chython/scripts/gen_font_metrics.py at 3.3 · chython/chython

Repository navigation

History

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

#!/usr/bin/env python3

# -*- coding: utf-8 -*-

# This file is part of chython.

# chython is free software; you can redistribute it and/or modify

# it under the terms of the GNU Lesser General Public License as published by

# the Free Software Foundation; either version 3 of the License, or

# (at your option) any later version.

# This program is distributed in the hope that it will be useful,

# but WITHOUT ANY WARRANTY; without even the implied warranty of

# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the

# GNU Lesser General Public License for more details.

# You should have received a copy of the GNU Lesser General Public License

# along with this program; if not, see <https://www.gnu.org/licenses/>.

"""Derive `chython/depict/metrics/*.tsv` from the AFM files matplotlib ships.

WHY A SHIPPED TABLE. The depictor must know a label's width before it places anything -- to centre a

symbol, to trim a bond at the label's box, to size a page. Asking the renderer at draw time is not an

option: SVG has no measurement API a library can call, and the three backends would each measure

differently. So the metrics are DATA, generated once, checked in, and identical for all three.

WHY THESE TWO FAMILIES. Helvetica and Times-Roman are PDF base-14: the PDF backend names them and

embeds no font file, PostScript has them as standard fonts, and SVG resolves them through

`font-family`. One table is exact in all three outputs, which no shipped TTF would be.

WHY MATPLOTLIB IS NOT A DEPENDENCY. It is the source of the AFMs, so it is a dependency of THIS

SCRIPT, run by a developer, and of one drift test that skips when it is absent -- the `needs_inchi`

shape. Nothing at runtime imports it.

Adobe's AFM licence permits redistribution provided the copyright notices are retained, and each

generated TSV carries the notice in its header for that reason.

"""

from argparse import ArgumentParser

from pathlib import Path

# Three parts, in this order, because Adobe's readme.txt requires all three: the copyright notice

# retained; the permission paragraph UNMODIFIED (it says so itself, so it is a licence condition and

# not prose we may tighten); and a prominent note that this file is a modification -- a derived

# numeric extraction -- rather than an Adobe AFM. Do not reflow or reword ADOBE_PARAGRAPH.

ADOBE_PARAGRAPH = (

'This file and the 14 PostScript(R) AFM files it accompanies may be used, copied, and '

'distributed for any purpose and without charge, with or without modification, provided that '

'all copyright notices are retained; that the AFM files are not distributed without this file; '

'that all modifications to this file or any of the AFM files are prominently noted in the '

'modified file(s); and that this paragraph is not modified. Adobe Systems has no responsibility '

'or obligation to support the use of the AFM files.')

ADOBE_DERIVATION = ('MODIFICATION NOTICE: this file is not an Adobe AFM. It is a derived numeric '

'table extracted from Helvetica.afm / Times-Roman.afm as distributed with '

'matplotlib (mpl-data/fonts/pdfcorefonts), generated by '

'scripts/gen_font_metrics.py. Only the glyphs chython labels need are kept, '

'and only the advance width and bounding box of each.')

# family -> AFM basename. The two PDF base-14 faces the depictor names; adding a third means adding it

# here AND in `metrics.FAMILIES` AND in `PDF_BASE_FONT`, and the test below checks the three agree.

FAMILIES = {'helvetica': 'Helvetica.afm', 'times': 'Times-Roman.afm'}

HEADER = ('char', 'name', 'wx', 'llx', 'lly', 'urx', 'ury')

def afm_directory() -> Path:

from matplotlib import get_data_path

return Path(get_data_path(), 'fonts', 'pdfcorefonts')

def read_afm(path: Path) -> list:

"""Parse an AFM's CharMetrics block.

Lines look like `C 67 ; WX 722 ; N C ; B 44 -19 681 737 ;` -- the code, the advance width, the

glyph name and the bounding box, all in 1/1000 em. `C -1` means the glyph is in the font but not

in the default encoding, and those are kept: `minus` is `C -1` in Helvetica and a charge sign needs

it. The key is the glyph NAME, and encoding is this script's problem, not the runtime's.

"""

rows = []

inside = False

for raw in path.read_text(encoding='latin-1').splitlines():

line = raw.strip()

if line.startswith('StartCharMetrics'):

inside = True

continue

if line.startswith('EndCharMetrics'):

break

if not inside or not line:

continue

fields = {}

for part in line.split(';'):

part = part.strip()

if not part:

continue

key, _, value = part.partition(' ')

fields[key] = value.strip()

if 'N' not in fields or 'WX' not in fields:

continue

box = fields.get('B', '0 0 0 0').split()

rows.append((fields['N'], int(float(fields['WX'])), *(int(float(v)) for v in box)))

rows.sort()

return rows

def compile_tables(destination: Path) -> None:

for family, basename in FAMILIES.items():

rows = read_afm(afm_directory() / basename)

lines = [f'# {ADOBE_COPYRIGHT}', '#']

# The paragraph is wrapped for line length but never reworded: `_wrap_comment` splits on

# spaces only, so joining the emitted lines back reproduces it character for character.

# A test below asserts exactly that, because "this paragraph is not modified" is a licence

# condition and a stray reflow that drops a word would breach it silently.

for line in _wrap_comment(ADOBE_PARAGRAPH):

lines.append(f'# {line}')

lines.append('#')

for line in _wrap_comment(ADOBE_DERIVATION):

lines.append(f'# {line}')

lines.append('#')

lines.append(f'# Generated by scripts/gen_font_metrics.py from {basename}. Do not hand-edit.')

lines.append('# Units are 1/1000 em. `char` is the character the glyph encodes, or - when it '

'encodes none.')

lines.append('\t'.join(HEADER))

for name, wx, llx, lly, urx, ury in rows:

char = _character_for(name)

lines.append('\t'.join((char, name, str(wx), str(llx), str(lly), str(urx), str(ury))))

(destination / f'{family}.tsv').write_text('\n'.join(lines) + '\n', encoding='utf-8')

def _wrap_comment(text: str, width: int = 96) -> list:

"""Wrap on spaces only, so ' '.join(result) == text exactly.

Not `textwrap.fill`: that collapses runs of whitespace and can break on hyphens, either of which

would alter Adobe's paragraph. The licence says the paragraph is not modified, so the wrap has

to be information-preserving, and `test_the_adobe_paragraph_survives_the_round_trip` proves it is.

"""

out = []

current = ''

for word in text.split(' '):

if not current:

current = word

elif len(current) + 1 + len(word) <= width:

current += ' ' + word

else:

out.append(current)

current = word

if current:

out.append(current)

return out

def _character_for(name: str) -> str:

"""The character an AFM glyph name encodes, for the names the depictor can actually emit.

Not a general Adobe-glyph-list implementation. What a chemical label contains is letters, digits,

the four charge and bond signs, brackets, a dot and a comma -- plus the typographic minus, which is

what a charge should be set in rather than a hyphen. Anything else keeps `-` in the `char` column

and is reachable by NAME only, which is how `text_box` handles a glyph a caller asked for by name.

"""

if len(name) == 1:

return name

return _NAMED_GLYPHS.get(name, '-')

_NAMED_GLYPHS = {

'space': ' ', 'exclam': '!', 'quotedbl': '"', 'numbersign': '#', 'dollar': '$', 'percent': '%',

'ampersand': '&', 'quoteright': '’', 'parenleft': '(', 'parenright': ')', 'asterisk': '*',

'plus': '+', 'comma': ',', 'hyphen': '-', 'period': '.', 'slash': '/',

'zero': '0', 'one': '1', 'two': '2', 'three': '3', 'four': '4',

'five': '5', 'six': '6', 'seven': '7', 'eight': '8', 'nine': '9',

'colon': ':', 'semicolon': ';', 'less': '<', 'equal': '=', 'greater': '>', 'question': '?',

'at': '@', 'bracketleft': '[', 'backslash': '\\', 'bracketright': ']', 'asciicircum': '^',

'underscore': '_', 'quoteleft': '‘', 'braceleft': '{', 'bar': '|', 'braceright': '}',

'asciitilde': '~', 'endash': '–', 'emdash': '—', 'minus': '−',

'periodcentered': '·', 'bullet': '•', 'degree': '°', 'plusminus': '±',

'multiply': '×', 'dagger': '†',

}

def main():

parser = ArgumentParser(description=__doc__)

parser.add_argument('verb', choices=['compile'])

parser.add_argument('-o', '--output', type=Path,

default=Path(__file__).resolve().parent.parent / 'chython' / 'depict' / 'metrics')

args = parser.parse_args()

compile_tables(args.output)

if __name__ == '__main__':

main()

FilesExpand file tree

gen_font_metrics.py

Latest commit

History

gen_font_metrics.py

File metadata and controls