Files
akbasic/tools/docs_screenshots.sh

234 lines
7.1 KiB
Bash
Raw Permalink Normal View History

Generate the documentation's figures from the listings they illustrate Chapters 6 and 8 described what a verb draws in prose. Eight figures now show it, and each one is produced by running the BASIC listing printed immediately above it -- so a picture cannot drift away from the code beside it, which is the way a screenshot goes wrong and the way nothing notices. tools/screenshot.c is a second SDL host, much smaller than the frontend: dummy video driver, software renderer, run to completion, read the target back, write a PNG. It draws no text layer on purpose, so a READY in the corner is not noise in a figure about BOX and no font has to be resolved. tools/docs_screenshots.sh reads the new screenshot=NAME fence tag straight out of the markdown. size=WxH is the second tag, and SCALE's figure uses it: the point being made is a 320x200 listing filling a larger window, which cannot be made on a 320x200 surface. Two gates, answering different questions. docs_examples fails a tagged block with no image, in both configurations, so a figure cannot be added and forgotten. docs_screenshots -- a CTest, AKGL build only -- re-renders every figure and compares byte for byte, so a listing edited without regenerating fails. Only the second catches a stale picture. The PNGs are checked in because a reader on the forge has no build tree, and docs/images/README.md says loudly that they are generated. Regenerating is never part of a build: the target is run deliberately, so a make cannot put eight binary diffs in front of whoever ran it. Drawing the BOX figure caught a defect in TODO.md itself. Deviation 16 claimed in bold that BOX fills on a negative angle while its own paragraph said the fill was filed rather than implemented. BOX cannot fill, and filled_rect is reached by no verb as a result. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-01 07:37:04 -04:00
#!/bin/bash
#
# Regenerate every figure in the documentation from the listing shown beside it.
#
# A chapter that says "CIRCLE draws ellipses" and shows a listing wants a picture
# of what that listing draws, and the failure mode of a picture is that it stops
# being of the code next to it. Nothing tells you: a screenshot taken by hand in
# 2026 still looks like a screenshot in 2027, long after the verb it illustrates
# has changed.
#
# So the figure is not an asset, it is *output*. A block tagged
#
# ```basic requires=akgl screenshot=circle
#
# is run by tools/screenshot.c and written to docs/images/circle.png, and the
# chapter shows that file. Regenerate and the picture follows the code. The
# generated PNGs are checked in on purpose -- a reader on the forge has no build
# tree -- and tests/docs_examples.sh fails if a tagged block has no image, so a
# new figure cannot be forgotten.
#
# **This is not run by the build.** `cmake --build build-akgl --target
# docs_screenshots` is deliberate; see the target's comment in CMakeLists.txt.
#
# Exit status is the number of figures that failed, the house convention.
set -u
ROOT=""
TOOL=""
CHECK=0
FAILURES=0
usage()
{
cat >&2 <<'EOF'
usage: docs_screenshots.sh --root DIR --tool PATH [--check] [FILE...]
--root DIR repository root; images are written to DIR/docs/images
--tool PATH the built akbasic_screenshot
--check render to a scratch directory and compare, changing nothing
FILE... which documents to regenerate (default: docs/*.md)
EOF
exit 2
}
while [ $# -gt 0 ]; do
case "$1" in
--root) ROOT="$2"; shift 2 ;;
--tool) TOOL="$2"; shift 2 ;;
--check) CHECK=1; shift ;;
--help|-h) usage ;;
--*) echo "unknown option $1" >&2; usage ;;
*) break ;;
esac
done
[ -n "${ROOT}" ] || usage
[ -n "${TOOL}" ] || usage
[ -x "${TOOL}" ] || { echo "FAIL: no screenshot tool at ${TOOL}" >&2; exit 2; }
# Absolute, because the loop below cd's into a sandbox to run each program --
# a listing that loads an asset resolves it relative to its own directory.
case "${TOOL}" in
/*) ;;
*) TOOL="${PWD}/${TOOL}" ;;
esac
cd "${ROOT}" || exit 2
ROOT="${PWD}"
DOCS=("$@")
if [ ${#DOCS[@]} -eq 0 ]; then
DOCS=(docs/*.md)
fi
IMAGES="${ROOT}/docs/images"
mkdir -p "${IMAGES}" || exit 2
WORK="$(mktemp -d)"
trap 'rm -rf "${WORK}"' EXIT
# --check renders somewhere else entirely and compares, so a run that finds a
# stale figure does not also fix it. A test that repairs what it is measuring
# passes the second time for the wrong reason.
if [ "${CHECK}" -eq 1 ]; then
OUTDIR="${WORK}/rendered"
mkdir -p "${OUTDIR}" || exit 2
else
OUTDIR="${IMAGES}"
fi
# The value of attribute $1 in an info string $2, or empty. Same shape as the
# `attr` in tests/docs_examples.sh, and deliberately so -- the two read the same
# fence tags and a second dialect of them would be a bug waiting to happen.
attr()
{
local name="$1" info="$2" word
for word in ${info}; do
case "${word}" in
"${name}"=*) echo "${word#*=}"; return 0 ;;
esac
done
echo ""
}
Rewrite the two game tutorials as instructions rather than commentary Chapters 17 and 18 read as a code review of a finished listing: they explained why each decision had been made, walked through the project's own history, and led with what had once been broken. A reader who wanted to build the game got the reasoning and had to reconstruct the program. Both are now step-by-step. Each opens with a picture of the finished game and a bullet list of the steps, each bullet is a section, and each section states its goal, shows the code, and says how to check it. Chapter 17 is sixteen steps and Chapter 18 is thirteen, and the last of each is the assembly: the order of the file, the full declaration block, and the routines the earlier steps referred to. Project history is gone -- it belongs in Chapter 14 and in git -- and where a listing has to do something awkward, the tutorial shows how first and names the `TODO.md` item that will make it unnecessary second. **Three defects had no entry anywhere**, which the rewrite found by trying to state each rule as a rule. §6 item 35: `a - b + c` computes `a - (b + c)`, because `subtraction()` sits above `addition()` as its own precedence level and the inner loop eats the `+`. Item 36: only one unparenthesised `AND` or `OR` is matched, which is item 12's `if`-where-`while` on the one operator pair item 12 did not reach. Item 37: a `GOTO` out of a `FOR` or a `DO` leaks the loop's scope, so a main loop written that way stops on the thirty-second lost life -- which is why both games are built out of `LABEL` and `GOTO`, and it is a workaround rather than a preference. Chapter 3 gains the identifier rule the third trial ran into: there is no underscore in a name, and the error says `UNKNOWN TOKEN _`. **`tools/screenshot.c` learned to draw the text layer**, behind a new `text=1` fence attribute, because Chapter 17's game is characters in the grid and a figure without that layer is two sprites on black. It opens the bundled font at the size the standalone frontend uses, so a figure's cell size is the reader's cell size, and it uses the akgl sink alone rather than a tee so the program's output lands in the picture instead of on the stdout the caller reads to decide a figure failed. Both new figures -- `breakout-game.png` and `breakout-game-artwork.png` -- are generated from listings in the chapters like every other one. Verified by handing each chapter, alone, to an agent on a much smaller model and telling it to build the game from the tutorial text with the `examples/` tree off limits. The first pass scored 3.5 and 3 out of 10 and named what was missing: routines referred to but never shown, the third level layout, the sprite `DATA`, the font table, edits to earlier routines that were never marked as edits. Those are now in. The second pass built a 658-line Chapter 17 game that plays itself for ninety seconds with the score at 1890 and no error line, and the third built a 1053-line Chapter 18 game with 61 labels, no invented routines, no gaps found, and forty seconds clean. 9/10 and 8/10. One real bug in the new prose, caught in review: Chapter 18's `HITBAR` did not set the `HIT#` that `BALLPADDLE` reads to decide whether the paddle already caught the ball. Both suites green in both configurations, `docs_examples` and `docs_screenshots --check` pass. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EwxGB6TdoVvZ11KQQME9cL
2026-08-02 08:08:23 -04:00
# The font a `text=1` figure is drawn with. The bundled one, at the size the
# standalone frontend opens it at, so a figure of a text-grid program has the
# same cell size the reader's own window will have.
FONT="${ROOT}/assets/fonts/C64_Pro_Mono-STYLE.ttf"
Generate the documentation's figures from the listings they illustrate Chapters 6 and 8 described what a verb draws in prose. Eight figures now show it, and each one is produced by running the BASIC listing printed immediately above it -- so a picture cannot drift away from the code beside it, which is the way a screenshot goes wrong and the way nothing notices. tools/screenshot.c is a second SDL host, much smaller than the frontend: dummy video driver, software renderer, run to completion, read the target back, write a PNG. It draws no text layer on purpose, so a READY in the corner is not noise in a figure about BOX and no font has to be resolved. tools/docs_screenshots.sh reads the new screenshot=NAME fence tag straight out of the markdown. size=WxH is the second tag, and SCALE's figure uses it: the point being made is a 320x200 listing filling a larger window, which cannot be made on a 320x200 surface. Two gates, answering different questions. docs_examples fails a tagged block with no image, in both configurations, so a figure cannot be added and forgotten. docs_screenshots -- a CTest, AKGL build only -- re-renders every figure and compares byte for byte, so a listing edited without regenerating fails. Only the second catches a stale picture. The PNGs are checked in because a reader on the forge has no build tree, and docs/images/README.md says loudly that they are generated. Regenerating is never part of a build: the target is run deliberately, so a make cannot put eight binary diffs in front of whoever ran it. Drawing the BOX figure caught a defect in TODO.md itself. Deviation 16 claimed in bold that BOX fills on a negative angle while its own paragraph said the fill was filed rather than implemented. BOX cannot fill, and filled_rect is reached by no verb as a result. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-01 07:37:04 -04:00
# One figure: write the listing to a file, run it, keep the PNG.
render()
{
Rewrite the two game tutorials as instructions rather than commentary Chapters 17 and 18 read as a code review of a finished listing: they explained why each decision had been made, walked through the project's own history, and led with what had once been broken. A reader who wanted to build the game got the reasoning and had to reconstruct the program. Both are now step-by-step. Each opens with a picture of the finished game and a bullet list of the steps, each bullet is a section, and each section states its goal, shows the code, and says how to check it. Chapter 17 is sixteen steps and Chapter 18 is thirteen, and the last of each is the assembly: the order of the file, the full declaration block, and the routines the earlier steps referred to. Project history is gone -- it belongs in Chapter 14 and in git -- and where a listing has to do something awkward, the tutorial shows how first and names the `TODO.md` item that will make it unnecessary second. **Three defects had no entry anywhere**, which the rewrite found by trying to state each rule as a rule. §6 item 35: `a - b + c` computes `a - (b + c)`, because `subtraction()` sits above `addition()` as its own precedence level and the inner loop eats the `+`. Item 36: only one unparenthesised `AND` or `OR` is matched, which is item 12's `if`-where-`while` on the one operator pair item 12 did not reach. Item 37: a `GOTO` out of a `FOR` or a `DO` leaks the loop's scope, so a main loop written that way stops on the thirty-second lost life -- which is why both games are built out of `LABEL` and `GOTO`, and it is a workaround rather than a preference. Chapter 3 gains the identifier rule the third trial ran into: there is no underscore in a name, and the error says `UNKNOWN TOKEN _`. **`tools/screenshot.c` learned to draw the text layer**, behind a new `text=1` fence attribute, because Chapter 17's game is characters in the grid and a figure without that layer is two sprites on black. It opens the bundled font at the size the standalone frontend uses, so a figure's cell size is the reader's cell size, and it uses the akgl sink alone rather than a tee so the program's output lands in the picture instead of on the stdout the caller reads to decide a figure failed. Both new figures -- `breakout-game.png` and `breakout-game-artwork.png` -- are generated from listings in the chapters like every other one. Verified by handing each chapter, alone, to an agent on a much smaller model and telling it to build the game from the tutorial text with the `examples/` tree off limits. The first pass scored 3.5 and 3 out of 10 and named what was missing: routines referred to but never shown, the third level layout, the sprite `DATA`, the font table, edits to earlier routines that were never marked as edits. Those are now in. The second pass built a 658-line Chapter 17 game that plays itself for ninety seconds with the score at 1890 and no error line, and the third built a 1053-line Chapter 18 game with 61 labels, no invented routines, no gaps found, and forty seconds clean. 9/10 and 8/10. One real bug in the new prose, caught in review: Chapter 18's `HITBAR` did not set the `HIT#` that `BALLPADDLE` reads to decide whether the paddle already caught the ball. Both suites green in both configurations, `docs_examples` and `docs_screenshots --check` pass. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EwxGB6TdoVvZ11KQQME9cL
2026-08-02 08:08:23 -04:00
local name="$1" body="$2" size="$3" where="$4" setup="$5" text="$6"
local w=320 h=200 out="${OUTDIR}/${name}.png" log="" font=""
Generate the documentation's figures from the listings they illustrate Chapters 6 and 8 described what a verb draws in prose. Eight figures now show it, and each one is produced by running the BASIC listing printed immediately above it -- so a picture cannot drift away from the code beside it, which is the way a screenshot goes wrong and the way nothing notices. tools/screenshot.c is a second SDL host, much smaller than the frontend: dummy video driver, software renderer, run to completion, read the target back, write a PNG. It draws no text layer on purpose, so a READY in the corner is not noise in a figure about BOX and no font has to be resolved. tools/docs_screenshots.sh reads the new screenshot=NAME fence tag straight out of the markdown. size=WxH is the second tag, and SCALE's figure uses it: the point being made is a 320x200 listing filling a larger window, which cannot be made on a 320x200 surface. Two gates, answering different questions. docs_examples fails a tagged block with no image, in both configurations, so a figure cannot be added and forgotten. docs_screenshots -- a CTest, AKGL build only -- re-renders every figure and compares byte for byte, so a listing edited without regenerating fails. Only the second catches a stale picture. The PNGs are checked in because a reader on the forge has no build tree, and docs/images/README.md says loudly that they are generated. Regenerating is never part of a build: the target is run deliberately, so a make cannot put eight binary diffs in front of whoever ran it. Drawing the BOX figure caught a defect in TODO.md itself. Deviation 16 claimed in bold that BOX fills on a negative angle while its own paragraph said the fill was filed rather than implemented. BOX cannot fill, and filled_rect is reached by no verb as a result. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-01 07:37:04 -04:00
if [ -n "${size}" ]; then
w="${size%x*}"
h="${size#*x}"
fi
Rewrite the two game tutorials as instructions rather than commentary Chapters 17 and 18 read as a code review of a finished listing: they explained why each decision had been made, walked through the project's own history, and led with what had once been broken. A reader who wanted to build the game got the reasoning and had to reconstruct the program. Both are now step-by-step. Each opens with a picture of the finished game and a bullet list of the steps, each bullet is a section, and each section states its goal, shows the code, and says how to check it. Chapter 17 is sixteen steps and Chapter 18 is thirteen, and the last of each is the assembly: the order of the file, the full declaration block, and the routines the earlier steps referred to. Project history is gone -- it belongs in Chapter 14 and in git -- and where a listing has to do something awkward, the tutorial shows how first and names the `TODO.md` item that will make it unnecessary second. **Three defects had no entry anywhere**, which the rewrite found by trying to state each rule as a rule. §6 item 35: `a - b + c` computes `a - (b + c)`, because `subtraction()` sits above `addition()` as its own precedence level and the inner loop eats the `+`. Item 36: only one unparenthesised `AND` or `OR` is matched, which is item 12's `if`-where-`while` on the one operator pair item 12 did not reach. Item 37: a `GOTO` out of a `FOR` or a `DO` leaks the loop's scope, so a main loop written that way stops on the thirty-second lost life -- which is why both games are built out of `LABEL` and `GOTO`, and it is a workaround rather than a preference. Chapter 3 gains the identifier rule the third trial ran into: there is no underscore in a name, and the error says `UNKNOWN TOKEN _`. **`tools/screenshot.c` learned to draw the text layer**, behind a new `text=1` fence attribute, because Chapter 17's game is characters in the grid and a figure without that layer is two sprites on black. It opens the bundled font at the size the standalone frontend uses, so a figure's cell size is the reader's cell size, and it uses the akgl sink alone rather than a tee so the program's output lands in the picture instead of on the stdout the caller reads to decide a figure failed. Both new figures -- `breakout-game.png` and `breakout-game-artwork.png` -- are generated from listings in the chapters like every other one. Verified by handing each chapter, alone, to an agent on a much smaller model and telling it to build the game from the tutorial text with the `examples/` tree off limits. The first pass scored 3.5 and 3 out of 10 and named what was missing: routines referred to but never shown, the third level layout, the sprite `DATA`, the font table, edits to earlier routines that were never marked as edits. Those are now in. The second pass built a 658-line Chapter 17 game that plays itself for ninety seconds with the score at 1890 and no error line, and the third built a 1053-line Chapter 18 game with 61 labels, no invented routines, no gaps found, and forty seconds clean. 9/10 and 8/10. One real bug in the new prose, caught in review: Chapter 18's `HITBAR` did not set the `HIT#` that `BALLPADDLE` reads to decide whether the paddle already caught the ball. Both suites green in both configurations, `docs_examples` and `docs_screenshots --check` pass. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EwxGB6TdoVvZ11KQQME9cL
2026-08-02 08:08:23 -04:00
# `text=1` says this program's output is characters rather than drawing, so
# the tool has to render the grid -- see the comment at the head of
# tools/screenshot.c. Everything else gets no font and no text layer.
if [ -n "${text}" ]; then
if [ ! -r "${FONT}" ]; then
echo "FAIL ${where}: text=1 needs a font at ${FONT}" >&2
FAILURES=$((FAILURES + 1))
return
fi
font="${FONT}"
fi
Generate the documentation's figures from the listings they illustrate Chapters 6 and 8 described what a verb draws in prose. Eight figures now show it, and each one is produced by running the BASIC listing printed immediately above it -- so a picture cannot drift away from the code beside it, which is the way a screenshot goes wrong and the way nothing notices. tools/screenshot.c is a second SDL host, much smaller than the frontend: dummy video driver, software renderer, run to completion, read the target back, write a PNG. It draws no text layer on purpose, so a READY in the corner is not noise in a figure about BOX and no font has to be resolved. tools/docs_screenshots.sh reads the new screenshot=NAME fence tag straight out of the markdown. size=WxH is the second tag, and SCALE's figure uses it: the point being made is a 320x200 listing filling a larger window, which cannot be made on a 320x200 surface. Two gates, answering different questions. docs_examples fails a tagged block with no image, in both configurations, so a figure cannot be added and forgotten. docs_screenshots -- a CTest, AKGL build only -- re-renders every figure and compares byte for byte, so a listing edited without regenerating fails. Only the second catches a stale picture. The PNGs are checked in because a reader on the forge has no build tree, and docs/images/README.md says loudly that they are generated. Regenerating is never part of a build: the target is run deliberately, so a make cannot put eight binary diffs in front of whoever ran it. Drawing the BOX figure caught a defect in TODO.md itself. Deviation 16 claimed in bold that BOX fills on a negative angle while its own paragraph said the fill was filed rather than implemented. BOX cannot fill, and filled_rect is reached by no verb as a result. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-01 07:37:04 -04:00
# The same setup scripts tests/docs_examples.sh uses, run in the same place
# relative to the program. A listing that loads `ship.png` needs one whether
# it is being checked or being photographed.
if [ -n "${setup}" ]; then
if [ ! -x "${ROOT}/tests/docs_setups/${setup}.sh" ]; then
echo "FAIL ${where}: no setup script ${setup}.sh" >&2
FAILURES=$((FAILURES + 1))
return
fi
( cd "${WORK}" && "${ROOT}/tests/docs_setups/${setup}.sh" ) || {
echo "FAIL ${where}: setup ${setup} failed" >&2
FAILURES=$((FAILURES + 1))
return
}
fi
printf '%s' "${body}" > "${WORK}/${name}.bas"
# **stdout only.** The tool puts the interpreter's sink on stdout and libakgl
# logs its registry chatter -- "Actor akbasic:actor:1 initialized" -- to
# stderr, so merging the two would make every sprite figure look like a
# failing program. stderr is shown when the tool itself refuses, which is
# when it is worth reading.
Rewrite the two game tutorials as instructions rather than commentary Chapters 17 and 18 read as a code review of a finished listing: they explained why each decision had been made, walked through the project's own history, and led with what had once been broken. A reader who wanted to build the game got the reasoning and had to reconstruct the program. Both are now step-by-step. Each opens with a picture of the finished game and a bullet list of the steps, each bullet is a section, and each section states its goal, shows the code, and says how to check it. Chapter 17 is sixteen steps and Chapter 18 is thirteen, and the last of each is the assembly: the order of the file, the full declaration block, and the routines the earlier steps referred to. Project history is gone -- it belongs in Chapter 14 and in git -- and where a listing has to do something awkward, the tutorial shows how first and names the `TODO.md` item that will make it unnecessary second. **Three defects had no entry anywhere**, which the rewrite found by trying to state each rule as a rule. §6 item 35: `a - b + c` computes `a - (b + c)`, because `subtraction()` sits above `addition()` as its own precedence level and the inner loop eats the `+`. Item 36: only one unparenthesised `AND` or `OR` is matched, which is item 12's `if`-where-`while` on the one operator pair item 12 did not reach. Item 37: a `GOTO` out of a `FOR` or a `DO` leaks the loop's scope, so a main loop written that way stops on the thirty-second lost life -- which is why both games are built out of `LABEL` and `GOTO`, and it is a workaround rather than a preference. Chapter 3 gains the identifier rule the third trial ran into: there is no underscore in a name, and the error says `UNKNOWN TOKEN _`. **`tools/screenshot.c` learned to draw the text layer**, behind a new `text=1` fence attribute, because Chapter 17's game is characters in the grid and a figure without that layer is two sprites on black. It opens the bundled font at the size the standalone frontend uses, so a figure's cell size is the reader's cell size, and it uses the akgl sink alone rather than a tee so the program's output lands in the picture instead of on the stdout the caller reads to decide a figure failed. Both new figures -- `breakout-game.png` and `breakout-game-artwork.png` -- are generated from listings in the chapters like every other one. Verified by handing each chapter, alone, to an agent on a much smaller model and telling it to build the game from the tutorial text with the `examples/` tree off limits. The first pass scored 3.5 and 3 out of 10 and named what was missing: routines referred to but never shown, the third level layout, the sprite `DATA`, the font table, edits to earlier routines that were never marked as edits. Those are now in. The second pass built a 658-line Chapter 17 game that plays itself for ninety seconds with the score at 1890 and no error line, and the third built a 1053-line Chapter 18 game with 61 labels, no invented routines, no gaps found, and forty seconds clean. 9/10 and 8/10. One real bug in the new prose, caught in review: Chapter 18's `HITBAR` did not set the `HIT#` that `BALLPADDLE` reads to decide whether the paddle already caught the ball. Both suites green in both configurations, `docs_examples` and `docs_screenshots --check` pass. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EwxGB6TdoVvZ11KQQME9cL
2026-08-02 08:08:23 -04:00
if ! log="$(cd "${WORK}" && "${TOOL}" "${name}.bas" "${out}" "${w}" "${h}" "${font}" \
Generate the documentation's figures from the listings they illustrate Chapters 6 and 8 described what a verb draws in prose. Eight figures now show it, and each one is produced by running the BASIC listing printed immediately above it -- so a picture cannot drift away from the code beside it, which is the way a screenshot goes wrong and the way nothing notices. tools/screenshot.c is a second SDL host, much smaller than the frontend: dummy video driver, software renderer, run to completion, read the target back, write a PNG. It draws no text layer on purpose, so a READY in the corner is not noise in a figure about BOX and no font has to be resolved. tools/docs_screenshots.sh reads the new screenshot=NAME fence tag straight out of the markdown. size=WxH is the second tag, and SCALE's figure uses it: the point being made is a 320x200 listing filling a larger window, which cannot be made on a 320x200 surface. Two gates, answering different questions. docs_examples fails a tagged block with no image, in both configurations, so a figure cannot be added and forgotten. docs_screenshots -- a CTest, AKGL build only -- re-renders every figure and compares byte for byte, so a listing edited without regenerating fails. Only the second catches a stale picture. The PNGs are checked in because a reader on the forge has no build tree, and docs/images/README.md says loudly that they are generated. Regenerating is never part of a build: the target is run deliberately, so a make cannot put eight binary diffs in front of whoever ran it. Drawing the BOX figure caught a defect in TODO.md itself. Deviation 16 claimed in bold that BOX fills on a negative angle while its own paragraph said the fill was filed rather than implemented. BOX cannot fill, and filled_rect is reached by no verb as a result. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-01 07:37:04 -04:00
2>"${WORK}/${name}.err")"; then
echo "FAIL ${where}: ${name} did not render" >&2
cat "${WORK}/${name}.err" >&2
FAILURES=$((FAILURES + 1))
return
fi
# A program that raised prints its error line and still exits zero, because
# a BASIC error is the script's and not the host's. For a figure that is
# still a failure: an image of a blank screen is worse than no image.
if [ -n "${log}" ]; then
echo "FAIL ${where}: ${name} rendered, but the program reported:" >&2
echo "${log}" >&2
FAILURES=$((FAILURES + 1))
return
fi
if [ "${CHECK}" -eq 1 ]; then
if [ ! -r "${IMAGES}/${name}.png" ]; then
echo "FAIL ${where}: docs/images/${name}.png does not exist" >&2
FAILURES=$((FAILURES + 1))
return
fi
if ! cmp -s "${out}" "${IMAGES}/${name}.png"; then
echo "FAIL ${where}: docs/images/${name}.png is not what that listing draws" >&2
echo " regenerate with: cmake --build <akgl build> --target docs_screenshots" >&2
FAILURES=$((FAILURES + 1))
return
fi
echo " ${name}.png matches"
return
fi
echo " ${name}.png (${w}x${h})"
}
for DOC in "${DOCS[@]}"; do
[ -r "${DOC}" ] || { echo "FAIL: no document at \"${DOC}\"" >&2; exit 2; }
# Walk the file collecting fenced blocks. Only `screenshot=` ones are kept;
# everything else is somebody else's problem, and docs_examples.sh is that
# somebody.
IN_BLOCK=0
INFO=""
BODY=""
START=0
LINENO=0
while IFS= read -r LINE; do
LINENO=$((LINENO + 1))
case "${LINE}" in
'```'*)
if [ "${IN_BLOCK}" -eq 0 ]; then
IN_BLOCK=1
INFO="${LINE#'```'}"
BODY=""
START="${LINENO}"
else
IN_BLOCK=0
NAME="$(attr screenshot "${INFO}")"
if [ -n "${NAME}" ]; then
render "${NAME}" "${BODY}" "$(attr size "${INFO}")" \
Rewrite the two game tutorials as instructions rather than commentary Chapters 17 and 18 read as a code review of a finished listing: they explained why each decision had been made, walked through the project's own history, and led with what had once been broken. A reader who wanted to build the game got the reasoning and had to reconstruct the program. Both are now step-by-step. Each opens with a picture of the finished game and a bullet list of the steps, each bullet is a section, and each section states its goal, shows the code, and says how to check it. Chapter 17 is sixteen steps and Chapter 18 is thirteen, and the last of each is the assembly: the order of the file, the full declaration block, and the routines the earlier steps referred to. Project history is gone -- it belongs in Chapter 14 and in git -- and where a listing has to do something awkward, the tutorial shows how first and names the `TODO.md` item that will make it unnecessary second. **Three defects had no entry anywhere**, which the rewrite found by trying to state each rule as a rule. §6 item 35: `a - b + c` computes `a - (b + c)`, because `subtraction()` sits above `addition()` as its own precedence level and the inner loop eats the `+`. Item 36: only one unparenthesised `AND` or `OR` is matched, which is item 12's `if`-where-`while` on the one operator pair item 12 did not reach. Item 37: a `GOTO` out of a `FOR` or a `DO` leaks the loop's scope, so a main loop written that way stops on the thirty-second lost life -- which is why both games are built out of `LABEL` and `GOTO`, and it is a workaround rather than a preference. Chapter 3 gains the identifier rule the third trial ran into: there is no underscore in a name, and the error says `UNKNOWN TOKEN _`. **`tools/screenshot.c` learned to draw the text layer**, behind a new `text=1` fence attribute, because Chapter 17's game is characters in the grid and a figure without that layer is two sprites on black. It opens the bundled font at the size the standalone frontend uses, so a figure's cell size is the reader's cell size, and it uses the akgl sink alone rather than a tee so the program's output lands in the picture instead of on the stdout the caller reads to decide a figure failed. Both new figures -- `breakout-game.png` and `breakout-game-artwork.png` -- are generated from listings in the chapters like every other one. Verified by handing each chapter, alone, to an agent on a much smaller model and telling it to build the game from the tutorial text with the `examples/` tree off limits. The first pass scored 3.5 and 3 out of 10 and named what was missing: routines referred to but never shown, the third level layout, the sprite `DATA`, the font table, edits to earlier routines that were never marked as edits. Those are now in. The second pass built a 658-line Chapter 17 game that plays itself for ninety seconds with the score at 1890 and no error line, and the third built a 1053-line Chapter 18 game with 61 labels, no invented routines, no gaps found, and forty seconds clean. 9/10 and 8/10. One real bug in the new prose, caught in review: Chapter 18's `HITBAR` did not set the `HIT#` that `BALLPADDLE` reads to decide whether the paddle already caught the ball. Both suites green in both configurations, `docs_examples` and `docs_screenshots --check` pass. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EwxGB6TdoVvZ11KQQME9cL
2026-08-02 08:08:23 -04:00
"${DOC}:${START}" "$(attr setup "${INFO}")" \
"$(attr text "${INFO}")"
Generate the documentation's figures from the listings they illustrate Chapters 6 and 8 described what a verb draws in prose. Eight figures now show it, and each one is produced by running the BASIC listing printed immediately above it -- so a picture cannot drift away from the code beside it, which is the way a screenshot goes wrong and the way nothing notices. tools/screenshot.c is a second SDL host, much smaller than the frontend: dummy video driver, software renderer, run to completion, read the target back, write a PNG. It draws no text layer on purpose, so a READY in the corner is not noise in a figure about BOX and no font has to be resolved. tools/docs_screenshots.sh reads the new screenshot=NAME fence tag straight out of the markdown. size=WxH is the second tag, and SCALE's figure uses it: the point being made is a 320x200 listing filling a larger window, which cannot be made on a 320x200 surface. Two gates, answering different questions. docs_examples fails a tagged block with no image, in both configurations, so a figure cannot be added and forgotten. docs_screenshots -- a CTest, AKGL build only -- re-renders every figure and compares byte for byte, so a listing edited without regenerating fails. Only the second catches a stale picture. The PNGs are checked in because a reader on the forge has no build tree, and docs/images/README.md says loudly that they are generated. Regenerating is never part of a build: the target is run deliberately, so a make cannot put eight binary diffs in front of whoever ran it. Drawing the BOX figure caught a defect in TODO.md itself. Deviation 16 claimed in bold that BOX fills on a negative angle while its own paragraph said the fill was filed rather than implemented. BOX cannot fill, and filled_rect is reached by no verb as a result. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-01 07:37:04 -04:00
fi
fi
;;
*)
if [ "${IN_BLOCK}" -eq 1 ]; then
BODY="${BODY}${LINE}"$'\n'
fi
;;
esac
done < "${DOC}"
done
if [ "${FAILURES}" -ne 0 ]; then
echo "${FAILURES} figure(s) failed" >&2
fi
exit "${FAILURES}"