googlefonts / roboto-2

The Roboto family of fonts
Apache License 2.0
3.85k stars 309 forks source link

subset webfonts when running make #238

Closed m4rc1e closed 7 years ago

m4rc1e commented 7 years ago

When running make web, the Fontmake file didn't apply subset_for_web.py. The webfonts hosted at https://fonts.google.com/specimen/Roboto have been subsetted.

Unless anyone can point me to a set of webfonts which have the full character set, I think make should apply this.

m4rc1e commented 7 years ago

Ok, once the files have been subsetted, running run_web_tests.py fails on 3 tests

FAIL: test_c2sc_coverage (__main__.TestFeatures)
Tests that c2sc is supported for a required set.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/marc/Documents/googlefonts/manual_font_cleaning/roboto/env/lib/python2.7/site-packages/nototools/unittests/font_tests.py", line 461, in test_c2sc_coverage
    self.run_sub_coverage_test('c2sc', self.c2sc_reqs_path)
  File "/Users/marc/Documents/googlefonts/manual_font_cleaning/roboto/env/lib/python2.7/site-packages/nototools/unittests/font_tests.py", line 451, in run_sub_coverage_test
    (feature, u''.join(chars_with_no_sub).encode('UTF-8'))))
AssertionError: c2sc feature is not applied correctly to '&ABCDEFGHIJKLMNOPQRSTUVWXYZÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞĀĂĄĆĈĊČĎĐĒĔĖĘĚĜĞĠĢĤĦĨĪĬĮIJĴĶĹĻĽĿŁŃŅŇŊŌŎŐŒŔŖŘŚŜŞŠŢŤŦŨŪŬŮŰŲŴŶŸŹŻŽǺǼǾȘΆΈΉΊΌΎΏΑΒΓΔΕΖΗΘΙΚΛΜΝΞΟΠΡΣΤΥΦΧΨΩΪΫЁЂЃЄЅІЇЈЉЊЋЌЎЏАБВГДЕЖЗИЙКЛМНОПРСТУФХЦЧШЩЪЫЬЭЮЯҐẀẂẄẞỲ'

======================================================================
FAIL: test_smcp_coverage (__main__.TestFeatures)
Tests that smcp is supported for a required set.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/marc/Documents/googlefonts/manual_font_cleaning/roboto/env/lib/python2.7/site-packages/nototools/unittests/font_tests.py", line 456, in test_smcp_coverage
    self.run_sub_coverage_test('smcp', self.smcp_reqs_path)
  File "/Users/marc/Documents/googlefonts/manual_font_cleaning/roboto/env/lib/python2.7/site-packages/nototools/unittests/font_tests.py", line 451, in run_sub_coverage_test
    (feature, u''.join(chars_with_no_sub).encode('UTF-8'))))
AssertionError: smcp feature is not applied correctly to '&abcdefghijklmnopqrstuvwxyzßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿāăąćĉċčďđēĕėęěĝğġģĥħĩīĭįıijĵķĺļľŀłńņňŋōŏőœŕŗřśŝşšţťŧũūŭůűųŵŷźżžǻǽǿșάέήίαβγδεζηθικλμνξοπρςστυφχψωϊϋόύώабвгдежзийклмнопрстуфхцчшщъыьэюяёђѓєѕіїјљњћќўџґẁẃẅỳ'

======================================================================
FAIL: test_unic_coverage (__main__.TestFeatures)
Tests that unic is supported for a required set.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/marc/Documents/googlefonts/manual_font_cleaning/roboto/env/lib/python2.7/site-packages/nototools/unittests/font_tests.py", line 466, in test_unic_coverage
    self.run_sub_coverage_test('unic', self.unic_reqs_path)
  File "/Users/marc/Documents/googlefonts/manual_font_cleaning/roboto/env/lib/python2.7/site-packages/nototools/unittests/font_tests.py", line 451, in run_sub_coverage_test
    (feature, u''.join(chars_with_no_sub).encode('UTF-8'))))
AssertionError: unic feature is not applied correctly to 'AEMNUabcdefghijklmnopqrstuvwxyzÀÁÂÃÄÅÈÉÊËÑÙÚÜàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿĀāĂ㥹ćĉċčďđĒēĔĕĖėĘęĚěĝğġģĥħĩīĭįijĵķĺļľŀŃńŅņŇňŋōŏőœŕŗřśŝşšţťŧŨũŪūŬŭŮůŰűŲųŵŷźżžǺǻǽǿșẁẃẅỳ'

.

davelab6 commented 7 years ago

Unless anyone can point me to a set of webfonts which have the full character set, I think make should apply this.

I agree; the Google Fonts API doesn't want to increase filesize and thus latency - so make web should always subset

m4rc1e commented 7 years ago

Webfont tests have been updated.

davelab6 commented 7 years ago

Great! Per @roozbehp request, I have assigned @dougfelt as a reviewer, who I hope can use the Github review system to confirm this is good to merge :)

dougfelt commented 7 years ago

LGTM

davelab6 commented 7 years ago

Thank you Doug!!

On Mar 19, 2017 4:58 PM, "dougfelt" notifications@github.com wrote:

LGTM

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/google/roboto/pull/238#issuecomment-287647010, or mute the thread https://github.com/notifications/unsubscribe-auth/AAP9y0Lg2xAi58OgfAsPr7eM2juao_SKks5rnZcGgaJpZM4MeBbz .