mirror of
https://github.com/hanzoai/papers.git
synced 2026-10-22 17:20:17 +00:00
zooqueenandHanzo Dev
44ec401a31
`make -k all` returned 0 over any number of broken papers, so CI was green the
whole time it existed. Four things had to be true for that:
- every recipe line ended in `|| true` and the rule printed "FAIL" instead of
exiting, so make never saw a failure;
- nonstopmode emits a PDF for almost any input, so "the PDF exists" is not
the same question as "the paper compiled" — zen-nano_whitepaper produced
one through 23 LaTeX errors;
- pdfs/ was tracked, so a checkout handed the "check every paper came out"
step 100 of the PDFs it was about to look for;
- discovery was three wildcards, so 69 of the repo's 219 papers were
compiled by nothing at all, defense/ by a second copy of the same
swallowing loop written inline in ci.yml.
The rule is now one latexmk per paper and its exit status is the gate:
-halt-on-error (no PDF from a broken paper), -bibtex (a declared bibliography
must resolve rather than being skipped when the .bib is absent), -Werror (an
undefined citation is a defect, not a log note), -cd (each paper compiles from
its own directory, so zen/ finds references.bib and defense/ finds
hanzo-defense.sty). .DELETE_ON_ERROR so a failed run leaves nothing behind,
and pdfs/ is untracked because it is output.
Run against the tree it was guarding, the honest gate fails 24 papers:
6 \begin{itemize} with its backslash-b eaten, leaving a bare `egin{itemize}`
and no \end — artist, designer-instruct, eco, nano, next, omni
3 the same list with the opener deleted outright, so five \item stood in no
environment — artist-edit, designer-thinking, scribe
4 \and in a cover page's raw \@author: it expands to \end{tabular} and closed
a tabular nothing had opened, and \thanks left hyperref pointing at a
footnote anchor that was never created; plus five stray \end{table} at the
end of those documents — director, foley, musician, video
9 \color{gray} under \usepackage{color}, which defines eight names and not
that one — xcolor is the superset the working papers already load
6 \bibliography{references} naming a database that was never written, so
every citation rendered [?] — zen/references.bib now holds all 27 works
9 \bibliographystyle+\bibliography in papers with no \cite at all — dropped
3 characters pdflatex has no glyph for: CO₂ -> CO\textsubscript{2},
↔ -> $\leftrightarrow$, Persian letters -> their code points, the
SentencePiece marker -> \textvisiblespace
1 \IF{...} \COMMENT{...}: algorithmic's \IF opens a list and the comment
landed in it before any \item; the comment slot is \IF[...]{...}
1 \ref{eq:anchor_update} to an equation the paper never writes
All 219 papers now compile from a clean tree: `make -k all` exits 0.
Co-authored-by: Hanzo Dev <[email protected]>
118 lines
4.9 KiB
Makefile
118 lines
4.9 KiB
Makefile
# Makefile for Hanzo Papers
|
|
# Compiles every standalone .tex paper in the repo to a PDF under pdfs/.
|
|
#
|
|
# Usage:
|
|
# make -k all # Build every paper, report all failures
|
|
# make pdfs/hanzo-aso.pdf # Build one paper
|
|
# make list # Every paper, one path per line
|
|
# make clean # Remove aux files
|
|
# make clean-all # Remove aux + PDFs
|
|
#
|
|
# A paper that does not compile fails. Every recipe line here used to end in
|
|
# `|| true` and the rule printed "FAIL" instead of exiting, so `make all`
|
|
# returned 0 over any number of broken papers — the gate in
|
|
# .hanzo/workflows/ci.yml was green for as long as it existed. Use -k to build
|
|
# the rest after one breaks; that is what -k is for, and it is the only reason
|
|
# the old rule had to swallow the status.
|
|
#
|
|
# Docker (no local TeX install needed):
|
|
# docker run --rm -v "$$PWD:/workdir" -w /workdir texlive/texlive:latest make -k all
|
|
|
|
# A paper is any .tex with a \documentclass, wherever it lives. This is the one
|
|
# definition of the paper set; scripts/gen-index.sh reads it from here rather
|
|
# than restating it. Excluded: sections/ and shared/ hold fragments with no
|
|
# \documentclass of their own, site/ is the website, pdfs/ is output, and
|
|
# defense/logo-convert.tex is a helper that rasterises the logo through inkscape.
|
|
#
|
|
# The previous discovery was three wildcards over the root, zen/ and defense/,
|
|
# which is 150 of the 219 papers in the repo. The other 69 live in per-paper
|
|
# directories and were compiled by nothing.
|
|
ALL_TEX := $(shell grep -rl '^\\documentclass' --include='*.tex' . \
|
|
| sed 's|^\./||' \
|
|
| grep -v -e '^sections/' -e '/sections/' -e '^shared/' -e '^site/' \
|
|
-e '^pdfs/' -e 'node_modules/' -e '^defense/logo-convert\.tex$$' \
|
|
| sort)
|
|
|
|
# Map to output PDFs under pdfs/
|
|
ALL_PDFS := $(patsubst %.tex,pdfs/%.pdf,$(ALL_TEX))
|
|
|
|
# A recipe that exits nonzero leaves no PDF behind, so a stale artifact can
|
|
# never stand in for a build that did not happen.
|
|
.DELETE_ON_ERROR:
|
|
|
|
# Default: build everything
|
|
.PHONY: all
|
|
all: $(ALL_PDFS)
|
|
@echo ""
|
|
@echo "=== Build complete ==="
|
|
@echo "Papers compiled: $(words $(ALL_PDFS))"
|
|
|
|
# One paper, one rule, wherever it lives. latexmk decides how many passes are
|
|
# needed and runs bibtex; its exit status is the answer to "did this paper
|
|
# compile?", which is the whole gate.
|
|
#
|
|
# -halt-on-error stop at the first error rather than nonstopmode's "carry on
|
|
# and emit a PDF anyway" — a broken paper leaves no PDF
|
|
# -bibtex run bibtex whenever a paper declares a bibliography, so
|
|
# naming a .bib that does not exist fails instead of being
|
|
# quietly skipped (latexmk's default)
|
|
# -Werror an undefined citation or cross-reference is a defect in a
|
|
# paper, not a note in a log nobody reads
|
|
# -cd compile each paper from its own directory, so a paper's
|
|
# relative paths mean the same thing to make as they do to a
|
|
# human opening the file: zen/ finds references.bib, defense/
|
|
# finds hanzo-defense.sty, the root finds shared/
|
|
# -outdir absolute, because -cd moved us: PDF and aux land together
|
|
# under pdfs/, never at the repo root
|
|
pdfs/%.pdf: %.tex
|
|
@mkdir -p $(dir $@)
|
|
@echo "Compiling $< ..."
|
|
@latexmk -pdf -halt-on-error -bibtex -Werror -interaction=nonstopmode -quiet \
|
|
-cd -outdir=$(CURDIR)/$(dir $@) $< > $(dir $@)$(*F).latexmk 2>&1 \
|
|
|| { echo " FAIL $<"; \
|
|
grep -A5 '^!' $(dir $@)$(*F).log; \
|
|
sed -n '/^Latexmk: Summary/,/^----/p' $(dir $@)$(*F).latexmk; \
|
|
false; }
|
|
@echo " OK $@"
|
|
|
|
# Clean auxiliary files (keep PDFs). The second sweep clears what the previous
|
|
# rule left at the repo root; latexmk writes beside the PDF and never there.
|
|
.PHONY: clean
|
|
clean:
|
|
@mkdir -p pdfs
|
|
@find pdfs -type f ! -name '*.pdf' -delete
|
|
@find . -maxdepth 1 \( -name '*.aux' -o -name '*.log' -o -name '*.out' \
|
|
-o -name '*.toc' -o -name '*.bbl' -o -name '*.blg' \
|
|
-o -name '*.fls' -o -name '*.fdb_latexmk' \) -delete
|
|
@echo "Cleaned."
|
|
|
|
# Clean everything
|
|
.PHONY: clean-all
|
|
clean-all:
|
|
@echo "Removing all generated files..."
|
|
@rm -rf pdfs
|
|
@echo "Done"
|
|
|
|
# Every paper, one per line. scripts/gen-index.sh reads this.
|
|
.PHONY: list
|
|
list:
|
|
@printf '%s\n' $(ALL_TEX)
|
|
|
|
.PHONY: help
|
|
help:
|
|
@echo "Hanzo Papers Makefile"
|
|
@echo ""
|
|
@echo "Targets:"
|
|
@echo " make -k all - Compile all $(words $(ALL_TEX)) papers, report every failure"
|
|
@echo " make pdfs/<name>.pdf - Compile a specific paper"
|
|
@echo " make list - List all discovered papers"
|
|
@echo " make clean - Remove auxiliary files"
|
|
@echo " make clean-all - Remove all generated files"
|
|
@echo ""
|
|
@echo "Docker:"
|
|
@echo " docker run --rm -v \"\$$PWD:/workdir\" -w /workdir texlive/texlive:latest make -k all"
|
|
|
|
.PHONY: index
|
|
index:
|
|
@scripts/gen-index.sh
|