adopt-project.py, shape_materialize.py and the assembly-root template: after #1121 and #1105, one GNU_MAKE_READS and one lookup, their NEXT NOTE and collision line as true as #1207's, and the NITs 1, 3b, 3c, 4 and 5 of #1105's second review #1240

(lane openRepoShape-5. These are #1132's items that #1207 leaves, because they live only in lane 4's #1105 (#1048) and #1121 (#1047). #1132 is shut when #1207 lands, so this issue tracks them: #1207's review, issuecomment-6374516382, SHOULD-2.)

Problem

#1207 does the parts of #1132 that hold on main. Lane 4 lands #1105 and #1121 as reviewed, with no fix round, #1121 first, and hands their items to lane 5 (#473 issuecomment-6350568104: "Your list above is the record of what moves to lane 5"). Once both land, main has the following. Letters name them below. "Item N" is #1132's own numbering, and "RV1105b" is #1105's second review (issuecomment-6321479057).

• A. Two tuples and three lookups (item 4). shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105 adds a second GNU_MAKE_READS to shape_materialize.py, and a test that holds the two equal. shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's _predict_collisions and adopt-project.py: execute's NEXT prints python3 scripts/bootstrap.py where GNU make at the root reads a kept GNUmakefile or makefile, and its NOTE names that file and what it takes for make bootstrap to run there (#1047) #1121's _next_steps each write their own next(... for name in GNU_MAKE_READS ...), beside _makefile_follow_up's. adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207 adds _make_reads_first, which all three can share.
• B. shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's collision line calls a kept DIRECTORY a file (item 2). At a root that keeps Makefile and GNUmakefile/, the Makefile's collision follow-up says "GNU make at the root reads the source's GNUmakefile, ... so the shape's make targets ... reach make there only once GNUmakefile has them, or includes a Makefile that does". GNU make 4.3 stops at that directory for every target: make: *** GNUmakefile: Is a directory. Stop., exit 2. adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207 fixed the same claim in the CONTRACTS_DIR line and in NEXT.
• C. shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's collision line and adopt-project.py: execute's NEXT prints python3 scripts/bootstrap.py where GNU make at the root reads a kept GNUmakefile or makefile, and its NOTE names that file and what it takes for make bootstrap to run there (#1047) #1121's NEXT NOTE name python3 alone (item 3). Both say "python3 scripts/bootstrap.py is what the shape's make bootstrap runs". The template's scripts/bootstrap.py says "(python on Windows)", and adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207's BOOTSTRAP_COMMAND says it everywhere else.
• D. adopt-project.py: execute's NEXT prints python3 scripts/bootstrap.py where GNU make at the root reads a kept GNUmakefile or makefile, and its NOTE names that file and what it takes for make bootstrap to run there (#1047) #1121's NOTE beside a kept Makefile leaves out the merge. For a kept FILE GNUmakefile beside a kept Makefile, it says the targets "run there only once GNUmakefile has them, or includes a Makefile that does". Where GNUmakefile includes Makefile, the root's Makefile is still the source's own until the follow-up merges the two. adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207's review found the same gap in adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207's directory NOTE (its NIT-3), and adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207 now names the merge there.
• E. RV1105b's NIT-1, the link case. Four comments say the merge alone brings the targets "only where that file includes Makefile". That is untrue where the file is a symlink to Makefile. RV1105b measured Makefile + GNUmakefile -> Makefile: there, the merge alone made make bootstrap run the shape's recipe.
• F. RV1105b's NIT-3b: "the words execute's NEXT uses (adopt-project.py: execute's NEXT prints make bootstrap where GNU make at the root reads a kept makefile or GNUmakefile, which has no bootstrap rule #1047)". This is in collision_follow_up's docstring, and the test module says "The words are those of execute's NEXT". It is true only once adopt-project.py: execute's NEXT prints python3 scripts/bootstrap.py where GNU make at the root reads a kept GNUmakefile or makefile, and its NOTE names that file and what it takes for make bootstrap to run there (#1047) #1121 lands. adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207 changes NEXT's words, so check it again once all three have landed.
• G. RV1105b's NIT-3c: Materialized.follow_ups(). shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's docstring says it "would say this of them". It passes no reads_first, so it would not, and nothing calls it.
• H. RV1105b's NIT-4: no row ties shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's template clause to make. Its mutant K9 put both the template and ADOPTED_SENTENCES back to "and never the shape's Makefile", and all 118 rows passed.
• I. RV1105b's NIT-5, cosmetic. shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's template paragraph leaves a short line, "does. A SCAFFOLDED". Its test docstring breaks mid-sentence: "The tail says what the".
• J. _make_reads_first_clause's has argument goes live. On adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207 it is reached only with does, so adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207's mutant M18 is equivalent there. It goes live once adopt-project.py: execute's NEXT prints python3 scripts/bootstrap.py where GNU make at the root reads a kept GNUmakefile or makefile, and its NOTE names that file and what it takes for make bootstrap to run there (#1047) #1121's NOTE words its file case through that helper.

Measured

This was measured offline at #1105 b28f40de0, #1121 1c5173758, main e04826db4 and #1207 75c90b6db, with git grep and the lane's text probe. The probe loads each head's adopt-project.py and shape_materialize.py, and renders the three texts for a set of roots. Its extracted trees match each head byte for byte.

At #1207 75c90b6db, _make_reads_first_clause(first, reach, "has them") for GNUmakefile gives #1121's own reads clause byte for byte. So #1121's file case can go through it with no change to its words.

Suggested fix

Do this after #1121, #1105 and #1207 have landed. #1207 lands after both and resolves its conflict with each. Where that resolution already does an entry, its round says which one, and this issue keeps the rest.

• A: adopt-project.py imports GNU_MAKE_READS from shape_materialize, as it already imports BOOTSTRAP. The equality test becomes an is test. _make_reads_first is the one lookup for _makefile_follow_up, _predict_collisions' reads_first and _next_steps, and _make_reads_first_clause moves to shape_materialize.py.
• B: collision_follow_up takes reads_first with its /, and says the directory stops GNU make, in _make_reads_first_clause's words.
• C: BOOTSTRAP_COMMAND goes into adopt-project.py: execute's NEXT prints python3 scripts/bootstrap.py where GNU make at the root reads a kept GNUmakefile or makefile, and its NOTE names that file and what it takes for make bootstrap to run there (#1047) #1121's NOTE and shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's meanwhile. The command lines a person pastes stay python3 scripts/bootstrap.py.
• D: beside a kept Makefile, adopt-project.py: execute's NEXT prints python3 scripts/bootstrap.py where GNU make at the root reads a kept GNUmakefile or makefile, and its NOTE names that file and what it takes for make bootstrap to run there (#1047) #1121's NOTE says the targets run there once the follow-up merges the two, and only once the file has them, as adopt-project.py, shape_materialize.py and the assembly-root template: a kept GNUmakefile/ is a directory GNU make stops at, python3 scripts/bootstrap.py names python on Windows, and the template says what to run where make reads another file #1207's directory NOTE now does.
• E: "only where that file includes Makefile or is a link to it".
• F: reword it to what NEXT says once all three PRs have landed.
• G: delete Materialized.follow_ups(), or have it pass reads_first, and make the docstring match.
• H: in the clone rows where make reads the shape's Makefile through include Makefile, assert that the pinned AGENTS-shape.md does not say "never the shape's Makefile".
• I: rewrap both. A template byte change makes every project re-pin, so make the template one together with the next template change.

The local branch fix/1132-template-bootstrap-stacked (9eb5cce0b) is a model of the end state on #1105 + #1121 + main. It was never pushed and never tested, and its words differ from #1207's in places.

Verify

• git grep -n 'GNU_MAKE_READS =' -- adopt-project.py scripts/ gives one line. git grep -n 'for name in GNU_MAKE_READS' gives only _make_reads_first's lookup and the clause's list.
• Rows for shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's collision line at roots that keep Makefile + GNUmakefile/ and Makefile + makefile/: the line says GNU make stops at the directory, and make -n bootstrap beside the template's Makefile exits 2 with Is a directory..
• Rows for adopt-project.py: execute's NEXT prints python3 scripts/bootstrap.py where GNU make at the root reads a kept GNUmakefile or makefile, and its NOTE names that file and what it takes for make bootstrap to run there (#1047) #1121's NOTE and shape_materialize.py and the pinned AGENTS-shape.md: where the root keeps a GNUmakefile or a makefile, which GNU make reads first, the Makefile collision follow-up no longer says the shape's make targets arrive with the merge, and the pinned file says make reads the first of the three and the block reaches it only once that file holds it (#1048) #1105's line: each names python on Windows, and each NOTE beside a kept Makefile names the merge.
• A row for the link case, measured with GNUmakefile -> Makefile.
• A row that fails under K9.
• Mutants for each, each applied alone and restored with its sha256 checked.
• These pass on git 2.43 and 2.47.1: tests/test_adopt_make_bootstrap_followups.py, tests/test_adopt_next_make_reads_first.py, tests/test_adopt_collision_make_reads.py, tests/test_pinned_copy_text.py, tests/test_adopt_e2e.py and the hygiene files.

Lane: openreposhape-5 (openRepoShape-5)