Something is wrong with the text width calculation

I am using pdf2json@4.0.3.

The x and y coordinates seems ok, but the w is a little large compared to the width of the page.

{
  "Width": 37.208,
  "Height": 52.62,
  "HLines": [],
  "VLines": [],
  "Fills": [],
  "Texts": [
    {
      "x": 11.181,
      "y": 6.631,
      "w": 228.612,
      "clr": 0,
      "sw": 0.36196875,
      "A": "left",
      "R": [
        {
          "T": "Intellectual Property Center, 28 Upper McKinley Rd.",
          "S": -1,
          "TS": [0,12.96,0,0]
        }
      ]
    },
    {
      "x": 8.369,
      "y": 7.343999999999999,
      "w": 318.272,
      "clr": 0,
      "sw": 0.36196875,
      "A": "left",
      "R": [
        {
          "T": "McKinley Hill Town Center, Fort Bonifacio, Taguig City 1634, Philippines",
          "S": -1,
          "TS": [0,12.96,0,0]
        }
      ]
    }
  ],
  "Fields": [],
  "Boxsets": []
}

Start by reproducing the report with pdf2json@4.0.3 using the supplied page dimensions, text strings, and reported x, y, and w values. Compare the calculated text widths with the page Width and the rendered result; done means the w values accurately represent the text width without changing the correct coordinates.