This counter is intended for use in GitHub README markdown files. It has the following features:

• It is accessed in a <img> tag in the markdown file.
• It will only count for specified IDs. IDs are a means to identify which README was viewed.
• It does not display the current count. This is intentional.
• It's small, less than 115kb in size.
  • Code + JSON = ~4k
  • Images = ~111k - The only reason it's this big is because of the "error" images, they're supposed to be obvious. The actual image that used in counting is less than 1k in size.

You will need:

• A web server, HTTP or HTTPS should be ok.
• The server needs PHP >5.6.
• You will need access to your server to copy files and create a folder.

These set up instructions will verify that it's working. However for live use some editing is required, this will be described later.

1) Copy the following files to a desired location on your server. Typically this will be in a public_html folder or in a sub-folder(recommended) of public_html.

• counters.json
• mdcount.php
• All PNG files, blank-10x10.png, err-10x10.png, id_err.png, id_oops.png, and test-246x246.png
• NOTE: It is not necessary to copy the mdimg folder to the server!

2) In the folder where you copied the files to create a new folder named logs.

Open example.html in your editor and:

1) Find the <img> tags and edit the src attributes.

• Replace all occurrences of [your-server]/[some-folder] with your server's domain and the sub-folder where you placed the files.

2) Then on your PC run the example.html file by double-clicking it or by dropping it onto an open browser window.

Test Run Expected Results:

a) At the top of the page you should see this:

b) A little further down the page you will see where there does not appear to be an image. It's there, it's just small (10px X 10px) and transparent.

c) The last two images demonstrate how errors are handled. The two types of errors that can occur are:

• An invalid ID
• The ID query is missing

The images are large on purpose. They're meant to be obvious when something is wrong.

d) In the folder where you placed the files you will have a new file named testtest_count.log.

e) In the sub-folder named logs you should have 2 new files, example_1_count.log and example_2_count.log.

Each of the *.log files should have a count of 1 in them.

The counter log files will be found in the folder ./logs. They are JSON formatted and contain the current count and the date and time of the visit.

Here is an example:

{
  "count": 2,
  "time": 1616961746,
  "dtime": [
    "20210328",
    "150226"
  ]
}

Use within a GitHub markdown file is easy. You cah add the counter to a README.md file or any other *.md file in any of your repositories.

The following line can go anywhere in the markdown file but at the bottom is recommended.

<img src="http://[your-server]/[some-folder]/mdcount.php?id=testtest">

The counters.json file contains IDs that insure that each use has its own counter log file. The counters.json file in this repository looks like this:

{
  "valid":[
    "TESTTEST",
    "example_1",
    "example_2"
  ]
}

Then:

<img src="http://[your-server]/[some-folder]/mdcount.php?id=example_1">

will create this file - logs/example_1_count.log, and each time the containing file is opened it will increment the count in that log file.

Let's say you have a README in your repository named My Repo. First add it to the counters.json file like this:

{
  "valid":[
    "TESTTEST",
    "example_1",
    "example_2",
    "My_Repo"
  ]
}

And to count use this:

<img src="http://[your-server]/[some-folder]/mdcount.php?id=My_Repo">

The resulting file will be logs/my_repo_count.log.

NOTE: Replace any spaces with an underscore or hyphen. The incoming IDs and those in the counters.json file can be mixed case because they are converted to lower case before they're checked. For typical use with Github and markdown files the IDs should match repository names.

The images.json file contains the file names of the PNG files and their path relative to the sub-folder where you placed the images and mdcounter.php.

{
    "path": "./",
    "testimg": "test-246x246.png",
    "countimg": "blank-10x10.png",
    "errimg": "id_err.png",
    "oopsimg": "id_oops.png"
}

The "path": "./" member may also be edited if necessary.

Other than a "normal view" here are somethings that will trigger the counter:

• WYSIWYG markdown editors. Each time you open the markdown file it will trigger a count.
• Refreshing GitHub repository main page while viewing it.

It is possible to retrieve the counter data for a single counter or for all of them. This can be accomplished with -

Return data for a specific counter -

GET http[s]://your-server/path-to-file/mdcountdata.php?id=sensornet

The data is returned in JSON -

[
    {
        "id": "sensornet",
        "data": {
            "count": 30,
            "time": 1616458038,
            "dtime": [
                "20210322",
                "190718"
            ]
        }
    }
]

Return just the count data for a specific counter -

GET http[s]://your-server/path-to-file/mdcountdata.php?id=sensornet&count

A single value is returned, no JSON formatting -

30

Return data for all counters -

GET http[s]://your-server/path-to-file/mdcountdata.php

The data is returned in JSON -

[
    {
        "id": "sensornet",
        "data": {
            "count": 30,
            "time": 1616458038,
            "dtime": [
                "20210322",
                "190718"
            ]
        }
    },
    {
        "id": "test",
        "data": {
            "count": 1,
            "time": 1616458038,
            "dtime": [
                "20210322",
                "190718"
            ]
        }
    }
]

Sorting Counter Data -

The counter data can be sorted by count before returning it to the caller.

The data is returned in ascending order -

GET http[s]://your-server/path-to-file/mdcountdata.php?csort=a

The data is returned in descending order -

GET http[s]://your-server/path-to-file/mdcountdata.php?csort=d

One of the following sorts can be chosen:

• csort - sort by count value
• tsort - sort by time of last hit
• isort - sort by counter ID

Limiting Number of Counters in Sort -

It is possible to limit the quantity of counters returned for any of the sorting queries.

GET http[s]://your-server/path-to-file/mdcountdata.php?csort=a|[&limit=[1-n]]

The value of limit can be from 1 to n. If n exceeds the number of counters then all counters will be returned for that sorting choice.

NOTE : limit only applies to sorting queries.

Errors -

If an error occurs it is also returned in JSON -

GET http[s]://your-server/path-to-file/mdcountdata.php?id=BLAH


[
    {
        "error": true,
        "msg": "BLAH not found in ./counters.json"
    }
]

The following files are used in report generation and viewing:

• mdreport.php - Retrieves the counter data and renders a Bootstrap 4.x table
  • mdhcreport.css - Additional required CSS for the table
  • embedreport.css - CSS required for an embedded report
  • reporthead.html - For non-embedded reports, this contains the page title and table heading
  • reportcaption.html - For non-embedded reports, this contains the table caption
  • mdreport-th.txt - Column heading text
  • stddefines.php - A collection of define() that make a number of PHP $_SERVER[] values available to the application. It contains components used for creating URLs to resources
• report.html - The minimum required HTML/CSS and JavaScript/jQuery to render and display the table

Retrieving the Report -

The code that retreives the report and handles clicks on the column headings is contained in report.html. The necessary CSS can be found in mdhcreport.css.

<body>
    <div id="repout">
    </div>
<script>
document.addEventListener("DOMContentLoaded", () => {
    /*
        Send a GET request and invoke a 
        callback function when completed.
    */
    function httpGet(url, callback, tickle = false) {
        var xmlhttp = new XMLHttpRequest();
        xmlhttp.onreadystatechange = function() {
            if(this.readyState == 4 && this.status == 200) {
                var resp = this.responseText;
                callback(resp);
            }
        };
        // bypass caching, useful when retrieving resources
        // that change frequently
        if((tickle === true) && (url.includes('?') === false)) {
            url = url + '?_=' + new Date().getTime();
        }
        xmlhttp.open('GET', url, true);
        xmlhttp.send();
    };

    //////////////////////////////////////////////////////////////////////////
    // sorting types
    var sorts = ['?csort=','?isort=','?tsort='];
    // sort directions
    var sdirs = ['a','d'];
    // URL to report generator
    var url = './mdreport.php';
    // optional limiting qty of counters returned
    var limit = '&limit=10';
    // 
    var datasrc = url+sorts[2]+sdirs[1];//+limit;

    // create a table with mdreport.php and place 
    // it within an specified element ID
    function createTable(elemid, ds) {
        httpGet(ds, (resp) => {
            $(elemid).html(resp);
            // wait for a column in this table to be clicked
            $(elemid+' .orderhover').click((col) => {
                var target = {
                    id: col.target.id,
                    ix: col.target.dataset.ix,
                    order: col.target.dataset.order
                };
                $(document).trigger('newtable', target);
            });
        },false);
    };

    // render the table of counters
    createTable('#repout', datasrc);

    // when a column heading is clicked 
    $(document).on('newtable', (e, target) => {
        $('.orderhover').off('click');
        // build the URL and query...
        var datasrc = url+sorts[target.ix - 1] + (target.order === 'a' ? 'd' : 'a');// + limit;
        // erase the previous table
        $('#repout').html('');
        // go get it!
        createTable('#repout', datasrc);
    });
});
</script>
</body>

Sample Report Screen Shot - (via report.html)

The "Hit Count", "Repository" and "Last Counted" headings can be clicked to select sorting criteria and direction (ascending vs descending). When a column heading is clicked the report caption will change to reflect the choice.

A demonstration of the counter report can be seen here.

NOTES :

• The links in the "Repository" column are built from the IDs found in counters.json. The ID text I used there is the name of the repository that the counter is intended for. The corresponding counter data files are named repo-name_counter.json.
• When the report is viewed the data shown is current.
• If the "Last Counted" column date and time seem to be incorrect then the time zone may need to be changed. Edit the timezone.json file to match your time zone.

It is possible to retreive counter data without using report.html. The purpose is to embed the table in another page and to be able to apply some alternative CSS.

To retrieve it -

GET http[s]://your-server/path-to-file/mdreport.php?tsort=d&limit=5&embed

The response is going to contain some CSS for the table, and just the table itself. You can use a browser and go to http[s]://your-server/path-to-file/mdreport.php?tsort=d&limit=5&embed. You will see something like this-

NOTE: The table is not sortable like the one rendered with report.html.

You could count just about anything. All you need is to do a GET of mdcount.php with a proper query and you got a counter!

This started out as an interesting afternoon project. I created it because I could not find exactly what I needed in the many hit counters out there.

And this project has evolved since then with the addition (and changes) of JSON formatted counter files, and sorted counter data retrieval. And the latest change was adding a 5th column to contain some GitHub statistics.