openRepoShape, setup.sh and install.sh: a URL, path or ref with a letter past ASCII is no control character under C either, as #760's byte rule decides in all seven checks that used [![:print:]], and every control character is still refused

install.sh

@@ -125,17 +125,30 @@ USAGE
 # setup.sh makes; there is no release at either, so this refuses them by
 # name. A control character is refused before anything is echoed, and
 # before it a character a terminal draws as something other than itself
-# (#837), named by its code point: first, because `[![:print:]]` is the
-# locale's, matching every byte past ASCII under `C` and none of those
-# characters under a UTF-8 locale, where the refusal below would echo them.
+# (#837), named by its code point: first, because the control check
+# matches none of those characters, in any locale, and the owner/name
+# refusal after it echoes the value.
+#
+# A CONTROL CHARACTER IS THE LAUNCHER'S RULE, #760's, which `main` holds
+# the directory to (#940): under `LC_ALL=C`, in a subshell, `[[:cntrl:]]`
+# or U+0080 to U+009F's two bytes (`\302`, then `\200` to `\237`), so the
+# answer is the same in every locale. It was `[![:print:]]`, which under
+# `C` matches every byte past ASCII: a letter past ASCII was refused as a
+# control character there, and as no owner/name under a UTF-8 locale. It is
+# no owner/name in either now, in those words. The trade is the launcher's
+# (#867): a byte that is not UTF-8, as a lone `0x9B` or an overlong form of
+# a control, passes, in every locale, and the refusal that names the value
+# prints it: a terminal that reads UTF-8 draws it as a character it cannot
+# read, and only one set to an 8-bit charset acts on it. `checked_version`
+# and the option check in `main` hold their values to the same rule.
 checked_repo() {
 	local misdrawn
 	misdrawn="$(LC_ALL=C; misdrawn_characters "$REPO")"
 	[ -z "$misdrawn" ] ||
 		die "\$OPENREPOSHAPE_REPO holds a character a terminal draws as something other than itself ($misdrawn): one that is invisible, is drawn as another character, ends a line or reorders the text around it. It must be an owner/name on github.com, such as opensoft/openRepoShape."
-	case "$REPO" in
-	*[![:print:]]*) die "\$OPENREPOSHAPE_REPO carries a control character; it must be an owner/name on github.com, such as opensoft/openRepoShape." ;;
-	esac
+	if (LC_ALL=C; case "$REPO" in *[[:cntrl:]]* | *$'\302'[$'\200'-$'\237']*) exit 0 ;; *) exit 1 ;; esac); then
+		die "\$OPENREPOSHAPE_REPO carries a control character; it must be an owner/name on github.com, such as opensoft/openRepoShape."
+	fi
 	case "$REPO" in
 	[A-Za-z0-9]*/[A-Za-z0-9._-]*) ;;
 	*) die "\$OPENREPOSHAPE_REPO is '$REPO', which is not an owner/name on github.com, such as opensoft/openRepoShape: this installer builds github.com URLs from it." ;;
@@ -151,7 +164,12 @@ checked_repo() {
 # shape, which is the only one that echoes it), plus the empty value, and,
 # between the length and the control character, a character a terminal
 # draws as something other than itself, as the launcher's `ref_fault` has
-# it (#837). Not shared, because this file arrives alone down a pipe.
+# it (#837). The control character is `checked_repo`'s rule, above, with
+# its trade (#940): a byte that is not UTF-8, as a lone `0x9B` or an
+# overlong form of a control, passes, and a version with a letter past
+# ASCII is the shape's to refuse, in its words, in every locale. A tag the
+# redirect names is printable ASCII by then (`latest_from_redirect`). Not
+# shared, because this file arrives alone down a pipe.
 #
 # THE SHAPE IS GIT'S RULE PART BY PART (#874), as `ref_fault`'s and
 # `setup.sh`'s are: no `..`, no empty part between slashes, no part that
@@ -169,9 +187,9 @@ checked_version() {
 	misdrawn="$(LC_ALL=C; misdrawn_characters "$ref")"
 	[ -z "$misdrawn" ] ||
 		die "$what holds a character a terminal draws as something other than itself ($misdrawn): one that is invisible, is drawn as another character, ends a line or reorders the text around it. No release tag, branch or commit holds one."
-	case "$ref" in
-	*[![:print:]]*) die "$what carries a control character, which is not a release tag, branch or commit." ;;
-	esac
+	if (LC_ALL=C; case "$ref" in *[[:cntrl:]]* | *$'\302'[$'\200'-$'\237']*) exit 0 ;; *) exit 1 ;; esac); then
+		die "$what carries a control character, which is not a release tag, branch or commit."
+	fi
 	case "$ref" in
 	[!A-Za-z0-9]* | *..* | */.* | *.lock | *.lock/* | *//* | */ | *. | *[!A-Za-z0-9._/-]*) die "$what is '$ref', which is not a release tag, branch or commit: a ref starts with an ASCII letter or digit, uses only ASCII letters, digits, '.', '_', '/', '-', has no '..' (git reads it as a revision range), no empty part between slashes ('//', or a '/' at the end), no part that starts with '.' or ends in '.lock' (the name git gives the lock file it writes beside a ref), and does not end in '.'." ;;
 	esac
@@ -189,7 +207,7 @@ checked_version() {
 # itself, never gathered into an array, so a reader and the analyser both
 # see them there. `tests/test_install_sh.py` holds every curl line to it.
 latest_from_redirect() {
-	local url="$1" headers line value status=0
+	local url="$1" headers line value tag status=0
 	FOUND=""
 	command -v curl >/dev/null 2>&1 || {
 		FAULT="curl is not installed"
@@ -213,13 +231,33 @@ latest_from_redirect() {
 			done
 			case "$value" in
 			*/releases/tag/?*)
-				FOUND="${value##*/releases/tag/}"
-				return 0
+				# The tag is the network's too, and `checked_version`, which
+				# it meets next, holds it to the byte rule a person's own
+				# values are held to (#940), whose trade passes a byte that
+				# is not UTF-8 and whose shape refusal prints the tag. So it
+				# is taken only when it is printable ASCII, under `LC_ALL=C`,
+				# which gives one answer in every locale. A release tag is
+				# ASCII, so this refuses no real one; any other tag names no
+				# release, unnamed, and `gh` is asked next.
+				tag="${value##*/releases/tag/}"
+				if (LC_ALL=C; case "$tag" in *[![:print:]]*) exit 1 ;; *) exit 0 ;; esac); then
+					FOUND="$tag"
+					return 0
+				fi
+				FAULT="it redirects to a location that names no release"
+				return 1
 				;;
 			esac
 			# The location is named only when it is text a terminal draws as
 			# it is: no control character, and no character it draws as
-			# something other than itself (#837).
+			# something other than itself (#837). It keeps the locale's
+			# `[![:print:]]`, not the byte rule a person's own values are
+			# held to (#940): it is the network's, naming it only lets the
+			# fault say more, and that class takes a byte that is not UTF-8,
+			# as a lone `0x9B` or an overlong form of a control, in no
+			# locale, where the byte rule's trade would print one from the
+			# network. The cost: under `C` a location with a letter past
+			# ASCII goes unnamed.
 			FAULT="it redirects to a location that names no release"
 			if [ -z "$(LC_ALL=C; misdrawn_characters "$value")" ]; then
 				case "$value" in
@@ -450,9 +488,12 @@ main() {
 			misdrawn="$(LC_ALL=C; misdrawn_characters "$1")"
 			[ -z "$misdrawn" ] ||
 				die "an option holds a character a terminal draws as something other than itself ($misdrawn): one that is invisible, is drawn as another character, ends a line or reorders the text around it. The one argument is a version, such as v1.0.1; with none, the newest release."
-			case "$1" in
-			*[![:print:]]*) die "an option carries a control character. The one argument is a version, such as v1.0.1; with none, the newest release." ;;
-			esac
+			# `checked_repo`'s rule for a control character, with its
+			# trade: a byte that is not UTF-8, as a lone `0x9B` or an
+			# overlong form of a control, passes (#940).
+			if (LC_ALL=C; case "$1" in *[[:cntrl:]]* | *$'\302'[$'\200'-$'\237']*) exit 0 ;; *) exit 1 ;; esac); then
+				die "an option carries a control character. The one argument is a version, such as v1.0.1; with none, the newest release."
+			fi
 			die "unknown option '$1'. The one argument is a version, such as v1.0.1; with none, the newest release."
 			;;
 		*)
@@ -474,11 +515,13 @@ main() {
 	# from the release that added the check, and only after this file had
 	# fetched it. The `case` is `install_commands`'s in `openRepoShape`,
 	# copied for the reason the drive-letter one below is, and the refusal
-	# says nothing of the value. `[[:cntrl:]]` and not `[![:print:]]`, which
-	# `checked_repo` uses: a directory with a letter the locale does not
-	# print (`café` under LC_ALL=C) is no fault of the directory's. The check
-	# runs under `LC_ALL=C` and names U+0080 to U+009F's two bytes (`\302`,
-	# then `\200` to `\237`), so a C1 control is refused in every locale,
+	# says nothing of the value. `[[:cntrl:]]` and not `[![:print:]]`: a
+	# directory with a letter the locale does not print (`café` under
+	# LC_ALL=C) is no fault of the directory's, and `checked_repo`,
+	# `checked_version` and the option check hold their values to these
+	# same bytes (#940). The check runs under `LC_ALL=C` and names U+0080
+	# to U+009F's two bytes (`\302`, then `\200` to `\237`), so a C1
+	# control is refused in every locale,
 	# whatever else the value holds, unless the locale reads those bytes as
 	# text (#858, below); U+2028 and U+2029, which only a UTF-8 locale's
 	# `[[:cntrl:]]` matched, are the next check's (#837).

openRepoShape

@@ -603,11 +603,12 @@ install_commands() {
 	# spelling below is. The control characters go first, and their refusal
 	# says nothing of the value, so that no refusal after it echoes one that
 	# can forge a line of output. It is `[[:cntrl:]]` and not
-	# `[![:print:]]`, which is what this file holds the repository and the
-	# ref to: a name with a letter the locale does not print, as `café`
-	# under LC_ALL=C, is no fault of the directory's. The check runs under
-	# `LC_ALL=C` and names U+0080 to U+009F's two bytes (`\302`, then `\200`
-	# to `\237`), so a C1 control is refused in every locale, whatever else
+	# `[![:print:]]`: a name with a letter the locale does not print, as
+	# `café` under LC_ALL=C, is no fault of the directory's, and the
+	# repository and the ref are held to these same bytes (#940,
+	# `checked_repo` below). The check runs under `LC_ALL=C` and names
+	# U+0080 to U+009F's two bytes (`\302`, then `\200` to `\237`), so a
+	# C1 control is refused in every locale, whatever else
 	# the value holds, unless the locale reads those bytes as text (#858,
 	# below); U+2028 and U+2029, which only a UTF-8 locale's `[[:cntrl:]]`
 	# matched, are the next check's (#837). Left to the locale,
@@ -794,11 +795,27 @@ find_python() {
 # sent by the fetch with it. An owner/name keeps it inside the URL, before
 # `.git`, so only those three forms lost it.
 #
+# A CONTROL CHARACTER IS #760's, AS `install_commands` READS ONE IN THE
+# DIRECTORY (#940): under `LC_ALL=C`, in a subshell, `[[:cntrl:]]` or
+# U+0080 to U+009F's two bytes (`\302`, then `\200` to `\237`), so the
+# answer is the same in every locale. It was `[![:print:]]`, which is the
+# locale's: under `C` and POSIX it matches every byte past ASCII, so a URL
+# or a path with a letter past ASCII (`/srv/café/openRepoShape.git`) was
+# refused there as carrying a control character, which it does not carry,
+# where a UTF-8 locale and the package's `--version` took it. The trade is
+# the directory's (#867): a byte that is not UTF-8, as a lone `0x9B` or an
+# overlong form of a control, passes, in every locale, as it passes the
+# package's check, and `--version` prints it: a terminal that reads UTF-8
+# draws it as a character it cannot read, and only one set to an 8-bit
+# charset acts on it. `setup.sh`, which is handed the clone URL as
+# `$OPENREPOSHAPE_REPO`, holds it to the same rule, and refuses U+2028 and
+# U+2029 there too, which this file refuses first by their code points
+# (below).
+#
 # A CHARACTER A TERMINAL DRAWS AS SOMETHING OTHER THAN ITSELF IS REFUSED
 # FIRST (#837), in words that name it by its code point and nothing else of
-# the value. First, because `[![:print:]]` is the locale's: under `C` it
-# matches every byte past ASCII, and under a UTF-8 locale none of these
-# characters, so after it the words would change with the locale.
+# the value. The control check after it matches none of these characters,
+# in any locale, so this is the check that refuses them.
 #
 # install.sh's `checked_repo` is stricter, an owner/name and nothing else,
 # because it builds only github.com URLs from the value.
@@ -808,9 +825,9 @@ checked_repo() {
 	misdrawn="$(LC_ALL=C; misdrawn_characters "$REPO")"
 	[ -z "$misdrawn" ] ||
 		die "OPENREPOSHAPE_REPO holds a character a terminal draws as something other than itself ($misdrawn): one that is invisible, is drawn as another character, ends a line or reorders the text around it. This tool will not put it on a \`git clone\` command line. Set it to a URL or a path."
-	case "$url$REPO" in
-	-* | *[![:print:]]*) die "OPENREPOSHAPE_REPO is not a value this tool will put on a \`git clone\` command line: it starts with '-' or carries a control character. Set it to a URL or a path." ;;
-	esac
+	if (LC_ALL=C; case "$url$REPO" in -* | *[[:cntrl:]]* | *$'\302'[$'\200'-$'\237']*) exit 0 ;; *) exit 1 ;; esac); then
+		die "OPENREPOSHAPE_REPO is not a value this tool will put on a \`git clone\` command line: it starts with '-' or carries a control character. Set it to a URL or a path."
+	fi
 }
 
 # A CHECKOUT OF THE STANDARD, made the way `setup.sh` self-bootstraps one:
@@ -874,9 +891,9 @@ shape_checkout() {
 # A CHARACTER A TERMINAL DRAWS AS SOMETHING OTHER THAN ITSELF (#837) is a
 # fault between the length and the control character, named by its code
 # point and never echoed, as `checked_repo` refuses one: before the
-# control check, whose `[![:print:]]` is the locale's, so the words do not
-# change with the locale, and before the shape, which echoes the value.
-# Those two are `ref_character_fault`'s, below.
+# control check, which matches none of these characters in any locale
+# (#940), and before the shape, which echoes the value. Those two are
+# `ref_character_fault`'s, below.
 #
 # THE SHAPE IS GIT'S RULE PART BY PART (#874), as `setup.sh`'s and
 # `install.sh`'s are: no `..`, no empty part between slashes, no part that
@@ -913,16 +930,23 @@ ref_fault() {
 # as it reads, and an install that does not record the ref takes one that
 # names no ref at all (workBenches', `record_ref`), so those two are left to
 # where a ref is used.
+#
+# THE CONTROL CHARACTER IS `checked_repo`'s RULE, #760's bytes under
+# `LC_ALL=C`, with its trade (#940): a byte that is not UTF-8, as a lone
+# `0x9B` or an overlong form of a control, passes. A ref with a letter past
+# ASCII (`café`) carries no control character in any locale, and is the
+# shape's to refuse where the ref is used, in the shape's words, as a UTF-8
+# locale had it.
 ref_character_fault() {
 	local ref="$1" what="$2" misdrawn
 	misdrawn="$(LC_ALL=C; misdrawn_characters "$ref")"
 	if [ -n "$misdrawn" ]; then
 		printf '%s' "$what holds a character a terminal draws as something other than itself ($misdrawn): one that is invisible, is drawn as another character, ends a line or reorders the text around it. No branch, tag or commit this tool will pass to \`git checkout\` holds one."
 		return 0
 	fi
-	case "$ref" in
-	*[![:print:]]*) printf '%s' "$what carries a control character, which is not a branch, tag or commit this tool will pass to \`git checkout\`." ;;
-	esac
+	if (LC_ALL=C; case "$ref" in *[[:cntrl:]]* | *$'\302'[$'\200'-$'\237']*) exit 0 ;; *) exit 1 ;; esac); then
+		printf '%s' "$what carries a control character, which is not a branch, tag or commit this tool will pass to \`git checkout\`."
+	fi
 }
 
 checked_ref() {