fontforge¶

Module attributes¶

fontforge.hooks¶

A dictionary which the user may fill to associate certain FontForge events with a python function to be run when those events happen. The function will be passed the font (or possibly glyph) for which the relevant event occurred.

Hook names

newFontHook: This function will be called when a new font has been created.

loadFontHook: This function will be called when a font is loaded from disk. (if a font has an “initScriptString” entry in its persistent dictionary, that script will be invoked before this function).

Other hooks are defined in a font’s own temporary and persistent dictionaries.

fontforge.splineCorner¶: A point type enumeration of value 0

fontforge.splineCurve¶: A point type enumeration of value 1

fontforge.splineHVCurve¶: A point type enumeration of value 2

fontforge.splineTangent¶: A point type enumeration of value 3

fontforge.spiroG4¶

A spiro point type enumeration of value 1.

A Spiro G4 curve point

fontforge.spiroG2¶

A spiro point type enumeration of value 2.

A Spiro G2 curve point

fontforge.spiroCorner¶

A spiro point type enumeration of value 3.

A Spiro corner point

fontforge.spiroLeft¶

A spiro point type enumeration of value 4.

A Spiro left “tangent” point

fontforge.spiroRight¶

A spiro point type enumeration of value 5.

A Spiro right “tangent” point

fontforge.spiroOpen¶

A spiro point type enumeration of value 6.

This may only be used on the first point in a spiro tuple. It indicates that the tuple describes an open contour.

fontforge.unspecifiedMathValue¶: A constant, used when the value is unspecified

Module functions¶

fontforge.getPrefs(pref_name)¶: returns the value of the named preference item

fontforge.setPrefs(pref_name, value)¶: sets the value of the named preference item

fontforge.hasSpiro()¶: Returns a boolean, True if Raph Levien’s spiro package is available for use in FontForge.

fontforge.savePrefs()¶: Saves the current preference settings

fontforge.loadPrefs()¶: Loads the user’s default preference settings. Not done automatically in a script.

fontforge.defaultOtherSubrs()¶: Sets the type1 PostScript OtherSubrs to the default value

fontforge.readOtherSubrsFile(filename)¶: Sets the type1 PostScript OtherSubrs to the stuff found in the file.

fontforge.loadEncodingFile(filename[, encname])¶: Loads an encoding file, returns the name of the encoding or None. When loading encodings in Unicode consortium format, an encname has to be specefied or the encoding will be ignored and None will be returned.

fontforge.loadNamelist(filename)¶: Loads a namelist

fontforge.loadNamelistDir(dirname)¶: Loads all namelist files in the directory

fontforge.preloadCidmap(filename, registry, order, supplement)¶: Loads a FontForge cidmap file (first three args are strings, last is an integer)

fontforge.printSetup(type[, printer|cmd|file, width, height])¶

Prepare to print a font sample. The first argument may be one of:

lp: Queues postscript output to the printer using lp. You may use the optional second argument to specify the printer name.

lpr: Queues postscript output to the printer using lpr. You may use the optional second argument to specify the printer name.

ghostview: Displays the output using ghostview (or gv). The second argument is ignored.

command: Use a custom shell command to print the output. The second argument should contain the command and its arguments.

ps-file: Dump the postscript output to a file. The second argument specifies the filename.

The third and fourth arguments are optional and specify the page size (in points) for the output. The third argument is the page width and the fourth is the height. These settings remain until changed.

fontforge.nameFromUnicode(uni[, namelist])¶: Finds the glyph name associated with a given unicode codepoint. If a namelist is specified the name will be taken from that.

fontforge.UnicodeAnnotationFromLib(n)¶: Returns the Unicode Annotations for this value as described by www.unicode.org. If there is no unicode annotation for this value, or no library available, then return empty string “”. It can execute with no current font.

fontforge.UnicodeBlockCountFromLib(n)¶: Return the number of Unicode Blocks as described by www.unicode.org. Currently, the blocks are {0..233}, spanning unicode values {uni0..uni10FFFF}. If there is no value, or no library available, then return -1. This can execute with no current font.

fontforge.UnicodeBlockEndFromLib(n)¶: Returns the Unicode Block end value as described by www.unicode.org. Currently, the blocks are {0..233}, spanning unicode values {uni0..uni10FFFF}. If there is no value, or no library available, then return -1. This can execute with no current font.

fontforge.UnicodeBlockNameFromLib(n)¶: Returns the Unicode Block Name as described by www.unicode.org. Currently, the blocks are {0..233}, spanning unicode values {uni0..uni10FFFF}. If there is no value, or no library available, then return empty string “”. This can execute with no current font.

fontforge.UnicodeBlockStartFromLib(n)¶: Returns the Unicode Block start value as described by www.unicode.org. Currently, the blocks are {0..233}, spanning unicode values {uni0..uni10FFFF}. If there is no value, or no library available, then return -1. This can execute with no current font.

fontforge.unicodeFromName(glyphname)¶: Looks up glyph name in its dictionary and if it is associated with a unicode code point returns that number. Otherwise it returns -1

fontforge.UnicodeNameFromLib(n)¶: Returns the Unicode Name for this value as described by www.unicode.org. If there is no unicode name for this value, or no library available, then return empty string “”. It can execute with no current font.

fontforge.UnicodeNamesListVersion()¶: Return the Unicode Nameslist Version (as described by www.unicode.org). libuninameslist is released on a schedule that depends on when www.unicode.org releases new information. These dates do not match FontForge release dates, therefore users might not keep this optional library upto current updates. This instruction can be used to test if the Nameslist library is recent for your script. This function currently works only for libuninameslist ver_0.3.20130501 or later, else it returns empty string “”. This can execute with no current font.

fontforge.UnicodeNames2GetCntFromLib()¶: Return the Total Count of all Names that were corrected with a new name. Errors and corrections happen, therefore names can be corrected in the next Unicode Nameslist version. If there is no libuninameslist ver 0.5 or later available, then return -1.

fontforge.UnicodeNames2GetNxtFromLib(n)¶: Errors and corrections happen, therefore names can be corrected in the next Unicode Nameslist version. With val==unicode value, this function returns -1 if no Names2 exists, or the Nth table location for this unicode value listed in libuninameslist that was corrected to a new name. If there is no libuninameslist ver 0.5 or later, then return -1.

fontforge.UnicodeNames2NxtUniFromLib(n)¶: Errors and corrections happen, therefore names can be corrected in the next Unicode Nameslist version. This function returns the Next Names2 listed in libuninameslist internal table that was corrected to a new name. The internal table of Unicode values is of size UnicodeNames2GetCntFromLib(). If there is no libuninameslist ver 0.5 or later, then return -1.

fontforge.UnicodeNames2FrmTabFromLib(n)¶: Errors and corrections happen, therefore names can be corrected in the next Unicode Nameslist version. This function returns the Next Names2 listed in libuninameslist internal table that was corrected to a new name. The internal table of Unicode values is of size UnicodeNames2GetCntFromLib(). If there is no libuninameslist ver 0.5 or later, then return None.

fontforge.UnicodeNames2FromLib(val)¶: Errors and corrections happen, therefore names can be corrected in the next Unicode Nameslist version. This function returns the Names2 or None based on the unicode value given. If there is no libuninameslist ver 0.5 or later, then return None.

fontforge.scriptFromUnicode(n)¶

Return the script tag for the given Unicode codepoint. So, for ord('Q'), it would return latn. This is most useful with font.addLookup(), like:

# Add a `mark` lookup for an arbitrary glyph...
script = fontforge.scriptFromUnicode(glyph.unicode)
font.addLookup("l1", "gpos_mark2base", None, (("mark",((script,("dflt")),)),))
font.addLookupSubtable("l1", "l1-1")
font.addAnchorClass("l1-1", "top")

fontforge.IsFraction(n)¶: Return 1 if n is a unicode fraction (either a vulgar fraction or other fraction) as described by www.unicode.org. Return 0 if there is no fraction for this value. It can execute with no current font.

fontforge.IsLigature(n)¶: Return 1 if n is a ligature as described by www.unicode.org. Return 0 if there is no unicode ligature for this value. It can execute with no current font.

fontforge.IsOtherFraction(n)¶: Return 1 if n is a unicode fraction (not defined as vulgar fraction) as described by www.unicode.org. Return 0 if there is no fraction for this value. It can execute with no current font.

fontforge.IsVulgarFraction(n)¶: Return 1 if n is a unicode vulgar fraction as described by www.unicode.org. Return 0 if there is no fraction for this value. It can execute with no current font.

fontforge.ucFracChartGetCnt()¶: Returns total count of Fractions found in the Unicode chart as described by www.unicode.org. It can execute with no current font. Note: Count depends on chart version built into FontForge.

fontforge.ucLigChartGetCnt()¶: Returns total count of Ligatures found in the Unicode chart as described by www.unicode.org. It can execute with no current font. Note: Count depends on chart version built into FontForge.

fontforge.ucLigChartGetLoc(val)¶: Returns n for FontForge internal table Unicode val=Ligature[n]. If val does not exist in table, then return -1. Can execute with no current font.

Note

Count depends on chart version built into FontForge.

fontforge.ucLigChartGetNxt(n)¶: Returns FontForge internal table Unicode Ligature[n]. Return -1 if n<0 or n>=ucLigChartGetCnt(). It can execute with no current font.

Note

Count depends on chart version built into FontForge.

fontforge.ucOFracChartGetCnt()¶: Returns total count of non-Vulgar Fractions found in the Unicode chart as described by www.unicode.org. It can execute with no current font.

Note

Count depends on chart version built into FontForge.

fontforge.ucOFracChartGetLoc(val)¶: Returns n for FontForge internal table Unicode val=OtherFraction[n]. If val does not exist in table, then return -1. Can execute with no current font.

Note

Count depends on chart version built into FontForge.

fontforge.ucOFracChartGetNxt(n)¶: Returns FontForge internal table Unicode (non-vulgar) Fraction[n]. Return -1 if n<0 or n>=ucOFracChartGetCnt(). Can execute with no current font.

Note

Count depends on chart version built into FontForge.

fontforge.ucVulChartGetCnt()¶: Returns total count of Vulgar Fractions found in the Unicode chart as described by www.unicode.org. It can execute with no current font.

Note

Count depends on chart version built into FontForge.

fontforge.ucVulChartGetLoc(val)¶: Returns n for FontForge internal table Unicode val=VulgarFraction[n]. If val does not exist in table, then return -1. Can execute with no current font.

Note

Count depends on chart version built into FontForge.

fontforge.ucVulChartGetNxt(n)¶: Returns FontForge internal table Unicode Vulgar Fraction[n]. Returns -1 if n<0 or n>=ucVulChartGetCnt(). Can execute with no current font.

Note

Count depends on chart version built into FontForge.

fontforge.SpiroVersion()¶: Returns the version of LibSpiro available to FontForge. Versions 0.1 to 0.5 do not have a method to indicate version numbers, but there is a limited method to estimate versions {‘0.0’..’0.5’}. ‘0.0’ if FontForge has no LibSpiro available. ‘0.1’ if LibSpiro 20071029 is available. ‘0.2’ if LibSpiro 0.2 to 0.5 is available. LibSpiro 0.6 and higher reports back it’s version available.

fontforge.version()¶: Returns FontForge’s version number. This will be a large number like 20070406.

fontforge.runInitScripts()¶: Runs the system or user initialization scripts, if not already run. This is primarily intended when importing FontForge into a python process.

fontforge.scriptPath()¶: Returns a tuple listing the directory paths which are searched for python scripts during FontForge initialization.

fontforge.fonts()¶: Returns a tuple of all fonts currently loaded into FontForge for editing

fontforge.activeFont()¶: If the script were invoked from the File->Execute Script… dialog, or invoked by a menu item in the font view, this returns the font that was active at the time. Otherwise it returns None.

fontforge.activeGlyph()¶: If the script were invoked from the File->Execute Script… dialog or a menu item from an outline glyph window or a glyph import/export command this returns the glyph that was active at the time. Otherwise it returns None.

fontforge.activeLayer()¶: This returns the currently active layer as an integer between 0 (inclusive) and the font/glyph’s layer count (exclusive). It may also be set to -1 if the current glyph window is displaying the font’s guideline layer.

fontforge.fontsInFile(filename)¶: Returns a tuple of all font names found in the specified file. The tuple may be empty if FontForge couldn’t find any.

fontforge.open(filename[, flags])¶: Opens a filename and returns the font it contains (if any). If the flags argument is 4, then ff will load all glyphs in the ‘glyf’ table of a ttc file (rather than just the glyphs used in the font picked). This will not load all ‘glyf’ tables though.

fontforge.parseTTInstrs(string)¶

Returns a binary string each byte of which corresponds to a truetype instruction. The input string should contain a set of instruction names as

"SRP0 MIRP[min,rnd,black]"

fontforge.unParseTTInstrs(sequence)¶: Reverse of the above. Converts a binary string into a human (sort of) readable string

fontforge.unitShape(n)¶: Returns a closed contour which is a regular n-gon. The contour will be inscribed in the unit circle. If n is negative, then the contour will be circumscribed around the unit circle. A value of 0 will produce a unit circle. If n==1 it is treated as if n were -4 – a circumscribed square where each side is 2 units long (this is for historical reasons). Behavior is undefined for n=2,-1,-2.

fontforge.registerGlyphSeparationHook(hook)¶: The GlyphSeparationHook is a python routine which FontForge will call when it wants to figure out the optical separation between two glyphs. If you never call this, or if you call it with a value of None FontForge will use a built-in default. This routine gets called during AutoWidth, AutoKern, and computing the optical left and right side bearings (for ‘lfbd’ and ‘rtbd’ features). For more information see its own section.

User Interface Module Functions¶

Users may define scripts to be run when menu items are invoked. Some of these scripts will want to ask users questions, so this section provides routines to determine if FontForge has a user interface, a command to add menu items, and various small standard dialogs to interact with the user. I do not currently provide a mechanism for allowing people to define special purpose dialogs (for example they might want to ask more than one question in a dialog, and I don’t support that).

When FontForge starts (if it’s a FontForge with python) it will look at the directories $(PREFIX)/share/fontforge/python and ~/.FontForge/python and attempt to run all files in those directories which end in ".py". Presumably these files will allow people to customize the user interface to suit their needs.

Currently it reads the files in directory order (which is generally somewhere between creation order and totally random). It will read the system directory before the user directory.

Example

import fontforge

def nameGlyph(junk, glyph):
   print(glyph.glyphname)

fontforge.registerMenuItem(nameGlyph, None, None, "Glyph", None, "Print Glyph Name")

def neverEnableMe(junk, glyph):
   return False

fontforge.registerMenuItem(nameGlyph, neverEnableMe, None, "Glyph", None, "SubMenu", "Print Glyph Name")

def importGlyph(junk, glyph, filename, toback):
   print("Import")
   print(glyph.glyphname)
   print(filename)

def exportGlyph(junk, glyph, filename):
   print("Import")
   print(glyph.glyphname)
   print(filename)

fontforge.registerImportExport(importGlyph, exportGlyph, None, "foosball", "foo", "foo,foobar")

The first call will define a menu item in the Tools menu of the Glyph window. The menu will be called “Print Glyph Name”. It has no shortcut to invoke it. It needs no external data. It is always enabled. And when activated it will invoke the function “nameGlyph” which prints the name of the glyph in the window from which the command is invoked.

The second call defines a menu item in a submenu of the Tools menu. This submenu is called “SubMenu”. This item will never be enabled – but if it were enabled it would again call “nameGlyph” to print the name of the current glyph.

The last provides a way to import and export files of type “foosball” (or it would if the routines did anything).

Not a very useful example.

fontforge.hasUserInterface()¶: Returns True if this session of FontForge has a user interface

fontforge.registerMenuItem(menu_function, enable_function, data, which_window, shortcut_string, {submenu_names, } menu_name_string)¶

If FontForge has a user interface this will add this menu item to FontForge’s Tool menu, either in the font or the outline glyph view (or both).

menu-function: This is the function that will be called when the menu item is activated. It will be passed two arguments, the first is the data value specified here (which may be None, indeed will probably usually be None), and the second is the glyph or font (depending on the window type) from which the menu item was activated. Its return value is ignored.

enable_function: This may be None – in which case the menu item will always be enabled. Otherwise it will be called before the menu pops up on the screen to determine whether this item should be enabled. It will be passed the same arguments as above. It should return True if the item should be enabled and False otherwise.

data: This can be whatever you want (including None). FontForge keeps track of it and passes it to both of the above functions. Use it if you need to provide some context for the menu item.

which_window: May be either of the strings "Font" or "Glyph" (or the tuple ("Font", "Glyph")) and it determines which type of window will have this menu item in its “Tools” menu.

shortcut-string: Warning

Deprecated?

May be None if you do not wish to supply a shortcut. Otherwise should be a string like “Menu Name|Cntl-H” (the syntax is defined in the translation section).

submenu-names: You may specify as many of these as you wish (including leaving them out altogether), this allows you to organize the Tools menu into submenus. (If a submenu of this name does not currently exist, FontForge will create it).

menu-name: The name that will appear in the menu for this item. This will only affect windows created after this command is executed. Normally the command will be executed at startup and so it will affect all windows.

fontforge.registerImportExport(import_function, export_function, data, name, extension[, extension_list])¶

This will add the capability to import or export files of a given type, presumably a way of specifying the splines in a given glyph.

import-function: The function to call to import a file into a glyph. It will be called with: The data argument (specified below), A pointer to the glyph into which the import is to happen, A filename, A flag indicating whether the import should go to the background layer or foreground. This function may be None. In which case there is no import method for this file type.

export-function: The function to call to export a glyph into a file. It will be called with: The data argument (see below), a pointer to the glyph, and a filename. This function may be None, in which case there is no export method for this file type.

data: Anything you like (including None). It will be passed to the import/export routines and can provide them with context if they need that. name The name to be displayed in the user interface for this file type. This may just be the extension, or it might be something more informative.

extension: This is the default extension for this file type. It is used by the export dialog to pick an extension for the generated filename.

extension-list: Some file types have more than one common extension (eps files are usually named “eps”, but I have also seen “ps” and “art” used). The import dialog needs to filter all possible filenames of this file type. This argument should be a comma separated list of extensions. It may be omitted, in which case it defaults to being the same as the “extension” argument above.

fontforge.logWarning(msg)¶: Adds the message (a string) to FontForge’s Warnings window. (if you wish to display a % character you must represent it as two percents). If there is no user interface the output will go to stderr.

fontforge.postError(win_title, msg)¶: Creates a popup dialog to display the message (a string) in that dlg. (if you wish to display a % character you must represent it as two percents). If there is no user interface the output will go to stderr.

fontforge.postNotice(win_title, msg)¶: Creates a little window which will silently vanish after a minute or two and displays the message (a string) in that window. (if you wish to display a % character you must represent it as two percents). If there is no user interface the output will go to stderr.

fontforge.openFilename(question[, def_name[, filter]])¶

All arguments are strings. The first is a question asked to the user (for which a filename to open is presumed to be the answer). The second is optional and provides a default filename. The third is optional and provides a filter (like “*.sfd”)

The result is either a filename or None if the user canceled the dialog.

Throws an exception if there is no user interface.

fontforge.saveFilename(question[, def_name[, filter]])¶

All arguments are strings. The first is a question asked to the user (for which a filename to save something to is presumed to be the answer). The second is optional and provides a default filename. The third is optional and provides a filter (like “*.sfd”)

The result is either a filename or None if the user canceled the dialog.

Throws an exception if there is no user interface.

fontforge.ask(title, question, answers[, def, cancel])¶

Allows you to ask the user a multiple choice question. It pops up a dialog posing the question with a list of buttons ranged underneath it – one for each answer.

The first argument is the dialog’s title, the second is the question to be asked, the third is a tuple of strings – each string will become a button, the fourth and fifth arguments are option, the fourth is the index in the answer array that will be the default answer (the one invoked if the user presses the [Return] key), and the fifth is the answer invoked if the user presses the [Escape] key. If omitted the default answer will be the first, and the cancel answer will be the last.

The function returns the index in the answer array of the answer chosen by the user.

Throws an exception if there is no user interface.

fontforge.askChoices(title, question, answers[, default=, multiple=])¶

Similar to the above allows you to ask the user a multiple choice question. It pops up a dialog posing the question with a scrollable list of choices – one for each answer.

The first argument is the dialog’s title, the second is the question to be asked, the third is a tuple of strings – each string will become a button, the fourth and fifth arguments are option, the fourth is the index in the answer array that will be the default answer (the one invoked if the user presses the [Return] key). If omitted the default answer will be the first.

The fifth argument means that multiple options can be selected. If true, the fourth argument should be a tuple of Boolean values or a single integer index into the answer tuple. So, if there are three options, it should look like (True, False, True), which would select the first and last option.

The function returns the index in the answer array of the answer chosen by the user. If the user cancels the dialog, a -1 is returned.

default and multiple may be passed by position if desired.

Throws an exception if there is no user interface.

fontforge.askString(title, question[, def_string])¶

Allows you to as the user a question for which a string is the answer.

The first argument is the dialog’s title, the second is the question to be asked, the third is optional and specified a default answer.

The function returns the string the user typed or None if they cancelled the dialog.

Throws an exception if there is no user interface.

Point¶

class fontforge.point([x, y, on_curve, type, selected])¶

Creates a new point. Optionally specifying its x,y location, on-curve status and selected status. x and y must be supplied together.

point.x¶: The x location of the point

point.y¶: The y location of the point

point.on_curve¶: Whether this is an on curve point or an off curve point (a control point)

point.selected¶: The value of this member also determines the selected status in the UI on setting a layer. FontForge usually retains the selection status of any point between and that of an on-curve point when saving, whether or not a UI is present.

point.type¶

For an on-curve point, its FontForge point type.

There are four types: fontforge.splineCorner, fontforge.splineCurve, fontforge.splineHVCurve and fontforge.splineTangent.

A new point will have type splineCorner. When assigning a layer to glyph.layers, glyph.background or glyph.foreground the type value is ignored. To influence the type FontForge will associate with an on-curve point use glyph.setLayer().

point.interpolated¶

True if FontForge treats this (quadratic, on-curve) point as interpolated. All interpolated points should be mid-way between their off-curve points, but some such points are not treated as interpolated. This flag is ignored when setting a layer.

Older versions of FontForge omitted interpolated points. This was equivalent to executing the following on a contour:

c[:] = [ p for p in c if not p.interpolated ]

This member will be false for a point marked “Never interpolate” in FontForge but there is currently no way of setting or preserving that mark when a layer is replaced using the Python interfaces. A “round trip” through a Python contour therefore clears that mark on any points that have it, and FontForge treats mid-placed and omitted on_curve points as equivalent.

point.name¶: The point name (generally there is no name)

point.dup()¶: Returns a copy of the current point.

point.transform(tuple)¶: Transforms the point by the transformation matrix

point.__reduce__()¶: This function allows the pickler to work on this type. I don’t think it is useful for anything else.

Contour¶

A contour is a collection of points. A contour may be either based on cubic or quadratic splines.

If based on cubic splines there should be either 0 or 2 off-curve points between every two on-curve points. If there are no off-curve points then we have a line between those two points. If there are 2 off-curve points we have a cubic bezier curve between the two end points.

If based on quadratic splines things are more complex. Again, two adjacent on-curve points yield a line between those points. Two on-curve points with an off-curve point between them yields a quadratic bezier curve. However if there are two adjacent off-curve points then an on-curve point will be interpolated between them. (This should be familiar to anyone who has read the truetype ‘glyf’ table docs).

For examples of what these splines can look like see the section on bezier curves.

A contour may be open in which case it is just a long wiggly line, or closed when it is more like a circle with an inside and an outside. Unless you are making stroked fonts all your contours should eventually be closed.

Contours may also be expressed in terms of Raph Levien’s spiro points. This is an alternate representation for the contour, and is not always available (Only if fontforge.hasSpiro() is True. If available the spiro member will return a tuple of spiro control points, while assigning to this member will change the shape of the contour to match the new spiros.

Two contours may be compared to see if they describe similar paths.

class fontforge.contour¶: Creates a new contour.

contour.is_quadratic¶: Whether the contour should be interpretted as a set of quadratic or cubic splines. Setting this value has the side effect of converting the point list to the appropriate format.

contour.closed¶: Whether the contour is open or closed.

contour.name¶: The contour name (generally there is no name).

contour.spiros¶

This is an alternate representation of a curve. This member is only available if fontforge.hasSpiro() is True. Returns a tuple of spiro control points. Each of these is itself a tuple of four elements; an x,y location, a type field, and a set of flags. The type field takes on values (which are predefined constants in the fontforge module):

fontforge.spiroG4
fontforge.spiroG2
fontforge.spiroCorner
fontforge.spiroLeft
fontforge.spiroRight
fontforge.spiroOpen

For more information on what these point types mean see Raph Levien’s work.

The flags argument is treated as a bitmap of which currently on one bit (0x1) is defined. This indicates that this point is selected in the UI.

When you assign a tuple of spiro control points to this member, the point list for the Bezier interpretation of the contour will change. And when you change the Bezier interpretation the set of spiro points will change.

contour.dup()¶: Returns a deep copy of the contour. That is, it copies the points that make up the contour.

contour.isEmpty()¶: Returns whether the contour is empty (contains no points)

contour.boundingBox()¶: Returns a tuple representing a rectangle (xmin,ymin, xmax,ymax) into which the contour fits. It is not guaranteed to be the smallest such rectangle, but it will often be.

contour.getSplineAfterPoint(pos)¶: Returns a tuple of two four-element tuples. These tuples are x and y splines for the curve after the specified point.

contour.draw(pen)¶: Draw the contour to the pen argument.

contour.__reduce__()¶: This function allows the pickler to work on this type. I don’t think it is useful for anything else.

contour.__iter__()¶: Returns an iterator for the contour which will return the points in order.

Sequence Protocol

Does not support the repeat concept.

len(c): The number of points in the contour

c[i]: The i th point on the contour. You may assign a point or point initializer to this (keeping the number of points constant) or use del c[i] to remove the entry (reducing the number by one).

c[i:j]: The contour containing points between i and j; i must be >= j. Alternatively, c[j:i:-1] returns those points in reverse order (larger strides are not supported).

c[i:j] = d: The points between i and j are replaced by those in d; i must be >= j. d can be a contour or a sequence of point initializer tuples, as in [(1,1,False),(2,1)]. If c[j:i:-1] is used instead the points of d are assigned in reverse order.

c + d: A contour concatenating c and d. d may be or encode either another contour or a point.

c += d: Appends d to c. d may be or encode either another contour or a point.

p in c: When p is a point, returns whether some point (p.x, p.y) is in the contour c. p can also be a tuple of two numbers.

Contour construction

contour.moveTo(x, y)¶: Adds an initial, on-curve point at (x,y) to the contour

contour.lineTo(x, y[, pos])¶: Adds an line to the contour. If the optional third argument is give, the line will be added after the pos’th point, otherwise it will be at the end of the contour.

contour.cubicTo((cp1x, cp1y)(cp2x, cp2y)(x, y)[, pos])¶: Adds a cubic curve to the contour. If the optional third argument is give, the line will be added after the pos’th point, otherwise it will be at the end of the contour.

contour.quadraticTo((cpx, cpy)(x, y)[, pos])¶: Adds a quadratic curve to the contour. If the optional third argument is give, the line will be added after the pos’th point, otherwise it will be at the end of the contour.

contour.insertPoint(point[, pos])¶: Adds point to the contour. If the optional third argument is give, the line will be added after the pos’th point, otherwise it will be at the end of the contour. The point may be either a point or a point initializer tuple.

contour.makeFirst(pos)¶: Rotate the point list so that the pos’th point becomes the first point

contour.isClockwise()¶: Returns whether the contour is drawn in a clockwise direction. A return value of -1 indicates that no consistant direction could be found (the contour self-intersects).

contour.reverseDirection()¶: Reverse the order in which the contour is drawn (turns a clockwise contour into a counter-clockwise one). See also layer.correctDirection().

contour.similar(other_contour[, error])¶

Checks whether this contour is similar to the other one where error is the maximum distance (in em-units) allowed for the two contours to diverge.

This is like the comparison operator, but that doesn’t allow you to specify an error bound.

contour.xBoundsAtY(ybottom[, ytop])¶: Finds the minimum and maximum x positions attained by the contour when y is between ybottom and ytop (if ytop is not specified it is assumed the same as ybottom). If the contour does not have any y values in the specified range then ff will return None.

contour.yBoundsAtX(xleft[, xright])¶: Finds the minimum and maximum y positions attained by the contour when x is between xleft and xright (if xright is not specified it is assumed the same as xleft). If the contour does not have any x values in the specified range then ff will return None.

Contour manipulation

contour.addExtrema([flags, emsize])¶

Extrema should be marked by on-curve points. If a curve lacks a point at an extrema this command will add one. Flags may be one of the following strings:

all: Add all missing extrema

only_good: Only add extrema on longer splines (with respect to the em-size)

only_good_rm: As above but also merge away on-curve points which are very close to, but not on, an added extremum

contour.cluster([within, max])¶

Moves clustered coordinates to a standard central value.

Layer¶

A layer is a collection of contours. All the contours must be the same order (all quadratic or all cubic). Currently layers do not contain references.

Layers may be compared to see if their contours are similar.

class fontforge.layer¶: Creates a new layer

layer.is_quadratic()¶: Whether the contours should be interpretted as a set of quadratic cubic splines. Setting this value has the side effect of converting the contour list to the appropriate format.

layer.__iter__()¶: Returns an iterator for the layer which will return the contours in order.

layer.__reduce__()¶: This function allows the pickler to work on this type. I don’t think it is useful for anything else.

layer.dup()¶: Returns a deep copy of the layer. That is, it will copy all the contours and all the points as well as copying the layer object itself.

layer.isEmpty()¶: Returns whether the layer is empty (contains no contour)

layer.addExtrema([flags, emsize])¶

Extrema should be marked by on-curve points. If a curve lacks a point at an extrema this command will add one. Flags may be one of the following strings:

all: Add all missing extrema

only_good: Only add extrema on longer splines (with respect to the em-size)

only_good_rm: As above but also merge away on-curve points which are very close to, but not on, an added extremum

layer.cluster([within, max])¶

Moves clustered coordinates to a standard central value.

Glyph Pen¶

This implements the Pen Protocol to draw a FontForge glyph. You create a glyphPen with glyph.glyphPen(). You then draw into it with the operators below.

This type may not be pickled.

Example

import fontforge
font = fontforge.open("Ambrosia.sfd") # Open a font
pen = font["B"].glyphPen()            # Create a pen to draw into glyph "B"
pen.moveTo((100,100))                 # draw a square
pen.lineTo((100,200))
pen.lineTo((200,200))
pen.lineTo((200,100))
pen.closePath()                       # end the contour

font["A"].draw(pen)                   # or you can copy from one glyph to another
                                      # by having a glyph draw itself into the pen
pen = None                            # Finalize the pen. This tells FontForge
                                      # that the drawing is done and causes
                                      # it to refresh the display (if a UI is active).

class fontforge.glyphPen¶

glyphPen.moveTo((x, y))¶: With one exception this call begins every contor and creates an on curve point at (x,y) as the start point of that contour. This should be the first call after a pen has been created and the call that follows a glyphPen.closePath(), glyphPen.endPath().

glyphPen.lineTo((x, y))¶: Draws a line from the last point to (x,y) and adds that to the contour.

glyphPen.curveTo((cp1.x, cp1.y), (cp2.x, cp2.y), (x, y)) ((cp.x, cp.y), (x, y))¶

This routine has slightly different arguments depending on the type of the font. When drawing into a cubic font (PostScript) use the first set of arguments (with two control points – off curve points – between each on curve point). When drawing into a quadratic font (TrueType) use the second format with one control point between adjacent on-curve points.

The standard appears to support super-bezier curves with more than two control points between on-curve points. FontForge does not. Nor does FontForge allow you to draw a quadratic spline into a cubic font, nor vice versa.

glyphPen.qCurveTo([(cp.x, cp.y)]*, (x, y)) ([(cp.x, cp.y)]*, None))¶

This routine may only be used in quadratic (TrueType) fonts and has two different formats. It is used to express the TrueType idiom where an on-curve point mid-way between its control points may be omitted, leading to a run of off-curve points (with implied but unspecified on-curve points between them).

The first format allows an arbetary number of off-curve points followed by one on-curve point.

It is possible to have a contour which consists solely of off-curve points. When this happens the contour is NOT started with a glyphPen.moveTo(), instead the entire contour, all the off curve points, are listed in one call, and the argument list is terminated by a None to indicate there are no on-curve points.

glyphPen.closePath()¶: Closes the contour (connects the last point to the first point to make a loop) and ends it.

glyphPen.endPath()¶: Ends the contour without closing it. This is only relevant if you are stroking contours.

glyphPen.addComponent(glyph_name, transform)¶: Adds a reference (a component) to the glyph. The PostScript transformation matrix is a 6 element tuple.

Glyph¶

The glyph type refers to a glyph object. It has no independent life of its own, it always lives within a font. It has all the things you expect to be associated with a glyph: a glyph name, a unicode encoding, a drawing layer, GPOS/GSUB features…

This type may not be pickled.

This type may not be created directly – all glyphs are bound to a font and must be created through the font.

class fontforge.glyph¶: Note: Glyphs do not have an independent existence. They live in fonts. You may not create them stand-alone, only in the context of a font. See font.createChar().

glyph.activeLayer¶: Returns currently active layer in the glyph (as an integer). May be set to an integer or a layer name to change the active layer.

glyph.altuni¶

Returns additional unicode code points for this glyph. For a primary code point, see glyph.unicode .

Returns either None or a tuple of alternate encodings. Each alternate encoding is a tuple of

(unicode-value, variation-selector, reserved-field)

The first is an unicode value of this alternate code point. The second is an integer for variation selector and can be set to -1 if not used. The third is an empty field reserved for future use and currently must be set to zero.

glyph.altuni can be set to None to clear all alternates, or to a tuple. The elements of the tuple may be either integers (an alternate unicode value with no variation selector) or a tuple with up to 3 values in it as explained above.

glyph.anchorPoints¶

Returns the list of anchor points in the glyph. Each anchor point is a tuple of

(anchor-class-name, type, x,y [,ligature-index])

The first two are strings, the next two doubles, and the last (which is only present if type=="ligature") is an integer. Type may be

mark
base
ligature
basemark
entry
exit

glyph.anchorPointsWithSel¶

Same as the above, except also includes whether the anchor point is selected in the UI. Returns a tuple of all anchor points in the glyph. Each anchor point is a tuple of

(anchor-class-name, type, x,y, selected [,ligature-index])

The first two are strings, the next two doubles, then a boolean, and the last (which is only present if type=="ligature") is an integer. Type may be

mark
base
ligature
basemark
entry
exit

glyph.background¶: The glyph’s background layer. This is a copy of the glyph’s data. See also glyph.foreground and glyph.layers.

glyph.changed¶: Whether this glyph has been modified. This is (should be) maintained automatically, but you may set it if you wish.

glyph.color¶: The color of the glyph in the fontview. A 6 hex-digit RGB number or -1 for default. 0xffffff is white, 0x0000ff is blue, etc.

glyph.comment¶: Any comment you wish to associate with the glyph. UTF-8

glyph.dhints¶: A tuple with one entry for each diagonal stem hint. Each stem hint is itself represented by a tuple of three coordinate pairs (themselves tuples of two numbers), these three are: a point on one side of the stem, a point on the other side, and a unit vector pointing in the stem’s direction.

glyph.encoding¶

Returns the glyph’s encoding in the font’s encoding. (readonly)

If the glyph has multiple encodings, one will be picked at random.

If the glyph is not in the font’s encoding then a number will be returned beyond the encoding size (or in some cases -1 will be returned).

glyph.font¶: The font containing this glyph. (readonly)

glyph.foreground¶: The glyph’s foreground layer. This is a copy of the glyph’s data. See also glyph.background, glyph.layers and glyph.references.

glyph.glyphclass¶: An opentype glyphclass, one of automatic, noclass, baseglyph, baseligature, mark, component

glyph.glyphname¶: The name of the glyph

glyph.hhints¶: A tuple of all horizontal postscript hints. Each hint is itself a tuple of starting locations and widths.

glyph.horizontalComponents¶

A tuple of tuples.

This allows constructing very large versions of the glyph by stacking the componants together. Some components may be repeated so there is no bound on the size.

This is different from horizontalVariants which expects prebuilt glyphs of various fixed sizes.

The components are stacked in the order they appear in the (top-level) tuple. Each sub-tuple represents information on one component. The subtuple should contain: (String glyph-name, Boolean is-extender, Int startConnectorLength, Int endConnectorLength, Int fullAdvance). Any of these may be omitted (except the glyph name) and will be assumed to be 0 if so.

glyph.horizontalComponentItalicCorrection¶: The italic correction for any composite glyph made with the horizontalComponents.

glyph.horizontalVariants¶

A string containing a list of glyph names. These are alternate forms of the current glyph for use in typesetting math. Presumably the variants are of different sizes.

Although ff will always return a string of glyph names, you may assign to it with a tuple of glyphs and ff will convert that to corresponding names.

glyph.isExtendedShape¶: A boolean containing the MATH “is extended shape” field.

glyph.italicCorrection¶: The glyph’s italic correction field. Used by both TeX and MATH. The special value fontforge.unspecifiedMathValue means the value is unspecified (An unspecified value will not go into the output tables, a value of 0 will)

glyph.layer_cnt¶: The number of layers in this glyph. (Cannot be set)

glyph.layers¶: A dictionary like object containing the layers of the glyph. It may be indexed by either a layer name, or an integer between 0 and glyph.layer_cnt-1 to produce a layer object. Layer 0 is the background layer. Layer 1 is the foreground layer.

glyph.layerrefs¶: A dictionary like object containing the references in the layers of the glyph. It may be indexed by either a layer name, or an integer between 0 and glyph.layer_cnt-1 to produce a reference tuple object. Layer 0 is the background layer. Layer 1 is the foreground layer.

glyph.lcarets¶: A tuple containing the glyph’s ligature caret locations. Setting this will also either enable or disable the “Default Ligature Caret Count” flag depending from the number of elements in the tuple.

glyph.left_side_bearing¶: The left side bearing of the glyph. Setting this value will adjust all layers so that guides in the background etc will be adjusted with the rest of the glyph

glyph.manualHints¶: The glyph’s hints have been set by hand, and the glyph should not be autohinted without a specific request from the user. The “Don’t AutoHint” flag.

glyph.mathKern.bottomLeft¶: The glyph’s math kerning data associated with the bottom left vertex. This returns a tuple of two element tuples, each of which contains a kerning offset and an associated height (in the last entry the height term is meaningless, but present).

glyph.mathKern.bottomRight¶: The glyph’s math kerning data associated with the bottom right vertex. This returns a tuple of two element tuples, each of which contains a kerning offset and an associated height (in the last entry the height term is meaningless, but present).

glyph.mathKern.topLeft¶: The glyph’s math kerning data associated with the top left vertex. This returns a tuple of two element tuples, each of which contains a kerning offset and an associated height (in the last entry the height term is meaningless, but present).

glyph.mathKern.topRight¶: The glyph’s math kerning data associated with the top right vertex. This returns a tuple of two element tuples, each of which contains a kerning offset and an associated height (in the last entry the height term is meaningless, but present).

glyph.originalgid¶: The GID of this glyph in the font it was read from. (readonly)

glyph.persistent¶: Whatever you want (these data will be saved as a pickled object in the sfd file. It is your job to insure that whatever you put here can be pickled). See also the glyph.temporary field.

glyph.references¶: A tuple of tuples containing glyph-name and a transformation matrix for each reference in the foreground. See also glyph.foreground and glyph.layerrefs.

glyph.right_side_bearing¶: The right side bearing of the glyph

glyph.script¶: A string containing the OpenType 4 letter tag for the script associated with this glyph (readonly)

glyph.temporary¶

Whatever you want (these data will be lost once the font is closed)

Selection¶

This represents a font’s selection. You may index it with an encoding value (in the encoding ISO-646-US (ASCII) the character “A” has encoding index 65), or with a glyph’s name, or with a string like "uXXXXX" where XXXXX represent the glyph’s unicode codepoint in hex, or with a fontforge.glyph object. The value of indexing into a selection will be either True or False.

>>> print(fontforge.activeFont().selection[65])
True

This type may not be pickled.

class fontforge.selection¶

selection.byGlyphs¶

Returns another selection, just the same as this one except that its iterator function will return glyphs (rather than encoding slots) and will only return those entries for which glyphs exist.

This is read-only.

selection.__iter__()¶: Returns an iterator for the selection which will return all selected encoding slots in encoding order.

selection.all()¶: Select everything.

selection.none()¶: Deselect everything.

selection.changed()¶: Select all glyphs which have changed.

selection.invert()¶: Invert the selection.

selection.select(args)¶

There may be an arbitrary number of arguments. Each argument may be either:

A glyph name

Note: There need not be a glyph with this name in the font yet, but if you use a standard name (like “A”) fontforge will still know where that glyph should be.
An integer (this will be interpreted as either an encoding index or (default) a unicode code point depending on the flags).
A fontforge glyph.
A tuple of flags.

(If you wish to specify a single flag it must still be in a tuple, and you must append a trailing comma to the flag (so ("more",) rather than just ("more") ). FF needs the flags to be in a tuple otherwise it can’t distinguish them from glyph names)

unicode

Interpret integer arguments as unicode code points

encoding

Interpret integer arguments as encoding indeces.

more

Specified items should be selected

less

Specified items should be deselected.

singletons

Specified items should be interpreted individually and mean the obvious.

ranges

Specified items should be interpreted in pairs and represent all encoding slots between the start and end points specified by the pair. So .select(("ranges",None),"A","Z") would select all the upper case (latin) letters.

If the first argument is not a flag argument (or if it doesn’t specify either “more” or “less”) then the selection will be cleared. So .select("A") would produce a selection with only “A” selected, .select(("more",None),"A") would add “A” to the current selection, while .select(("less",None),"A") would remove “A” from the current selection.

Private¶

This represents a font’s postscript private dictionary. You may index it with one of the standard names of things that live in the private dictionary.

This type may not be pickled.

class fontforge.private¶

private.__iter__()¶: Returns an iterator for the dictionary which will return all entres.

private.guess(name)¶: Guess a value for this this entry in the private dictionary. If FontForge can’t make a guess it will simply ignore the request.

Math¶

This represents a font’s math constant table. Not all fonts have math tables, and checking this field will not create the underlying object, but examining or assigning to its members will create it.

This type may not be pickled.

Members

Any of the math constant names may be used as member names.

(These names begin with capital letters, not Python’s conventions but Microsoft’s)

These all take (16 bit) integer values.

I do not currently provide python access to any associated device tables.

math.ScriptPercentScaleDown¶: Percentage scale down for script level 1

math.ScriptScriptPercentScaleDown¶: Percentage scale down for script level 2

math.DelimitedSubFormulaMinHeight¶: Minimum height at which to treat a delimited expression as a subformula

math.DisplayOperatorMinHeight¶: Minimum height of n-ary operators (integration, summation, etc.)

math.MathLeading¶: White space to be left between math formulae to ensure proper line spacing.

math.AxisHeight¶: Axis height of the font

math.AccentBaseHeight¶: Maximum (ink) height of accent base that does not require raising the accents.

math.FlattenedAccentBaseHeight¶: Maximum (ink) height of accent base that does not require flattening the accents.

math.SubscriptShiftDown¶: The standard shift down applied to subscript elements. Positive for moving downward.

math.SubscriptTopMax¶: Maximum height of the (ink) top of subscripts that does not require moving subscripts further down.

math.SubscriptBaselineDropMin¶: Maximum allowed drop of the baseline of subscripts relative to the bottom of the base. Used for bases that are treated as a box or extended shape. Positive for subscript baseline dropped below base bottom.

math.SuperscriptShiftUp¶: Standard shift up applied to superscript elements.

math.SuperscriptShiftUpCramped¶: Standard shift of superscript relative to base in cramped mode.

math.SuperscriptBottomMin¶: Minimum allowed height of the bottom of superscripts that does not require moving them further up.

math.SuperscriptBaselineDropMax¶: Maximum allowed drop of the baseline of superscripts relative to the top of the base. Used for bases that are treated as a box or extended shape. Positive for superscript baseline below base top.

math.SubSuperscriptGapMin¶: Minimum gap between the superscript and subscript ink.

math.SuperscriptBottomMaxWithSubscript¶: The maximum level to which the (ink) bottom of superscript can be pushed to increase the gap between superscript and subscript, before subscript starts being moved down.

math.SpaceAfterScript¶: Extra white space to be added after each sub/superscript.

math.UpperLimitGapMin¶: Minimum gap between the bottom of the upper limit, and the top of the base operator.

math.UpperLimitBaselineRiseMin¶: Minimum distance between the baseline of an upper limit and the bottom of the base operator.

math.LowerLimitGapMin¶: Minimum gap between (ink) top of the lower limit, and (ink) bottom of the base operator.

math.LowerLimitBaselineDropMin¶: Minimum distance between the baseline of the lower limit and bottom of the base operator.

math.StackTopShiftUp¶: Standard shift up applied to the top element of a stack.

math.StackTopDisplayStyleShiftUp¶: Standard shift up applied to the top element of a stack in display style.

math.StackBottomShiftDown¶: Standard shift down applied to the bottom element of a stack. Positive values indicate downward motion.

math.StackBottomDisplayStyleShiftDown¶: Standard shift down applied to the bottom element of a stack in display style. Positive values indicate downward motion.

math.StackGapMin¶: Minimum gap between bottom of the top element of a stack, and the top of the bottom element.

math.StackDisplayStyleGapMin¶: Minimum gap between bottom of the top element of a stack and the top of the bottom element in display style.

math.StretchStackTopShiftUp¶: Standard shift up applied to the top element of the stretch stack.

math.StretchStackBottomShiftDown¶: Standard shift down applied to the bottom element of the stretch stack. Positive values indicate downward motion.

math.StretchStackGapAboveMin¶: Minimum gap between the ink of the stretched element and the ink bottom of the element above.

math.StretchStackGapBelowMin¶: Minimum gap between the ink of the stretched element and the ink top of the element below.

math.FractionNumeratorShiftUp¶: Standard shift up applied to the numerator.

math.FractionNumeratorDisplayStyleShiftUp¶: Standard shift up applied to the numerator in display style.

math.FractionDenominatorShiftDown¶: Standard shift down applied to the denominator. Positive values indicate downward motion.

math.FractionDenominatorDisplayStyleShiftDown¶: Standard shift down applied to the denominator in display style. Positive values indicate downward motion.

math.FractionNumeratorGapMin¶: Minimum tolerated gap between the ink bottom of the numerator and the ink of the fraction bar.

math.FractionNumeratorDisplayStyleGapMin¶: Minimum tolerated gap between the ink bottom of the numerator and the ink of the fraction bar in display style.

math.FractionRuleThickness¶: Thickness of the fraction bar.

math.FractionDenominatorGapMin¶: Minimum tolerated gap between the ink top of the denominator and the ink of the fraction bar.

math.FractionDenominatorDisplayStyleGapMin¶: Minimum tolerated gap between the ink top of the denominator and the ink of the fraction bar in display style.

math.SkewedFractionHorizontalGap¶: Horizontal distance between the top and bottom elements of a skewed fraction.

math.SkewedFractionVerticalGap¶: Vertical distance between the ink of the top and bottom elements of a skewed fraction.

math.OverbarVerticalGap¶: Distance between the overbar and the ink top of the base.

math.OverbarRuleThickness¶: Thickness of the overbar.

math.OverbarExtraAscender¶: Extra white space reserved above the overbar.

math.UnderbarVerticalGap¶: Distance between underbar and the (ink) bottom of the base.

math.UnderbarRuleThickness¶: Thickness of the underbar.

math.UnderbarExtraDescender¶: Extra white space reserved below the underbar.

math.RadicalVerticalGap¶: Space between the ink to of the expression and the bar over it.

math.RadicalDisplayStyleVerticalGap¶: Space between the ink top of the expression and the bar over it in display style.

math.RadicalRuleThickness¶: Thickness of the radical rule in designed or constructed radical signs.

math.RadicalExtraAscender¶: Extra white space reserved above the radical.

math.RadicalKernBeforeDegree¶: Extra horizontal kern before the degree of a radical if such be present.

math.RadicalKernAfterDegree¶: Negative horizontal kern after the degree of a radical if such be present.

math.RadicalDegreeBottomRaisePercent¶: Height of the bottom of the radical degree, if such be present, in proportion to the ascender of the radical sign.

math.MinConnectorOverlap¶: Minimum overlap of connecting glyphs during glyph construction.

math.exists()¶: Returns whether the font currently has an underlying math table associated with it. Note that examining or assigning to one of the members will create such a table.

math.clear()¶: Removes any underlying math table from the font.

Font¶

The font type refers to a fontforge font object. It generally contains a list of glyphs, an encoding to order those glyphs, a fontname, a list of GPOS/GSUB lookups and many other things.

This type may not be pickled.

class fontforge.font¶: Creates a new font.

font.activeLayer¶: Returns currently active layer in the font (as an integer). May be set to an integer or a layer name to change the active layer.

font.ascent¶: The font’s ascent

font.bitmapSizes¶

A tuple with an entry for each bitmap strike attached to the font. Each strike is identified by pixelsize (if the strike is a grey scale font it will be indicated by (bitmap-depth<<16)|pixelsize.

When setting this value pass in a tuple of the same format. Any existing strike not specified in the tuple will be removed. Any new sizes will be created (but not rasterized – use font.regenBitmaps() for that).

font.capHeight¶: (readonly) Computes the Cap Height (the height of capital letters such as “E”). A negative number indicates the value could not be computed (the font might have no capital letters because it was lower case only, or didn’t include glyphs for a script with capital letters).

font.changed¶: Bit indicating whether the font has been modified. This is (should be) maintained automatically, but you may set it if you wish.

font.cidcopyright¶: Copyright message of the cid-keyed font as a whole (ie. not the current subfont).

font.cidfamilyname¶: Family name of the cid-keyed font as a whole (ie. not the current subfont).

font.cidfontname¶: Font name of the cid-keyed font as a whole (ie. not the current subfont).

font.cidfullname¶: Full name of the cid-keyed font as a whole (ie. not the current subfont).

font.cidordering¶

font.cidregistry¶

font.cidsubfont¶

Returns the number index of the current subfont in the cid-keyed font (or -1 if this is not a cid-keyed font).

May be set to an index (an integer) or a subfont fontname (a string) to change the current subfont. (To find the name of the current subfont, simply use .fontname).

font.cidsubfontcnt¶: Returns the number of subfonts in this cid-keyed font (or 0 if it is not a cid-keyed font)

font.cidsubfontnames¶: Returns a tuple of the subfont names in this cid-keyed font (or None if it is not a cid-keyed font)

font.cidsupplement¶

font.cidversion¶

font.cidweight¶: Weight of the cid-keyed font as a whole

font.comment¶: A comment associated with the font. Can be anything.

font.copyright¶: PostScript copyright notice

font.cvt¶

Returns a sequence object containing the font’s cvt table. Changes made to this object will be made to the font (this is a reference not a copy).

The object has one additional method cvt.find(value[,low,high]) which finds the index of value in the cvt table (or -1 if not found). If low and high are specified then the index will be between [low,high).

font.default_base_filename¶: The default base for the filename when generating a font

font.descent¶: The font’s descent

font.design_size¶: Size (in pica points) for which this font was designed.

font.em¶: The em size of the font. Setting this will scale the entire font to the new size.

font.encoding¶: The name of the current encoding. Setting it will change the encoding used for indexing

font.familyname¶: PostScript font family name

font.fondname¶: Mac fond name

font.fontlog¶: A comment associated with the font. Can be anything.

font.fontname¶

PostScript font name

Note that in a CID keyed font this will be the name of the current subfont. Use cidfontname for the name of the font as a whole.

font.fullname¶: PostScript font name

font.gasp¶

Returns a tuple of all gasp table entries. Each item in the tuple is itself a tuple composed of a ppem (an integer) and a tuple of flags. The flags are chosen from:

gridfit
antialias
symmetric-smoothing
gridfit+smoothing

font.gasp_version¶: The version of the ‘gasp’ table. Currently this may be 0 or 1.

font.gpos_lookups¶: Returns a tuple of all positioning lookup names in the font. This member cannot be set.

font.gsub_lookups¶: Returns a tuple of all substitution lookup names in the font. This member cannot be set.

font.guide¶: A copy of the font’s guide layer

font.hasvmetrics¶

font.head_optimized_for_cleartype¶

font.hhea_ascent¶

font.hhea_ascent_add¶

font.hhea_descent¶

font.hhea_descent_add¶

font.hhea_linegap¶

font.horizontalBaseline¶

Returns a tuple of tuples containing the horizontal baseline information in the font (the ‘BASE’ table). If there is no information None will be returned, otherwise the format of the tuple is:

((tuple of baseline tags used), (tuple of script information))

The (tuple of baseline tags used) is simply a tuple of 4 letter strings as ("hang", "ideo", "romn") these are standard baseline tag names as defined in the opentype spec. The number of entries here, and their order is important as there will be subsequent tuples (in the script tuple) which use the same ordering.

The (tuple of script information) is again a tuple of script information tuples.

A script information tuple looks like

(script-tag,default-baseline-tag, (tuple of baseline positions), (tuple of language extents))

If there are no baseline tags defined (an empty tuple), then the default-baseline-tag and the (tuple of baseline positions) will be None. Otherwise both tags will be 4 character strings, and the (tuple of baseline positions) will be a tuple of numbers (in the same order as the (tuple of baseline tags used) above) specifying the relative positions of each baseline for this script.

A (tuple of language extents) is a tuple of language extent tuples.

A language extent tuple is

(language-tag,min-extent,max-extent, (tuple of feature extents))

language-tag is a 4 letter string specifying an opentype language, min/max-extent are numbers specifying how far above and below the baseline characters go in this script/language.

A (tuple of feature extents> is a tuple of feature extent tuples.

A feature extent tuple is

(feature-tag,min-extent,max-extent)

feature-tag is a 4 letter string specifying an opentype (GPOS or GSUB) feature tag, min/max-extent are numbers specifying how far above and below the baseline characters go in this script/language with the feature applied.

Example:

(("hang","ideo","romn"),
  (("cyrl","romn",(1405,-288,0),()),
   ("grek","romn",(1405,-288,0),()),
   ("latn","romn",(1405,-288,0),
     (("dflt",-576,1913,
       (("NoAc",-576,1482),
        ("ENG ",-576,1482))
     ),
   )
  )
 )
)

(Note: The comma after the dflt tuple puts it into a one-element tuple.)

font.is_cid¶: Indicates whether the font is a cid-keyed font or not. (Read-only)

font.is_quadratic¶

Deprecated. Whether the contours should be interpretted as a set of quadratic or cubic splines. Setting this value has the side effect of converting the entire font into the other format

Now each layer may have its own setting for this value, which should be set on the font’s font.layers.

font.isnew¶: A flag indicating that this is a new font

font.italicangle¶

font.macstyle¶

Bit 0: Bold (if set to 1)

Bit 1: Italic (if set to 1)

Bit 2: Underline (if set to 1)

Bit 3: Outline (if set to 1)

Bit 4: Shadow (if set to 1)

Bit 5: Condensed (if set to 1)

Bit 6: Extended (if set to 1)

Bits 7-15: Reserved (set to 0).

(source)

font.layer_cnt¶: The number of layers in the font. (Read only. Can change using add and del operations on the font.layers array)

font.layers¶

Returns a dictionary like object with information on the layers of the font – a name and a boolean indicating whether the layer is quadratic or not.

You may remove a layer with

del font.layers["unneeded layer"]

You may add a new layer with

font.layers.add("layer-name",is_quadratic[, is_background])

You may change a layer’s name with

font.layers["layer"].name = "new-name"

You may change the type of splines in a layer with

font.layers["layer"].is_quadratic = True

You may change whether it is a background layer by

font.layers["layer"].is_background = True

Note: The layers that live in the font are different from layers that live in a glyph. These objects do not have the Layer type documented earlier.

font.loadState¶

A bitmask indicating non-fatal errors found when loading the font. (readonly)

0x01: Bad PostScript entry in ‘name’ table

0x02: Bad ‘glyf’ or ‘loca’ table

0x04: Bad ‘CFF ‘ table

0x08: Bad ‘hhea’, ‘hmtx’, ‘vhea’ or ‘vmtx’ table

0x10: Bad ‘cmap’ table

0x20: Bad ‘EBLC’, ‘bloc’, ‘EBDT’ or ‘bdat’ (embedded bitmap) table

0x40: Bad Apple GX advanced typography table

0x80: Bad OpenType advanced typography table (GPOS, GSUB, GDEF, BASE)

0x100

Bad OS/2 version number

Windows will reject all fonts with a OS/2 version number of 0 and will reject OT-CFF fonts with a version number of 1

font.maxp_FDEFs¶: The number of function definitions used by the tt program

font.maxp_IDEFs¶: The number of instruction definitions used by the tt program

font.maxp_maxStackDepth¶: The maximum stack depth used by the tt program

font.maxp_storageCnt¶: The number of storage locations used by the tt program

font.maxp_twilightPtCnt¶: The number of points in the twilight zone of the tt program

font.maxp_zones¶: The number of zones used in the tt program

font.multilayer¶

font.onlybitmaps¶: A flag indicating that this font only contains bitmaps. No outlines.

font.os2_codepages¶: A 2 element tuple containing the OS/2 Codepages field

font.os2_family_class¶

font.os2_fstype¶

font.os2_panose¶

font.os2_strikeypos¶

font.os2_strikeysize¶

font.os2_subxoff¶

font.os2_subxsize¶

font.os2_subyoff¶

font.os2_subysize¶

font.os2_supxoff¶

font.os2_supxsize¶

font.os2_supyoff¶

font.os2_supysize¶

font.os2_typoascent¶

font.os2_typoascent_add¶

font.os2_typodescent¶

font.os2_typodescent_add¶

font.os2_typolinegap¶

font.os2_use_typo_metrics¶

font.os2_unicoderanges¶: A 4 element tuple containing the OS/2 Unicode Ranges field

font.os2_vendor¶

font.os2_version¶

font.os2_weight¶

font.os2_weight_width_slope_only¶

font.os2_width¶

font.os2_winascent¶

font.os2_winascent_add¶

font.os2_windescent¶

font.os2_windescent_add¶

font.path¶: (readonly) Returns a string containing the name of the file from which the font was originally read (in this session), or if this is a new font, returns a made up filename in the current directory named something like “Untitled1.sfd”. See also font.sfd_path.

font.persistent¶

Whatever you want – though I recommend you store a dict here (these data will be saved as a pickled object in the sfd file. It is your job to ensure that whatever you put here can be pickled)

If you do store a dict then the following entries will be treated specially:

initScriptString

If present, and if this is a string, then each time the font is loaded from an sfd file, this string will be passed to the python interpreter.

Note

This is a string, not a function. Function code cannot be pickled. Since it is a string it will receive no arguments, but the current font will be available in the activeFont method of the fontforge module.

This string will be interpreted before the loadFontHook of the fontforge.hooks dictionary.

One possible behavior for this string is to define function hooks to be stored in the temporary dict described below.

font.math¶: Returns a math object which provides information on the font’s underlying math constant table. There is only one of these per font.

font.private¶

Returns a private dictionary-like object representing the PostScript private dictionary for the font. Changing entries in this object will change them in the font. (It’s a reference, not a copy).

There is an iterator associated with this entry.

font.privateState¶

Checks the (PostScript) Private dictionary and returns a bitmask of some common errors.

0x000001: Odd number of elements in either the BlueValues or OtherBlues array.

0x000002: Elements in either the BlueValues or OtherBlues are disordered.

0x000004: Too many elements in either the BlueValues or OtherBlues array.

0x000008: Elements in either the BlueValues or OtherBlues array are too close (must be at least 2*BlueFuzz +1 apart).

0x000010: Elements in either the BlueValues or OtherBlues array are not integers.

0x000020: Alignment zone height in either the BlueValues or OtherBlues array is too big for the value of BlueScale.

0x000100: Odd number of elements in either the FamilyBlues or FamilyOtherBlues array.

0x000200: Elements in either the FamilyBlues or FamilyOtherBlues are disordered.

0x000400: Too many elements in either the FamilyBlues or FamilyOtherBlues array.

0x000800: Elements in either the FamilyBlues or FamilyOtherBlues array are too close (must be at least 2*BlueFuzz +1 apart).

0x001000: Elements in either the FamilyBlues or FamilyOtherBlues array are not integers.

0x002000: Alignment zone height in either the FamilyBlues or FamilyOtherBlues array is too big for the value of BlueScale.

0x010000: Missing BlueValues entry.

0x020000: Bad BlueFuzz entry.

0x040000: Bad BlueScale entry.

0x080000: Bad StdHW entry.

0x100000: Bad StdVW entry.

0x200000: Bad StemSnapH entry.

0x400000: Bad StemSnapV entry.

0x800000: StemSnapH does not include StdHW.

0x1000000: StemSnapV does not include StdVW.

0x2000000: Bad BlueShift entry.

font.selection¶: Returns a reference to a array-like object representing the font's selection. There is one entry for each encoding slot (there may not be a glyph attached to every encoding slot). You may set this with a tuple of integers (or boolean values). There should not be more entries in the tuple than there are encoding slots in the current encoding. A True or non-0 value means the slot is selected.

font.sfd_path¶: (readonly) Returns a string (or None) containing the name of the sfd file associated with this font. Sometimes this will be the same as font.path.

font.sfnt_names¶

The strings in the sfnt ‘name’ table. A tuple of all MS names. Each name is itself a tuple of strings (language,strid,string). Language may be either the (english) name of the language/locale, or the number representing that language in Microsoft’s specification. Strid may be one of the (English) string names (Copyright, Family, SubFamily, etc.) or the numeric value of that item. The string itself is in UTF-8.

Mac names will be automagically created from MS names

font.sfntRevision¶

The font revision field stored in the 'head' table of an sfnt. This is documented to be a fixed 16.16 number (that is a 32 bit number with the binary point between bits 15 and 16).

The field may be unset (in which case when the font is generated, FontForge will guess a default value from one of the version strings).

The value returned with be None if the field is unset or a double.

You may set it to None which “unsets” it, or to a double value, or to an integer. The integer will be treated as a 32 bit integer and right shifted by 16 to get a 16.16 value).

font.size_feature¶

The OpenType ‘size’ feature has two formats. It may either represent the design size of the font (and nothing else) or the design size, and range (top and bottom point sizes for which this design works), a style id (used to represent this design size throughout the font family) and a set of language/string pairs used to represent this design size in the menu.

If no size information is specified in the font FontForge will return None.

If only the design size is specified, FontForge will return a tuple containing a single element: the point size for which the font was designed. (This is returned as a real number – the field can represent tenths of a point).

Otherwise FontForge returns a tuple containing five elements, the design size, the bottom of the design range, the top, the style id and a tuple of tuples. Each sub-tuple is a language/string pair. Language may be either the (english) name of the language/locale, or The string itself is in UTF-8.

font.strokedfont¶: is this a stroked font?

font.strokewidth¶: the stroke width of a stroked font

font.temporary¶

Whatever you want – though I recommend you store a dict here (these data will be lost once the font is closed)

If you do store a dict then the following entries will be treated specially:

generateFontPreHook: If present, and if this is a function it will be called just before a font is generated. It will be called with the font and the filename to which the font will be written.

generateFontPostHook: If present, and if this is a function it will be called just after a font is generated. It will be called with the font and the filename to which the font will be written.

font.texparameters¶

Returns a tuple of TeX font parameters. TeX font type followed by 22 parameters. Font type is one of:

text

mathsym

mathext

unset

In case of unset default values for font parameters will be returned.

font.uniqueid¶

font.upos¶: underline position

font.userdata¶: Warning

Deprecated name for font.temporary

font.uwidth¶: underline width

font.version¶: PostScript font version string

font.verticalBaseline¶: Same format as font.horizontalBaseline.

font.vertical_origin¶: Warning

Deprecated

font.vhea_linegap¶

font.weight¶: PostScript font weight string

font.woffMajor¶

The major version number of a woff file, an integer.

The field may be unset (in which case when the font is generated, FontForge will guess a default value from one of the version strings).

The value returned with be None if the field is unset or an integer.

You may set it to None which “unsets” it, or to an integer.

font.woffMinor¶

The minor version number of a woff file, an integer.

The field may be unset (in which case when the font is generated, FontForge will guess a default value from one of the version strings).

The value returned with be None if the field is unset or an integer.

You may set it to None which “unsets” it, or to an integer.

font.woffMetadata¶: Any metadata associated with a woff file. This is a utf8 string containing unparsed xml.

font.xHeight¶: (readonly) Computes the X Height (the height of lower case letters such as “x”). A negative number indicates the value could not be computed (the font might have no lower case letters because it was upper case only, or didn’t include glyphs for a script with lower case letters).

font.__iter__()¶: Returns an iterator for the font which will run through the font, in gid order, returning glyph names

font.__contains__()¶: Returns whether the font contains a glyph with the given name.

font.__len__()¶: The number of glyph slots in the current encoding

font.__getitem__(key)¶: If key is an integer, then returns the glyph at that encoding. If a string then returns the glyph with that name. May not be assigned to.

font.addAnchorClass(lookup_subtable_name, new_anchor_class_name)¶: Adds an anchor class to the specified (anchor) subtable.

font.addKerningClass(lookup_name, new_subtable_name, first_classes, second_classes, offsets[, after])¶

font.addKerningClass(lookup_name, new_subtable_name, separation, first_classes, second_classes[, onlyCloser, autokern, after])

font.addKerningClass(lookup_name, new_subtable_name, separation, class_distance, first_glyph_list, second_glyph_list[, onlyCloser, autokern, after])

font.addKerningClass(lookup_name, new_subtable_name, separation, class_distance[, onlyCloser, autokern, after])

Creates a new subtable and a new kerning class in the named lookup. The classes arguments are tuples of tuples of glyph names (each sub-tuble of glyph names is a kerning class). The offsets argument is a tuple of kerning offsets. There must be as many entries as

len(first-class)*len(second-class)

The optional after argument is used to specify the order of the subtable within the lookup.

The second format will cause FontForge to auto kern the subtable. The separation argument specifies the desired optical distance between any two glyphs (if this is specified as 0 then the kerning class will be designed so glyphs just touch each other). Again the user specifies two sets of predefined classes. If the (optional) onlyCloser flag is set true then only negative kerning values will be inserted into the table.

In the third format the user merely specifies two lists of glyphs to be used, fontforge will look for similarities among among the glyphs and guess at classes. The class-distance argument to determine how precise the classes should match (1 is very tight matching, 20 is rather loose).

In the last format the font’s selection will be used to specify the list of glyphs to be examined (and the same list will be used for both the left and right glyphs – but fontforge will probably find different classes).

font.addLookup(new_lookup_name, type, flags, feature_script_lang_tuple[, after_lookup_name)¶

Creates a new lookup with the given name, type and flags. It will tag it with any indicated features. The type of one of

gsub_single
gsub_multiple
gsub_alternate
gsub_ligature
gsub_context
gsub_contextchain
gsub_revesechain
morx_indic
morx_context
morx_insert
gpos_single
gpos_pair
gpos_cursive
gpos_mark2base
gpos_mark2ligature
gpos_mark2mark
gpos_context
gpos_contextchain
kern_statemachine

The flags argument is a tuple of strings, or None. At most one of these strings may be the name of a mark class. The others are:

right_to_left
ignore_bases
ignore_ligatures
ignore_marks

A feature-script-lang tuple is a tuple with one entry for each feature (there may be no entries if there are no features). Each entry is itself a two element tuple, the first entry is a string containing a 4 letter feature tag, and the second entry is another tuple (potentially empty) with an entry for each script for which the feature is active. Each entry here is itself a two element tuple. The first element is a 4 letter script tag and the second is a tuple of languages. Each entry in the language tuple is a four letter language. Example: (("liga",(("latn",("dflt")),)),)

The optional final argument allows you to specify the ordering of the lookup. If not specified the lookup will be come the first lookup in its table.

font.addLookupSubtable(lookup_name, new_subtable_name[, after_subtable_name])¶

Creates a new subtable within the specified lookup. The lookup name should be a string specifying an existing lookup. The subtable name should also be a string and should not match any currently existing subtable in the lookup. The optional final argument allows you to specify the ordering within the lookup. If not specified this subtable will be first in the lookup.

If you want to create a subtable in a contextual lookup, then use font.addContextualSubtable(). If you want to create a kerning class subtable, then use font.addKerningClass().

font.addContextualSubtable(lookup_name, new_subtable_name, type, rule[, afterSubtable=] [, bclasses=] [, mclasses=] [, fclasses=] [, bclassnames=] [, mclassnames=] [, fclassnames=])¶

Creates a new subtable within the specified contextual lookup (contextual, contextual chaining, or reverse contextual chaining). The lookup name should be a string specifying an existing lookup. The subtable name should also be a string and should not match any currently existing subtable in the lookup.

The type should be one of the strings “glyph”, “class”, “coverage” or “reversecoverage”. The rule should be a string specifying a string to match and a set of lookups to apply once the match has been made. (See below for more details).

The remaining arguments are optional, keyword arguments.

afterSubtable=, if present this should be followed by a string, the name of a subtable after which this one is to be placed in the lookup. If not specified this subtable will be first in the lookup.
bclasses=, fclasses=, mclasses= these three arguments specify sets of glyph classes for when type="class". They should be a tuple of thingies where each thingy is either a string containing a list of space separated glyph names, or another tuple containing a set of strings, each a glyph name. Note that the first class is magic and should usually be left as a null string.
bclassnames=, fclassnames=, mclassnames= These provide names for the glyph classes described above. These names are optional, but can be convenient. These are tuples of strings. There should be the same number of entries in bclassnames as there are in bclasses.

When type="glyph"

The rule should look something like:

glyph-name1 glyph-name2 | glyph-name3 @<lookup-name> | glyph-name4

The | s divide between backtrack, match and lookahead sections. So this example would match it the current glyph were named glyph-name3 and it were preceded by glyph-name2 and that by glyph-name1 and followed by glyph-name4. If the match were successful then the lookup named lookup-name would be applied. The @<> are litteral characters and should be present in the rule.

If the invoked lookup is a ligature lookup then it should be invoked after the first glyph that forms the lookup (rather than the last) and all glyphs that might make up the lookup should be in the match section. So…

e | f @<ff-lig> f l | o

would only apply the ff-lig lookup if the ffl were preceeded by e and followed by o.

When type="class"

The rule should look something like:

class-name1 class-name2 | class-name3 @<lookup-name> | class-name4

Very similar to the case of glyphs, except that instead of glyph names we have class names here. It is possible to have different sets of class names in the three different sections (backtrack, match and lookahead). If you don’t specify any class names then you must use numbers instead, each number refering to the class at that position in the tuple (the first class will be class 0, the second class 1, and so on).

When type="coverage"

The rule should look something like:

[g1 g2] [g3 g4] | [g5 g6 g7] @<lookup-name> | [g8 g9]

Each entry within brackets, [], represents a coverage table and should be a list of glyph names. The brackets are specified literally.

When type="reversecoverage"

The rule should look something like:

[g1 g2] [g3 g4] | [g5 g6 g7] => [rg1 rg2 rg3] | [g8 g9]

Very similar to normal coverage tables except that instead of specifying a lookup there are replacement glyphs inline. There must be the same number of replacement glyphs (rg1, rg2,rg3) as match glyphs (g5, g6, g7) and there may be only one coverage table in the match section.

Warning

This format has some limitations, if there are multiple lookups they will be applied in textual order (First lookup in the string is the first one applied). This limitation is also present in Adobe’s feature files so I hope it shan’t be a severe limitation.

font.addSmallCaps(scheight=None, capheight=None, lcstem=None, ucstem=None, symbols=None, letter_extension=None, symbol_extension=None, stem_height_factor=None)¶

This function uses keyword parameters. None are required, if omitted a default value will be used (generally found by analyzing the font).

For each selected letter, this function will create a corresponding small caps glyph. If you set the symbol keyword to True it will also create small caps variants of digits and symbols.

The outlines of the new glyph will be based on the outlines of the upper-case variant of the letter. These will then be scaled so that a glyph which was capheight high will now be scheight high, and so that stems which were ucstem wide will be lcstem wide. Normally the ratio of stem heights is the same as the ratio of stem widths, but you may change that with stem_height_factor.

When it creates a new glyph it will name that glyph by appending “.sc” to the original lower case letter name (so “a” would become “a.sc”) you may change the extension used with letter_extension. Similary symbols and digits will use the extension “taboldstyle”, but you may change that with symbol_extension.

When it creates a glyph it also creates two lookups one bound to the feature “c2sc” and the other to “smcp”. A mapping from the lower case letter to the small caps letter will be provided under “smcp”, while a mapping from the upper case to the small caps under “c2sc”. Symbols will have both lookup maps attached to the original glyph.

font.alterKerningClass(subtable_name, first_classes, second_classes, offsets)¶: Changes the kerning class in the named subtable. The classes arguments are tuples of tuples of glyph names (each sub-tuble of glyph names is a kerning class). The offsets argument is a tuple of kerning offsets. There must be as many entries as len(first-class)*len(second-class). The optional after argument is used to specify the order of the subtable within the lookup.

font.autoKern(subtable_name, separation[, minKern=, onlyCloser=, touch=])¶

font.autoKern(subtable_name, separation, glyph_list1, glyph_list2[, minKern=, onlyCloser=, touch=])

The named subtable must be a kerning pair subtable that already exists.

This command will automatically generate kerning pairs for the named subtable. If no glyph lists are specified it will look at all pairs of the glyphs in the selection; if glyph lists are specified then it will look at all pairs that can be made with one glyph from the first list and the second from the second list.

It will attempt to guess a good kerning value between the two glyphs – a value which will make the optical separation between the two appear to be separation em-units. If minkern is specified then and the (absolute value of the) kerning correction is less than this number then no kerning pair will be generated. If onlyCloser is set true then only negative kerning offsets will be generated (only thing which move two glyphs closer together). If touch is set to 1 then the kerning offset will not be based on optical distance but on the closest approach between two the two glyphs.

font.appendSFNTName(language, strid, string)¶: Adds a new (or replaces an old) string in the sfnt ‘name’ table. Language may be either the English name of the language/locale as a string, or the number representing that language in MicroSoft’s specification. Strid may be one of the (english) string names (Copyright, Family, SubFamily, etc.) or the numeric value of that item. The string itself is in UTF-8.

font.buildOrReplaceAALTFeatures()¶: Removes any existing AALT features (and any lookups solely controled by such features) and creates new ones containing all possible single and alternate substutions available for each glyph.

font.cidConvertByCMap(cmap_filename)¶

Converts a normal font into a CID-keyed font with one subfont using

the CMAP to determine the mapping.

font.cidConvertTo(registry, ordering, supplement)¶: Converts a normal font into a CID-keyed font with one subfont.

font.cidFlatten()¶: Converts a CID font into a normal font (glyphs will be in CID order).

font.cidFlattenByCMap(cmap_filename)¶: Converts a CID font into a normal font using the encoding specified in the CMAP file.

font.cidInsertBlankSubFont()¶: Adds a new (blank) sub-font into a cid-keyed font and changes the current sub-font to be it.

font.cidRemoveSubFont()¶: Removes the current subfont from a cid-keyed font.

font.close()¶

Frees memory for the current font.

Warning: Any python references to it will become invalid.

font.compareFonts(other_font, filename, flags_tuple)¶

This will compare the current font with the font in other-font (which must already have been opened). It will write the results to the filename, you may use “-” to send the output to stdout. The flags argument is a tuple of strings and controls what will be compared.

outlines: compare outlines

outlines-exactly: compare outlines exactly (otherwise allow slight errors and the unlinking of references)

warn-outlines-mismatch: warn if the outlines don’t exactly match (but are pretty close)

hints: compare hints

warn-refs-unlink: warn if references need to be unlinked before a match is found

strikes: compare bitmap strikes

fontnames: compare font names

gpos: compare glyph positioning

gsub: compare glyph substitutions

add-outlines: for any glyphs whose outlines differ, add the outlines of the glyph in the second font to the background of the glyph in the first

create-glyphs: if a glyph exists in the second font but not the first, create that glyph in the first and add the outlines from the second into the backgroun layer

font.createChar(uni[, name])¶

Create (and return) a character at the specified unicode codepoint in this font and optionally name it. If you wish to create a glyph with no unicode codepoint, set the first argument to -1 and specify a name.

If there is already a character at that (positive) codepoint then it is returned. If the optional name parameter is included and differs from its current name then the character is also renamed.

font.createInterpolatedGlyph(glyph1, glyph2, amount)¶: Create (and return) a glyph with the same unicode code point as glyph1. The glyph may not already exist. The contents of the glyph will be formed by interpolating between glyph1 and glyph2. If amount==0 the result will look like glyph1, or 1 then like glyph2.

font.createMappedChar(enc)¶
font.createMappedChar(name): Create (and return) a character at the specified encoding in this font. If there is already a character there, return it

font.find(contour[, error_bound, search_flags])¶

Searches the font for all glyphs containing the contour (or layer) and returns an iterator which returns those glyphs.

error-bound: defaults to 0.01.

search_flags: tuple of the strings: reverse, flips, rotate, scale.

When found, the glyph.temporary is set to a dict of:

{
  "findMatchedRefs": matched_refs_bit_map,
  "findMatchedContours": matched_contours_bit_map,
  "findMatchedContoursStart": matched_contours_start_bit_map,
}

font.findEncodingSlot(uni)¶

font.findEncodingSlot(name)

Tests whether a glyph with this codepoint or name is in the font’s encoding and returns the encoding slot. If the glyph is not present it returns -1.

(If a glyph with that name/unicode is in the font, but is not in the encoding, then an value beyond the end of the encoding will be returned).

font.glyphs([type])¶: Returns an iterator which will return the glyphs in the font. By default they will be returned in “GID” order, but if type is specified as “encoding” they will be returned in encoding order. If returned in encoding order it is possible that a glyph will be returned more than once if there are multiple encoding slots which reference it.

font.generate(filename[, bitmap_type=, flags=, bitmap_resolution=, subfont_directory=, namelist=, layer=])¶

Generates a font. The type is determined by the font’s extension. The bitmap type (if specified) is also an extension. If layer is specified, then the splines and references in that layer will be used instead of the foreground layer.

Flags is a tuple containing some of

afm: output an afm file

pfm: output a pfm file

tfm: output a tfm file

ofm: output a ofm file

composites-in-afm: Store composite info in the afm file

glyph-map-file: Output a glyph map file giving the mapping between output gid and glyphnames

short-post: Do not include glyphnames in a ttf/otf file

apple: output apple advanced typography tables

opentype: output opentype tables

old-kern: output an old style ‘kern’ with opentype tables

winkern: only add kern pairs for mapped glyphs (required for kerning in some/all versions of Windows)

dummy-dsig: output an empty DSIG table so MS will mark a font with .ttf extension as an OpenType font.

TeX-table: Include a ‘TeX ‘ table in an ttf/otf file

round: Round PS coordinates to integers

no-hints: Do not include PS hints

no-flex: Do not include PS flex hints

omit-instructions: Do not include TrueType instructions

PfEd-comments: Include font and glyph comments in the ‘PfEd’ table

PfEd-colors: Include glyph colors in the ‘PfEd’ table

PfEd-lookups: Include lookup names in the ‘PfEd’ table

PfEd-guidelines: Include guideline locations in the ‘PfEd’ table

PfEd-background: Include background (and spiro) layers in the ‘PfEd’ table

symbol: Generate an sfnt with a Symbol cmap entry rather than a Unicode entry.

fontforge¶

Module attributes¶

Module functions¶

User Interface Module Functions¶

Point¶

Contour¶

Layer¶

Glyph Pen¶

Glyph¶

Selection¶

Private¶

Math¶

Font¶

FontForge

Navigation

Related Topics