From: Lady Date: Sun, 19 May 2024 19:54:20 +0000 (-0400) Subject: Remove sed range expressions X-Git-Tag: 0.8.3~2 X-Git-Url: https://git.ladys.computer/Shushe/commitdiff_plain/991cf40919b58de83c8695ef344a1cefa4a8ccbb?ds=inline Remove sed range expressions These are technically only Posix in the Posix locale and have undefined meaning otherwise. It’s not the policy of ⛩️📰 书社 to require the Posix locale, so the safest thing is to just not use range expressions here. (Actually, this policy might be worth revisiting for things which definitely need to be operating on Unicode text.) --- diff --git a/GNUmakefile b/GNUmakefile index b7e1112..6eece6b 100644 --- a/GNUmakefile +++ b/GNUmakefile @@ -309,7 +309,7 @@ override diffprereqs = $(if $(subst $(shell $(CAT) $(call quote,$(BUILDDIR)/last override sedesc = $(subst /,[/],$(subst $$,\$$,$(subst *,\*,$(subst .,\.,$(subst [,\[,$(subst ^,\^,$(subst \,\\,$1))))))) # The command to use for percent‐decoding. -override perdeccmd := $(SED) 's/|/%7C/g;s/[\]/%5C/g;s/%[0-9A-Fa-f]\{2\}/|&|/g' | $(TR) '|' '\n' | $(SED) '/^%[0-9A-Fa-f]\{2\}$$/!s/%/|%25|/' | $(TR) '|' '\n' | $(AWK) '$$0!~/%/{printf "%s",$$0}/%/{d="0123456789ABCDEF";v=substr(toupper($$0),2,2);printf "\\%04o",(index(d,substr(v,1,1))-1)*16+index(d,substr(v,2,1))-1}' | $(xargsquote) | $(XARGS) -E '' $(PRINTF) '%b' +override perdeccmd := $(SED) 's/|/%7C/g;s/[\]/%5C/g;s/%[0123456789ABCDEFabcdef]\{2\}/|&|/g' | $(TR) '|' '\n' | $(SED) '/^%[0123456789ABCDEFabcdef]\{2\}$$/!s/%/|%25|/' | $(TR) '|' '\n' | $(AWK) '$$0!~/%/{printf "%s",$$0}/%/{d="0123456789ABCDEF";v=substr(toupper($$0),2,2);printf "\\%04o",(index(d,substr(v,1,1))-1)*16+index(d,substr(v,2,1))-1}' | $(xargsquote) | $(XARGS) -E '' $(PRINTF) '%b' # (callable) Percent‐decode the given strings. # @@ -324,7 +324,7 @@ override perdec = $(shell $(PRINTF) '%s\0450A' $(foreach encoded,$1,$(call quote # It is assumed that the given strings do not contain newlines. # # ☡ This variable creates a subshell every time it is computed. -override perenc = $(shell $(PRINTF) '%s\n' $(foreach unencoded,$1,$(call quote,$(unencoded))) | $(OD) -t x1 | $(SED) 's/^[0-9]*//;s/0[Aa]/%&/g;s/2[146-9A-Ea-e]/%&/g;s/3[0-9AaBbDd]/%&/g;s/40/%&/g;s/[46][1-9A-Fa-f]/%&/g;s/[57][0-9Aa]/%&/g;s/5[Ff]/%&/g;s/7[Ee]/%&/g;s/[ ][0-9A-Fa-f]\{2\}/%25&/g' | $(TR) -d ' \n' | $(TR) 'abcdef' 'ABCDEF' | $(perdeccmd)) +override perenc = $(shell $(PRINTF) '%s\n' $(foreach unencoded,$1,$(call quote,$(unencoded))) | $(OD) -t x1 | $(SED) 's/^[0123456789]*//;s/0[Aa]/%&/g;s/2[146789ABCDEabcde]/%&/g;s/3[0123456789AaBbDd]/%&/g;s/40/%&/g;s/[46][123456789ABCDEFabcdef]/%&/g;s/[57][0123456789Aa]/%&/g;s/5[Ff]/%&/g;s/7[Ee]/%&/g;s/[ ][0123456789ABCDEFabcdef]\{2\}/%25&/g' | $(TR) -d ' \n' | $(TR) 'abcdef' 'ABCDEF' | $(perdeccmd)) # (callable) Percent‐encode each component in the given paths. # @@ -472,14 +472,14 @@ endif override id = $(XMLLINT) --xpath '/*/*[local-name()="id" and namespace-uri()="urn:fdc:ladys.computer:20231231:Shu1She4"]/text()[1]' $(call quote,$1) 2> /dev/null || $(PRINTF) '%s\n' $(call quote,about:shushe?$(or $2,unknown)=$(call pathenc,$(basename $(notdir $1)))) # (callable) Sanitize and wrap the provided plaintext file in X·M·L, printing to `stdout´. -override wrapplaintext = { $(PRINTF) '%s\n%s' '' ''; } +override wrapplaintext = { $(PRINTF) '%s\n%s' '' ''; } # (callable) Check if the provided X·M·L file is X·M·L 1.1, and if so, coerce to X·M·L 1.0 as best as possible, printing the result (or the original file contents) to `stdout´. # # The X·M·L declaration will be dropped and character escapes for C0 control codes will be replaced with a literal `U+0091 PRIVATE USE ONE´, which is invalid in X·M·L 1.1, but valid X·M·L 1.0 (making the replacement obvious). # # This isn’t a perfect substitution (it makes some assumptions about the format of the underlying X·M·L), but it should be workable for most sensible, welformed files. -override serializexml = $(SED) "$$($(PRINTF) '%b' '/]*?>//\n s/&\0043x0*[1-8BCEFbcef];/\0302\0221/g\n s/&\0043x0*1[0-9A-Fa-f];/\0302\0221/g\n s/&\00430*[1-8];/\0302\0221/g\n s/&\00430*1[124-9];/\0302\0221/g\n s/&\00430*2[0-9];/\0302\0221/g\n s/&\00430*3[01];/\0302\0221/g\n}')" < $(call quote,$1) | $(SED) "$$($(PRINTF) '%b' ':a\n/^\\n*$$/{ $$d\n N\n ba\n}')" +override serializexml = $(SED) "$$($(PRINTF) '%b' '/]*?>//\n s/&\0043x0*[12345678BCEFbcef];/\0302\0221/g\n s/&\0043x0*1[0123456789ABCDEFabcdef];/\0302\0221/g\n s/&\00430*[12345678];/\0302\0221/g\n s/&\00430*1[12456789];/\0302\0221/g\n s/&\00430*2[0123456789];/\0302\0221/g\n s/&\00430*3[01];/\0302\0221/g\n}')" < $(call quote,$1) | $(SED) "$$($(PRINTF) '%b' ':a\n/^\\n*$$/{ $$d\n N\n ba\n}')" # ─ ¶ Phony Targets ───────────────────────────────────────────────────