PURPOSE: To reduce the data volume of font data and to shorten the substitution time.

CONSTITUTION: For example, font data for lateral writing is used as the mother, and only data, which is not included in font data for lateral writing, out of respective character pattern data of font data for vertical writing is extracted and is added to font data for lateral writing, and an identifier ID is added to each character pattern of this unified font data to generate a second table ST. Separately, a first table which prescribes relations between identifiers and character codes is generated for each of lateral writing (PTH) and vertical writing (PTV). When a character pattern for lateral writing is requested, the character code related to the request is used as the keyword to read out a corresponding identifier from the first table for lateral writing. This identifier is used as the keyword to read out a corresponding character pattern for lateral writing from the second correspondence table. For example, the first correspondence table for lateral writing is changed to generate the first correspondence table for vertical writing.