bug 1254884 allows boolean values to be dumped in config.data and set in config.status, but it's further down the road that there still are road blocks.

This allows to use True and False as values given to set_config/set_define in moz.configure files, while postponing having to deal with the long tail of things depending on the values from substs and defines. Ideally, everything would handle the bools just fine, but there are too many things involved to deal with this right now: scripts using buildconfig.{substs,defines}, scripts using ConfigEnvironment (e.g. process_define_files.py), ConfigEnvironment itself, etc. Review commit: https://reviewboard.mozilla.org/r/41659/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/41659/

Review commit: https://reviewboard.mozilla.org/r/41661/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/41661/

Review commit: https://reviewboard.mozilla.org/r/41663/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/41663/

Review commit: https://reviewboard.mozilla.org/r/41665/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/41665/

Review commit: https://reviewboard.mozilla.org/r/41667/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/41667/

Comment on attachment 8733231 [details] MozReview Request: Bug 1258618 - Serialize substs/configs and defines bools as '1' or '' in config.status Review request updated; see interdiff: https://reviewboard.mozilla.org/r/41659/diff/1-2/

Comment on attachment 8733232 [details] MozReview Request: Bug 1258618 - Allow to use bools as values given to add_old_configure_assignment Review request updated; see interdiff: https://reviewboard.mozilla.org/r/41661/diff/1-2/

Comment on attachment 8733233 [details] MozReview Request: Bug 1258618 - Use True instead of '1' for set_config Review request updated; see interdiff: https://reviewboard.mozilla.org/r/41663/diff/1-2/

Comment on attachment 8733234 [details] MozReview Request: Bug 1258618 - Use True instead of '1' for set_define Review request updated; see interdiff: https://reviewboard.mozilla.org/r/41665/diff/1-2/

Comment on attachment 8733235 [details] MozReview Request: Bug 1258618 - Use True instead of '1' for add_old_configure_assignment Review request updated; see interdiff: https://reviewboard.mozilla.org/r/41667/diff/1-2/

Comment on attachment 8733231 [details] MozReview Request: Bug 1258618 - Serialize substs/configs and defines bools as '1' or '' in config.status https://reviewboard.mozilla.org/r/41659/#review38277 nits. ::: configure.py:35 (Diff revision 2) > def config_status(config): > # Sanitize config data to feed config.status > + # Ideally, all the backend and frontend code would handle the booleans, but > + # there are so many things involved, that it's easier to keep config.status > + # untouched for now. > + def sanitized_bool(v): This is a little misleading, since the input doesn't have to be a `bool`. Perhaps `sanitize_bools`? I guess this doesn't escape this scope, so not a big deal. ::: configure.py:40 (Diff revision 2) > + def sanitized_bool(v): > + if v is True: > + return '1' > + if v is False: > + return '' > + return v Is this a chance to assert something else about our data? Is it always a string? Always '1' or ''?

Comment on attachment 8733233 [details] MozReview Request: Bug 1258618 - Use True instead of '1' for set_config https://reviewboard.mozilla.org/r/41663/#review38281

Comment on attachment 8733234 [details] MozReview Request: Bug 1258618 - Use True instead of '1' for set_define https://reviewboard.mozilla.org/r/41665/#review38283

Comment on attachment 8733235 [details] MozReview Request: Bug 1258618 - Use True instead of '1' for add_old_configure_assignment https://reviewboard.mozilla.org/r/41667/#review38285

Comment on attachment 8733232 [details] MozReview Request: Bug 1258618 - Allow to use bools as values given to add_old_configure_assignment https://reviewboard.mozilla.org/r/41661/#review38287

https://reviewboard.mozilla.org/r/41659/#review38277 > Is this a chance to assert something else about our data? Is it always a string? Always '1' or ''? Mmmmm I'd rather do that in a followup in set_config itself, that would give a better traceback pointing to where the unwanted typed value is being set.

https://hg.mozilla.org/integration/mozilla-inbound/rev/aa4bcd9eadcb https://hg.mozilla.org/integration/mozilla-inbound/rev/fa9a23bf9e17 https://hg.mozilla.org/integration/mozilla-inbound/rev/0f1ddce63248 https://hg.mozilla.org/integration/mozilla-inbound/rev/34d422497ef9 https://hg.mozilla.org/integration/mozilla-inbound/rev/43d66eea833b

https://hg.mozilla.org/mozilla-central/rev/aa4bcd9eadcb https://hg.mozilla.org/mozilla-central/rev/fa9a23bf9e17 https://hg.mozilla.org/mozilla-central/rev/0f1ddce63248 https://hg.mozilla.org/mozilla-central/rev/34d422497ef9 https://hg.mozilla.org/mozilla-central/rev/43d66eea833b