Structured Fax File Format Description

Contents:

Introduction

This article describe the format of a Structured Fax File.

Data types

Document header

Page header

Page data

Page data is coded line by line: data describes each pixel row. Lines are coded as records of variable length; each line is coded according to the element coding in the page header. At present, only modified Huffman coding is supported.

MH coding is bitoriented: the pixel bits are stored in the bits of code words, least significant first. No EOL code words or fill bits are included. If the data includes EOL code words, COMMON-ISDN-API ignores these.

• 1...216: a pixel row with 1...216 MH-coded bytes follows immediately
• 0: escape code for a pixel row with more than 216 MH-coded bytes. In this case, the following word in the range 218...32767 defines the number of MH-coded bytes which follow.
• 217...253: white space, skip 1...37 empty lines
• 254: start of page header (see above)
• 255: if followed by a byte with value 0, illegal line coding. Applications may choose whether to interpret this line as empty or as a copy of the previous line. If this byte is followed by a byte with a value 1...255, then 1...255 bytes of additional user information follow (reserved for future extensions).

Modified Huffman Coding

The bit stream start with bit 0 of the first data byte. The data coding the number of successive pixel with a specific color. The coding start with white color. Black and white are alternate, up to all pixels are decoded or the bit stream is ended. There are Make-up codes and Terminating codes. First is the Make-up code or not. Annotation: A application for decoding should handle more then one Make-up code in sequence. It exist exactly one Terminating code. After the Terminating code the color of pixel alter. The codes are differently for black and white. For line length more then 1728 pixel there are additional Make-up codes. This codes are equal for black and white.

List of references