) + ']';
+ });
+ }
+ }
+ }
+
+ return parts.join('');
+ }
+
+ var rewritten = [];
+ for (var i = 0, n = regexs.length; i < n; ++i) {
+ var regex = regexs[i];
+ if (regex.global || regex.multiline) { throw new Error('' + regex); }
+ rewritten.push(
+ '(?:' + allowAnywhereFoldCaseAndRenumberGroups(regex) + ')');
+ }
+
+ return new RegExp(rewritten.join('|'), ignoreCase ? 'gi' : 'g');
+ }
+
+ var PR_innerHtmlWorks = null;
+ function getInnerHtml(node) {
+ // inner html is hopelessly broken in Safari 2.0.4 when the content is
+ // an html description of well formed XML and the containing tag is a PRE
+ // tag, so we detect that case and emulate innerHTML.
+ if (null === PR_innerHtmlWorks) {
+ var testNode = document.createElement('PRE');
+ testNode.appendChild(
+ document.createTextNode('\n '));
+ PR_innerHtmlWorks = !/)[\r\n]+/g, '$1')
+ .replace(/(?:[\r\n]+[ \t]*)+/g, ' ');
+ }
+ return content;
+ }
+
+ var out = [];
+ for (var child = node.firstChild; child; child = child.nextSibling) {
+ normalizedHtml(child, out);
+ }
+ return out.join('');
+ }
+
+ /** returns a function that expand tabs to spaces. This function can be fed
+ * successive chunks of text, and will maintain its own internal state to
+ * keep track of how tabs are expanded.
+ * @return {function (string) : string} a function that takes
+ * plain text and return the text with tabs expanded.
+ * @private
+ */
+ function makeTabExpander(tabWidth) {
+ var SPACES = ' ';
+ var charInLine = 0;
+
+ return function (plainText) {
+ // walk over each character looking for tabs and newlines.
+ // On tabs, expand them. On newlines, reset charInLine.
+ // Otherwise increment charInLine
+ var out = null;
+ var pos = 0;
+ for (var i = 0, n = plainText.length; i < n; ++i) {
+ var ch = plainText.charAt(i);
+
+ switch (ch) {
+ case '\t':
+ if (!out) { out = []; }
+ out.push(plainText.substring(pos, i));
+ // calculate how much space we need in front of this part
+ // nSpaces is the amount of padding -- the number of spaces needed
+ // to move us to the next column, where columns occur at factors of
+ // tabWidth.
+ var nSpaces = tabWidth - (charInLine % tabWidth);
+ charInLine += nSpaces;
+ for (; nSpaces >= 0; nSpaces -= SPACES.length) {
+ out.push(SPACES.substring(0, nSpaces));
+ }
+ pos = i + 1;
+ break;
+ case '\n':
+ charInLine = 0;
+ break;
+ default:
+ ++charInLine;
+ }
+ }
+ if (!out) { return plainText; }
+ out.push(plainText.substring(pos));
+ return out.join('');
+ };
+ }
+
+ var pr_chunkPattern = new RegExp(
+ '[^<]+' // A run of characters other than '<'
+ + '|<\!--[\\s\\S]*?--\>' // an HTML comment
+ + '|' // a CDATA section
+ // a probable tag that should not be highlighted
+ + '|<\/?[a-zA-Z](?:[^>\"\']|\'[^\']*\'|\"[^\"]*\")*>'
+ + '|<', // A '<' that does not begin a larger chunk
+ 'g');
+ var pr_commentPrefix = /^<\!--/;
+ var pr_cdataPrefix = /^) into their textual equivalent.
+ *
+ * @param {string} s html where whitespace is considered significant.
+ * @return {Object} source code and extracted tags.
+ * @private
+ */
+ function extractTags(s) {
+ // since the pattern has the 'g' modifier and defines no capturing groups,
+ // this will return a list of all chunks which we then classify and wrap as
+ // PR_Tokens
+ var matches = s.match(pr_chunkPattern);
+ var sourceBuf = [];
+ var sourceBufLen = 0;
+ var extractedTags = [];
+ if (matches) {
+ for (var i = 0, n = matches.length; i < n; ++i) {
+ var match = matches[i];
+ if (match.length > 1 && match.charAt(0) === '<') {
+ if (pr_commentPrefix.test(match)) { continue; }
+ if (pr_cdataPrefix.test(match)) {
+ // strip CDATA prefix and suffix. Don't unescape since it's CDATA
+ sourceBuf.push(match.substring(9, match.length - 3));
+ sourceBufLen += match.length - 12;
+ } else if (pr_brPrefix.test(match)) {
+ //
tags are lexically significant so convert them to text.
+ // This is undone later.
+ sourceBuf.push('\n');
+ ++sourceBufLen;
+ } else {
+ if (match.indexOf(PR_NOCODE) >= 0 && isNoCodeTag(match)) {
+ // A will start a section that should be
+ // ignored. Continue walking the list until we see a matching end
+ // tag.
+ var name = match.match(pr_tagNameRe)[2];
+ var depth = 1;
+ var j;
+ end_tag_loop:
+ for (j = i + 1; j < n; ++j) {
+ var name2 = matches[j].match(pr_tagNameRe);
+ if (name2 && name2[2] === name) {
+ if (name2[1] === '/') {
+ if (--depth === 0) { break end_tag_loop; }
+ } else {
+ ++depth;
+ }
+ }
+ }
+ if (j < n) {
+ extractedTags.push(
+ sourceBufLen, matches.slice(i, j + 1).join(''));
+ i = j;
+ } else { // Ignore unclosed sections.
+ extractedTags.push(sourceBufLen, match);
+ }
+ } else {
+ extractedTags.push(sourceBufLen, match);
+ }
+ }
+ } else {
+ var literalText = htmlToText(match);
+ sourceBuf.push(literalText);
+ sourceBufLen += literalText.length;
+ }
+ }
+ }
+ return { source: sourceBuf.join(''), tags: extractedTags };
+ }
+
+ /** True if the given tag contains a class attribute with the nocode class. */
+ function isNoCodeTag(tag) {
+ return !!tag
+ // First canonicalize the representation of attributes
+ .replace(/\s(\w+)\s*=\s*(?:\"([^\"]*)\"|'([^\']*)'|(\S+))/g,
+ ' $1="$2$3$4"')
+ // Then look for the attribute we want.
+ .match(/[cC][lL][aA][sS][sS]=\"[^\"]*\bnocode\b/);
+ }
+
+ /**
+ * Apply the given language handler to sourceCode and add the resulting
+ * decorations to out.
+ * @param {number} basePos the index of sourceCode within the chunk of source
+ * whose decorations are already present on out.
+ */
+ function appendDecorations(basePos, sourceCode, langHandler, out) {
+ if (!sourceCode) { return; }
+ var job = {
+ source: sourceCode,
+ basePos: basePos
+ };
+ langHandler(job);
+ out.push.apply(out, job.decorations);
+ }
+
+ /** Given triples of [style, pattern, context] returns a lexing function,
+ * The lexing function interprets the patterns to find token boundaries and
+ * returns a decoration list of the form
+ * [index_0, style_0, index_1, style_1, ..., index_n, style_n]
+ * where index_n is an index into the sourceCode, and style_n is a style
+ * constant like PR_PLAIN. index_n-1 <= index_n, and style_n-1 applies to
+ * all characters in sourceCode[index_n-1:index_n].
+ *
+ * The stylePatterns is a list whose elements have the form
+ * [style : string, pattern : RegExp, DEPRECATED, shortcut : string].
+ *
+ * Style is a style constant like PR_PLAIN, or can be a string of the
+ * form 'lang-FOO', where FOO is a language extension describing the
+ * language of the portion of the token in $1 after pattern executes.
+ * E.g., if style is 'lang-lisp', and group 1 contains the text
+ * '(hello (world))', then that portion of the token will be passed to the
+ * registered lisp handler for formatting.
+ * The text before and after group 1 will be restyled using this decorator
+ * so decorators should take care that this doesn't result in infinite
+ * recursion. For example, the HTML lexer rule for SCRIPT elements looks
+ * something like ['lang-js', /<[s]cript>(.+?)<\/script>/]. This may match
+ * 'OcrApplication
Element Missed Instructions Cov. Missed Branches Cov. Missed Cxty Missed Lines Missed Methods Total 5 of 8 37 % 0 of 0 n/a 1 2 2 3 1 2 main(String[]) 0 % n/a 1 1 2 2 1 1 OcrApplication() 100 % n/a 0 1 0 1 0 1