rwachtler / itm13termstatistics Goto Github PK

Requirement

"Pro Wort mit vorkommenden Seiten und Anzahl auf vorkommender Seite"

Request parameter

word - String (e.g. politik)

Expected result

[
    {
        "derstandard.at": 156
    }, {
        "orf.at": 356
    }
]

How to reproduce

On index page try to add a wrong URL to the crawl list (like 12345), this shows an error modal. After dismissing, the input popup is not responding to further user actions. After closing the popup, the "New" button is not responding anymore, so no "Add to crawlist" pop-up can be opened again until complete page refresh.

Change input type of crawldepth in "New Page" to unsigned number.

Current output

{
    "data" : {
         "amount":15321,
         "word":"die"
    }
}

Expected output

{
    "hallo" : [
         {
             date : "01.11.2015",
             amount: 5
         },
         {
             date : "03.11.2015",
             amount: 20
         },
         {
             date : "15.11.2015",
             amount: 100
         }
    ]
}

How do we handle the case of re-crawling an (on this day) already crawled page? A few ideas:

Using the old resultfile in case of

• re-crawling within the same minute / hour
  -- the timestamp is in the filename, a check to see if the minute / hour is the same would be enough

Re-crawling in case of

• no idea

Some input would be nice.

If more than one crawler (= more than one website is crawled at the same time) is working at the same time, the analyzing process can't start until every crawler-thread (even the ones that don't belong to the same website) is finished.

Requirement

During the process of adding a new domain, a domain check should be performed which verifies that the user input is a valid URL.

NOTE: Should return a notification to the user, that the given URL is invalid. (optional)

There needs to be a notification when the user-triggered crawling of an already existing page is done, A simple implementation for that would be a new field in the database "is_crawling" with values "1" or "0" for true or false.

While the value is true the page is crawled again, when the value switches to false the user gets a notification that his triggered crawling is done either via e-mail or just on-page notification.

Requirement

"Pro Seite alle vorkommenden Wörter und Anzahl mit Zeiteinschränkung"

Request parameter

domainID - Numeric value
startdate - String (e.g. 01.11.2015)
enddate - String (e.g. 30.11.2015)

Expected result - (max. 10 most popular words)

[
    {
        "hallo": {
            stats: [{
                date: "01.11.2015",
                amount: "3"
            }, {
                date: "03.11.2015",
                amount: "10"
            }]
        },
        "politik": {
            stats: [{
                date: "01.11.2015",
                amount: "3"
            }, {
                date: "03.11.2015",
                amount: "10"
            }]
        }
    }
];

Right now the "SiteOverview" page shows "No data available in table" after adding a page to the list. It would be nice if there is a little status message that the page is crawled right now and there should be results in X minutes.

Display the URL of the page entry you want to edit in the head next to "Edit Entry" or as deactivated input field to give the user a hint which page he is editing right now.

The description field is nice, could be not enough of a hint tho.

After finished the crawling and analyzing the txt files should be deleted.

ä, ö, ü, ß etc. aren't rendered correctly. UrlEncode should be implemented in the adding Word section in the Analyzer and UrlDecode in the REST-Points