pdf-to-text-fragments

Extract all possible textual information from a PDF file. This is intended mainly for tabular data where positional as well as textual information is required. PDF uses two text string placement operators, Tj and TJ. Tj places equally spaced characters while TJ places variably spaced characters starting from an X, Y coordinate in arbitrary units. A text fragment consists of the X and Y coordinates of the text string along with the text string. A list of text fragments containing all the text on page is extracted and then sorted to put the text fragments into reading order. The list is then converted into strings; 'X value, Y value, text\n' which are concatenated and stored in a text file, 'Page n', where n is the page number. Data extraction becomes much easier because parsing can be based on both text value and text position. This is very useful for data sources which are available only in PDF form and are updated regularly. For example, insider trading data from SEDI.ca.