Skip to content

Kerning pairs from text

Retained catalogue

This page comes from the retained catalogue. Its launch instructions and native results describe the earlier tree, archived externally with historical prefix rescued/part2, rather than the current launcher. The documentation build does not rerun these examples.

Purpose. Paste a pair list, from a proofing session or another tool, and apply it to the current master:

# left right value
A V -80
@O_left T -40

@name is a FontLab kerning class. Values are rounded to whole units (FontLab stores whole units; halves round away from zero, as FontLab does) and the stored values are reported. Unreadable lines (including nan, infinities and values beyond ±32767, the 16-bit range of OpenType kerning) are listed, not guessed at, before anything is applied. glyph_names limits the run to pairs that involve those glyphs.

run(text=open("pairs.txt").read())
run(text=pairs, glyph_names=["A"], master="Bold")

FontLab API. fontpr.call("set_kerning_pair", ...), FontLab's MCP kerning tool from FontLab's own Python.

Native check (native.py, owned font): three pairs, read back with list_kerning_pairs, carry the values −80, −76 (from −75.5) and −60; the bad line is reported.

Provenance. The kind of task done by "Import Kerning" and "Add new pair" in the TypeRig GUI Kerning manager (typerig-lib/Scripts/TypeRig GUI/Manager/Kerning.py, MPL-2.0). Written from scratch: only the manager's menu labels were consulted, never its implementation.

Licence. Apache-2.0, as the repository.