Font::TTF::Scripts::AP - Memory representation of a TTFBuilder Attachment Point database (APDB)
use Font::TTF::Scripts::AP; $ap = Font::TTF::Scripts::AP->read_font($ttf_file, $ap_file, %opts); $ap->make_classes();
Reference to the Microsoft cmap within the font.
font
Reference to a font structure. read_font will cause at least the post, cmap, loca, and name tables to be read in.
read_font
An array of references to glyph data structures, indexed by glyphID. Stucture elements are:
Array of Unicode scalar values (decimal integers), if any, that map from cmap to this glyph.
Actual glyph ID from font.
Actual Postscript name from font.
This element is set by "make_names" or "make_classes" and is the replacement name returned by "make_name".
Note: The uni, gnum and post values are based on the UID, GID, and PSName fields of the APDB. If there are descrepancies between the APDB and the font's internal tables, then for calcuating the above three values, priority is given first to UID field, then PSName field, and finally GID.
uni
gnum
post
UID
GID
PSName
Reference to glyph structure read from font.
Line number in APDB where glyph is defined.
A hash of references to attachment point structures for this glyph, keyed by attachment point type (aka name). Each AP structure contains
The name (type in TTFBuilder terminology) of the attachment point
type
X and Y coordinates for the attachment point
Line number in APDB where this point is defined.
Present if the glyph is a composite. Is a reference to an array of component structures. Each component structure includes:
comma separated list of bounding box coordinates, i.e., x1, y1, x2, y2
x1, y1, x2, y2
Unicode scalar value, if any, of the component. (decimal integer)
Line number in APDB where this component is defined.
Note: The following instance variables contain the actual text read from the APDB. If there are descrepancies between the APDB and the font, these values may differ from corresponding values given above. Therefore these values should not be used except for diagnostic purposes.
Unicode scalar value, if any, as specified in the APDB. (string of hex digits)
Postscript name, if any, as specified in the APDB
Glyph id, if any, as specified in the APDB
Created by "make_classes", this is a hash keyed by class name returning an array of GIDs for glyphs that are in the class. Classes are identified by extensions (part after a '.') on the post name of each glyph. For each such extension, two classes are defined. The first is the class of all glyphs that have that extension (class name is the extension). The second is the class of nominal glyphs corresponding to the glyphs with that extension (class name is the extension but with the prefix 'no_').
Created by "make_classes", this is a hash keyed by attachment point name (as modified by "make_point") returning an array of GIDs for glyphs that have the given attachment point.
If defined, this variable will be updated by "make_classes". It is a hash, keyed by attachment point name (as modified by "make_point") returning a bit vec bit array, indexed by GID, each bit set to 1 if the corresponding glyph has the given attachment point.
Optionally created by "make_classes" if ligatures are requested and they exist. The base forms class is no_code while the ligatures are held in code.
If -errorfh not set, this accumulates any warning or error messages encountered.
-errorfh
Count of number of warnings or errors encountered.
Reads the TrueType font file $ttf_file and the attachment point database (APDB) file $ap_file, and builds a structure to represent the APDB.
$ttf_file
$ap_file
Options that may be supplied throught the %opts hash include:
%opts
A list of attachment point types to ignore. Can be a string containing comma- or space-separated names, or a ref to an array of strings.
If true, warn about attachment points that do not correspond to appropriate points on the outline of the glyph.
If this option is specified, read_font will warn if glyphs that should have outlines don't. The option value should be a list of names of glyphs that are known to have no outline (thus shouldn't generate warning). Can be a string containing comma- or space-separated names, or a ref to an array of strings.
A file handle to which warning messages are to be printed. If not supplied, warning messages are accumulated in WARNINGS.
WARNINGS
An alternative to "make_classes", this method just creates name records for all the glyphs in the font. That is, for every glyph record in glyphs, "make_names" invokes "make_name" and saves the result in the name element of the glyph.
glyphs
name
First, for every glyph record in glyphs, make_classes invokes make_name followed by, for every attachment point record in points, make_point . This gives sub-classes a chance to convert the names (of glyphs and points) to an alternate form (e.g., as might be useful in building Graphite source.) See GDL.pm for an example.
make_classes
make_name
points
make_point
make_classes then builds the classes and lists instance variables, and updates the vecs instance variable (if it is defined).
classes
lists
vecs
Options supported are:
Accepts one of two possible values: 'first' or 'last'. First creates ligature classes with the class based on the first element of the ligature and the contents of the class on the rest of the ligature. Last creates classes based on the last element of the ligature, thus grouping all glyphs with the same last ligature element together. Ligature classes are stored in $self-{'ligclasses'}>.
'first'
'last'
$self-
Ligature elements are separated by _ in the glyph name. Ligatures are only made if there are corresponding non ligature glyphs in the font. A final .text on the glyph name of a ligature is assumed to be associated with the whole ligature and not just the last element.
If set to 'comp', then treat glyph name extension as part of the final element.
'comp'
Given a glyph's name, USV, and a reference to its glyph structure, returns a replacement name, e.g., one that might be an acceptable identifier in a programming language. By default this returns $gname, but the function could be overridden when subclassing.
glyph
Given an an attachment point name and a reference to its glyph structure, returns a replacement name, e.g., one that might be an acceptable identifier in a programming language, or undef to indicate the attachment point should be omitted. By default this returns $pname, but the function could be overridden when subclassing.
Provides one possible way to map glyph names in a new font with an existing project.
Compares the list of glyphs found during "read_font" (the new glyph list) with the list from "parse_volt" in $dat (the old glyph list>, returning an array that provides a mapping between old and new glyph IDs. The array is indexed by old glyph ID and returns the new glyph ID.
This implementation works by sdiffing the two ordered lists of glyph names and then matching them up, first by name then by uni.
sdiff
Returns an XML representation of the AP database
TTFBuilder, Font::TTF::Font
Martin Hosken Martin_Hosken@sil.org.
Copyright (c) 1998-2013, SIL International (http://www.sil.org)
This module is released under the terms of the Artistic License 2.0. For details, see the full text of the license in the file LICENSE.
To install Font::Fret, copy and paste the appropriate command in to your terminal.
cpanm
cpanm Font::Fret
CPAN shell
perl -MCPAN -e shell install Font::Fret
For more information on module installation, please visit the detailed CPAN module installation guide.