private static void printHelp()
{
//helpFormatter.PrintHelp("java -jar [JARNAME] [URL]", helpHeader, options, helpFooter, true);
Console.WriteLine("SKS Web crawler/info extractor v0.1");
Console.Write("Required tags for their category are in "); CU.WCol("red", CU.r); Console.Write(" and recommended tags are in "); CU.WCol("yellow", CU.y);
Console.Write(CU.nl + CU.nl + "Enter the URL followed by tags. "); CU.WCol("Even if you aren't doing a scan, you need a URL", CU.r);
Console.WriteLine();
Console.Write(CU.nl + "Help options:");
Console.Write(CU.nl + "-h, --help display this help dialog");
Console.Write(CU.nl + "-ph,--pattern help help with regex patterns for search limiting");
Console.WriteLine();
Console.Write(CU.nl + "Crawl Options:");
Console.Write(CU.nl + "-bc,--backcrawl deep copy, enables discovery of hidden pages (slow)");
CU.WCol(CU.nl + "-c,--courtesy <int>", CU.y); Console.Write(" delay between page loads, in milliseconds");
CU.WCol(CU.nl + "-d,--depth <int>", CU.y); Console.Write(" depth of the search (default 10)");
CU.WCol(CU.nl + "-p,--pattern <regex>", CU.y); Console.Write(" regex pattern for restricting pages");
CU.WCol(CU.nl + "-t,--threads <int>", CU.y); Console.Write(" number of allowed threads. More threads = more aggressive (must be 2+)");
Console.Write(CU.nl + "-ul,--unlock unlocks crawler from target domain");
Console.Write(CU.nl + "-i,--iterative <int> scans urls iteratively in the form of url/1,2.. starting at <param>");
Console.Write(CU.nl + "-id,--iterative depth <int> Depth to scan to at each step of the iteration");
Console.Write(CU.nl + "-dc,--dont crawl do not execute crawl. (for the purpose of using other utilities only)");
Console.WriteLine();
Console.Write(CU.nl + "File Options:");
Console.Write(CU.nl + "-di,--images <regex> download images while crawling (takes regex for filtering)");
Console.Write(CU.nl + "-dt,--text <tag, regex> download text bodies from <tag> for analyzation, if the page matches <regex>");
Console.Write(CU.nl + "-g,--gallery only download files to one folder");
Console.Write(CU.nl + "-ddh,--HTML don't download HTML while crawling");
Console.Write(CU.nl + "-il,--include link include links to the parent page in text files");
//CU.WCol(CU.nl + "-l,--load <filename>", CU.y); Console.Write(" load data from previous scan, named <filename> ");
CU.WCol(CU.nl + "-o,--output <dir>", CU.y); Console.Write(" output location (defaults to exe location)");
Console.Write(CU.nl + "-O,--overwrite overwrite files when scan starts");
Console.WriteLine();
Console.Write(CU.nl + "Database Options:");
CU.WCol(CU.nl + "-dbip,--database ip <ip>", CU.r); Console.Write(" the IP address of the database to dump to");
Console.Write(CU.nl + "-dbc,--database check Check the database to prevent duplicate entries "); CU.WCol("(Slow and expensive)", CU.y);
Console.Write(CU.nl + "-dbi,--database images Save image locations to database, with tags defined by -dt");
Console.Write(CU.nl + "-dbl,--database links Save visited links into the DB, with tags defined by -dt");
Console.WriteLine();
Console.Write(CU.nl + "Data processing Options:");
Console.Write(CU.nl + "-m,--markov <int> generate a markov chain of <int> prefix Length and saves it.");
Console.Write(CU.nl + "-mp,--print markov <int> prints out <int> sentences from the chain (Must use -g)");
Console.WriteLine();
Console.Write(CU.nl + "Output Options:");
Console.Write(CU.nl + "-v,--verbose verbose mode");
Console.Write(CU.nl + "-vi,--visited print visited pages a after completion (n.i.);");
Console.WriteLine();
Console.Write(CU.nl + "Example usages:");
Console.Write(CU.nl + "Basic scan: java -jar [JARNAME] http://examplesite.com -di -d 5");
Console.Write(CU.nl + "Site Image Gallery: [URL] -di -ddh -g");
Console.Write(CU.nl + "Fullsize gallery of 4chan thread: [URL] -di ^((?!s.).)*$ -ddh -g -p .*/((?!#[spq]).)*");
Console.Write(CU.nl + "Booru tags on posts with urls: [URL] -g -il -ddh -dt title (.*)(/post/show/)(.*) -O -c 1000 -d 3");
Console.Write(CU.nl + "Iterative booru tag crawl: [BASEURL] -g -il -ddh -dt title -O -c 1000 -d 1000 -i <startpage>");
Console.Write(CU.nl + "Markov chain from 4chan board: [URL] -t 10 -d 15 -dt .post .* -m 2 -g -ddh -O -mp 40");
Console.Write(CU.nl + "Insert images into database with tags: [BOORUURL] -g -t 10 -di .*[/](_images/).* -ddh -d 10 -O -p .*[/]post/.* -ul -dt title -dbi)" + CU.nl);
}
public object LoadObject(string name, Type type)
{
object objOut = new object();
string fileName = baseFolder + name + ".xml";
DataContractSerializer serializer = new DataContractSerializer(type);
using (XmlReader reader = XmlReader.Create(fileName)) {
objOut = serializer.ReadObject(reader);
}
CU.WCol(fileName + " loaded successfully." + CU.nl, CU.g);
return(objOut);
}
public void SaveObject(string name, object objIn, bool overWrite)
{
int fileVersion = 1;
string fileName = baseFolder + name + ".xml";
//Overwrite prevention- ex: file, file_v1, file_v2, file_v3......
while (File.Exists(baseFolder + name) && !overWrite)
{
fileVersion++;
fileName = baseFolder + name + "_v " + fileVersion;
}
DataContractSerializer serializer = new DataContractSerializer(objIn.GetType());
using (XmlWriter writer = XmlWriter.Create(fileName)) {
serializer.WriteObject(writer, objIn);
}
CU.WCol(fileName + " saved successfully." + CU.nl, CU.g);
}
public int Run()
{
Console.WriteLine(threadName + " worker started!");
try {
//Executes iterative scan
if (data.iterative)
{
for (int i = 0; i < data.maxDepth; i++)
{
iteratorLocation++;
//Mod this into a more-than-2-thread operation later
if (sidedness)
{
while (iteratorLocation % 2 == 0)
{
iteratorLocation++;
}
}
else
{
while (iteratorLocation % 2 != 0)
{
iteratorLocation++;
}
}
pagesCrawled += Crawl(startURL + iteratorLocation, 0, sidedness);
}
}
else
{
pagesCrawled = Crawl(workingURL, 0, sidedness);
}
} catch (Exception e) {
CU.WCol(CU.nl + "something went terribly wrong, crawler died " + e, CU.r, CU.y);
}
isFinished = true;
Console.WriteLine(threadName + " finished");
return(pagesCrawled);
}
//Saves all elements in argument
public void DownloadElements(Element[] elements, FileInfo[] fileInfo = null)
{
int totalDownloaded = 0;
int totalBounced = 0;
//Returns if there are no files to be downloaded
if (fileInfo != null && fileInfo.Length == 0)
{
return;
}
for (int i = 0; i < elements.Length; i++)
{
Element content = elements[i];
string absURL;
string tag = content.Tag.ToString();
switch (tag)
{
case "img":
absURL = content.AbsUrl("src");
break;
case "a":
absURL = content.AbsUrl("href");
break;
default:
absURL = content.AbsUrl("src");
break;
}
data.mediaTrie.InsertURL(absURL);
FileInfo file;
//Doesn't recaculate file info if it doesn't have to
if (fileInfo == null)
{
int nameIndex = absURL.LastIndexOf('/');
//Name of the element
string elementName = Regex.Replace(absURL.Substring(nameIndex + 1), "[^A-Za-z.]", "");
//File location of the element
string elementLocation = absURL.Substring(0, nameIndex);
if (elementName.Length > 20)
{
elementName = elementName.Substring(elementName.Length - 20);
}
//Inserts hash into filename to avoid duplicates
string hashCode = Convert.ToString(content.GetHashCode());
elementName = elementName.Insert(0, hashCode);
if (!data.gallery)
{
file = new FileInfo(webStringUtils.UrlToDir(elementLocation) + elementName);
}
else
{
file = new FileInfo(data.outputFolder + elementName);
}
}
else
{
file = fileInfo[i];
}
//Defers downloading to the saver
Save(absURL, file);
//Sleeps to slow down image requests
Thread.Sleep(data.delay);
totalDownloaded++;
}
string report = "Downloaded " + totalDownloaded + " media files, denied " + totalBounced;
CU.WCol(CU.nl + report + CU.nl, CU.c);
}
//false = left, true = right
//This should probably be split into two or three classes that inherit
public int Crawl(string url, int currentDepth, bool sidedness)
{
//helpers at their base level do now download content
bool isHelper = false;
//Iterative searchers spawn recursors to assist
bool isIterativeRecursor = false;
if (data.iterative && !url.Substring(0, url.LastIndexOf('/') + 1).Equals(data.startURL))
{
isIterativeRecursor = true;
}
int newPages = 0;
//Early return if url is not in the domain or if it has been previously checked
if (!url.Equals(startURL))
{
//Checks to see if the URL has already been searched or if it's not in the domain- if so, terminate.
if (!Regex.IsMatch(url, data.crawlPattern) || !Regex.IsMatch(url, data.crawlDomain) || data.urlTrie.contains(url))
{
return(newPages);
}
//Bounces via DB if flag is checked
if (data.dataBaseCheck)
{
Task <bool> isInDB = Task.Run(() => TagDBDriver.entryExists(url));
isInDB.Wait();
data.urlTrie.InsertURL(url);
if (isInDB.Result)
{
return(newPages);
}
}
newPages++;
}
else if (url.Equals(startURL))
{
if (data.urlTrie.contains(url))
{
isHelper = true;
//isIterativeRecursor = true;
}
}
//Iterative helpers still check the URLTrie
if (isIterativeRecursor && (data.urlTrie.contains(url) || !Regex.IsMatch(url, data.crawlPattern)))
{
return(newPages);
}
data.urlTrie.InsertURL(url);
//Courtesy delaying
Thread.Sleep(data.delay);
if (data.verbose)
{
Console.WriteLine("Starting crawl on " + url);
}
Document HTMLDoc = null;
int errorIterator = 0;
while (HTMLDoc == null)
{
try {
HTMLDoc = getHTML(url);
} catch (Exception exception) {
//Helpers do not increase the iterator
if (data.iterative && !isIterativeRecursor)
{
if (exception.ToString().Contains("404"))
{
iteratorLocation++;
Console.WriteLine("404. Now on page " + iteratorLocation + ". Increasing index....");
}
else if (exception.ToString().Contains("503"))
{
iteratorLocation++;
Console.WriteLine("503. Now on page " + iteratorLocation + ". Increasing index....");
}
else
{
throw exception;
}
}
if (exception.ToString().Contains("429"))
{
//Handling for rate limit exceptions
errorIterator++;
if (errorIterator < 1)
{
Console.WriteLine(exception);
Console.WriteLine("Rate limited. waiting...");
Thread.Sleep(15000 + data.delay);
Console.WriteLine("Retrying...");
}
else
{
Console.WriteLine("Continued rate limiting. Thread waiting for one minute and increasing courtesy delay.");
Thread.Sleep(60000);
data.delay += 10000;
errorIterator = 0;
}
}
else
{
CU.WCol(CU.nl + "Could not load page. " + url + " : " + exception.Message + CU.nl, CU.r);
return(newPages);
}
}
}
//Grab links
Elements links = HTMLDoc.Select("a[href]");
int numberOfLinks = links.Count();
//Grabs the page title
string titleString = HTMLDoc.Title;
if (titleString != null)
{
titleString = HTMLDoc.Title;
}
else
{
titleString = "Untitled";
}
if (!data.verbose)
{
if (!isHelper)
{
Console.WriteLine(threadName + " Crawling " + url.Truncate(40) + "(" + titleString.Truncate(40) + ")");
}
else
{
Console.WriteLine(threadName + " Finishing " + url.Truncate(40) + "(" + titleString.Truncate(40) + ")");
}
}
if (data.verbose)
{
Console.WriteLine("Page name: " + titleString);
}
//Writes content to file
try {
//Refuse files if crawler is helper at level
if (!isHelper)
{
//Prep information for DB entries
FileInfo[] files = null;
string[] tags = null;
//Download HTML
if (data.downloadHTML)
{
Thread downloadHTML = new Thread(() => downloadManager.downloadHTML(HTMLDoc));
downloadHTML.Start();
}
//Download text within specified tags (-dt [tag])
if (data.downloadText)
{
Elements text = HTMLDoc.Select(data.textTag);
Task <string[]> downloadText = new Task <string[]>(() => downloadManager.downloadText(text));
downloadText.Start();
tags = downloadText.Result;
}
//Download images and links to images
if (data.downloadImages)
{
//Checks for links to images
Elements imageElements = HTMLDoc.Select("img");
if (imageElements != null)
{
//Append links to images as well
foreach (Element element in links)
{
if (Regex.IsMatch(element.AbsUrl("href"), ".*(.jpg|.png|.gif|.webm)"))
{
imageElements.Add(element);
}
}
}
Task <FileInfo[]> downloadImages = new Task <FileInfo[]>(() => downloadManager.DownloadElementsReturnNames(imageElements));
downloadImages.Start();
files = downloadImages.Result;
}
//Saves image locations to Database
if (data.dataBaseImages)
{
foreach (FileInfo file in files)
{
new Thread(() => TagDBDriver.insertImageWithTags(file.FullName, tags)).Start();
}
}
//Saves links to Database
if (data.dataBaseLinks)
{
new Thread(() => TagDBDriver.insertImageWithTags(url, tags)).Start();
}
}
} catch (Exception e) {
Console.WriteLine("Could not write to file: " + e);
}
//Checks if the search needs to recurse
if (numberOfLinks <= 0)
{
if (data.verbose)
{
Console.WriteLine("No links on page. Going back up...");
}
return(newPages);
}
//if the crawl is iterative, do not recurse
try {
//Recurses the algorithm if not at max depth
if (currentDepth + 1 > data.maxDepth && !data.iterative)
{
return(newPages);
}
//Do shallow recursion while in iterative mode
else if (currentDepth + 1 >= data.iterativeDepth && isIterativeRecursor)
{
return(newPages);
}
if (numberOfLinks > data.linkAssist && !data.assistTrie.contains(url) && !data.iterative)
{
data.assistTrie.InsertURL(url);
this.workingURL = url;
this.workingDepth = currentDepth;
createHelper();
}
//Right-handed search
//can these be one method?
int sizeLimit = (int)Math.Round(numberOfLinks / 2f);
if (sidedness)
{
for (int i = numberOfLinks - 1; i > 0; i--)
{
//Only search half at the entry depth
if (currentDepth == startDepth && i < sizeLimit)
{
break;
}
string currentLinkRight = links[i].AbsUrl("href");
//Checks to make sure that the URL isn't a page in-reference and that it doesn't link to another part of the page. Also ensures link validity.
//Also ignore links to other pages positioned along the iterative crawl
if (string.IsNullOrEmpty(currentLinkRight) || currentLinkRight.Equals(url) ||
(data.iterative && currentLinkRight.Substring(0, currentLinkRight.LastIndexOf('/') + 1).Equals(data.startURL)) ||
(currentLinkRight.Contains('#') && currentLinkRight.Substring(0, currentLinkRight.LastIndexOf('#')).Equals(url)) ||
links[i].HasAttr("rel"))
{
{
i--;
continue;
}
}
//Ensures the link can be connect to- if not, iterate to the next link
try {
WebRequest.Create(currentLinkRight);
} catch (Exception) {
i--;
continue;
}
newPages += Crawl(currentLinkRight, currentDepth + 1, sidedness);
}
} //Left-handed search
else
{
for (int i = 0; i < numberOfLinks - 1; i++)
{
if (currentDepth == startDepth && i > sizeLimit)
{
break;
}
string currentLinkLeft = links[i].AbsUrl("href");
string test = currentLinkLeft.Substring(0, currentLinkLeft.LastIndexOf('/') + 1);
if (string.IsNullOrEmpty(currentLinkLeft) || currentLinkLeft.Equals(url) ||
(data.iterative && currentLinkLeft.Substring(0, currentLinkLeft.LastIndexOf('/') + 1).Equals(data.startURL)) ||
(currentLinkLeft.Contains('#') && currentLinkLeft.Substring(0, currentLinkLeft.LastIndexOf('#')).Equals(url)) ||
links[i].HasAttr("rel"))
{
i++;
continue;
}
try {
WebRequest.Create(currentLinkLeft);
} catch (Exception) {
i++;
continue;
}
newPages += Crawl(currentLinkLeft, currentDepth + 1, sidedness);
}
}
//Backcrawl to hit missed directoies at the level
if (data.backCrawl)
{
while (url.Substring(8).Contains("/"))
{
Console.WriteLine("Backcrawling unfound urls...");
Crawl(url = url.Substring(0, url.LastIndexOf('/') - 1), currentDepth - 1, sidedness);
}
}
} catch (Exception e) {
CU.WCol(CU.nl + "Dead page: " + e, CU.r, CU.y);
CU.WCol(CU.nl + e.StackTrace, CU.r, CU.y);
Console.WriteLine("Now checking depth " + currentDepth + ", link: " + url);
}
return(newPages);
}
