Systems and methods for generating tables from print-ready digital source documents

A method for generating tables from print-ready digital source documents, comprising:

• receiving, by a processor of a computing device, a print-ready digital source document, the digital source document comprising at least one rendered page;
• identifying, by the processor, one or more text fragments in the at least one rendered page, each of the text fragments comprising text, spatial coordinates indicating the positioning of the text fragment on the rendered page, and an index assigned based on the spatial coordinates on the rendered page;
• generating, by the processor, a wrapping region collection comprising one or more wrapping regions,
• wherein each of the wrapping regions comprises one or more fragment runs, and
• wherein each of the one or more fragment runs comprises a subset of the one or more text fragments that are adjacent to one another and within a predetermined horizontal separation threshold and a vertical separation threshold;
• calculating, by the processor, for each of the one or more wrapping regions of the wrapping region collection, a tabular score, a narrative score, and a label score, wherein each of the tabular score, narrative score, and label score is a measure of qualification of a wrapping region as a tabular block type, as a narrative block type, or as a label block type, respectively;
• assigning, by the processor, a block type to each of the one or more wrapping regions based on the corresponding calculated tabular score, narrative score and label score;
• generating a wrapping region group set comprising one or more wrapping region groups,
• wherein each of the one or more wrapping region groups comprises a subset of the one or more wrapping regions that are spatially related to one another;
• generating, by the processor, a block set comprising one or more blocks,
• wherein each of the one or more blocks comprises a subset of the one or more wrapping region groups that are spatially related to one another;
• and
• generating, by the processor, one or more tables, each of the one or more tables comprising the text fragments corresponding to one of the one or more blocks,
• wherein each of the one or more tables comprises the corresponding text fragments, each organized into corresponding fields of the one or more tables;
• wherein the tabular score, the narrative score and the label score of each of the one or more wrapping regions are calculated based on one or more attributes selected from the group consisting of (i) a normalization ratio, (ii) a density ratio, (iii) an alignment ratio, (iv) a capital or non-alphabetic ratio, (v) a text fragment quantity, and (vi) a bold count;
• wherein (i) the normalization ratio is the number of normalized text fragments divided by the total number of text fragments within a wrapping region, (ii) the density ratio is the percentage of a bounding box area of a wrapping region occupied by text fragment bounding boxes, (iii) the alignment ratio is the number of normalized text fragments that fit within at least one alignment group divided by the total number of normalized text fragments within a wrapping region, (iv) the capital or non-alphabetic ratio is the number of normalized text fragments that start with either a capital letter or a non-alphabetic character divided by the number of normalized text fragments, (v) the text fragment quantity is a number of text fragments among the subset of the one or more text fragments corresponding to each of the one or more wrapping regions, and (vi) the bold count is the number of text fragments with bolded text.