Percent‐encode filenames when generating u·r·i’s

It’s not known or expected whether tools like `xmlcatalog` can handle
full leiris. It’s better and safer to just only use u·r·i’s for
identifying resources.

Note that this _does_ have implications on includes (they must also be
percent‐encoded). Ideally, it would be possible to run this conversion
in the transforms, but this probably is not possible in X·S·L·T 1.0.

diff --git a/GNUmakefile b/GNUmakefile
index b306ff25ebdd85fc258a14542a75c36e69de645f..0b644c2d2e06ec369e77ba39b618ebeb032340b4 100644 (file)
--- a/GNUmakefile
+++ b/GNUmakefile
@@ -14,6 +14,7 @@ override define makefileinfo
 ║│ version incompatibilities. The full list of program        │║
 ║│ requirements is as follows :—                             │║
 ║│                                                            │║
+║│ • awk                                                      │║
 ║│ • cat                                                      │║
 ║│ • cp                                                       │║
 ║│ • date                                                     │║
@@ -22,6 +23,7 @@ override define makefileinfo
 ║│ • find                                                     │║
 ║│ • mkdir (requires support for `-p´)                        │║
 ║│ • mv                                                       │║
+║│ • od (requires support for `-t x1´)                        │║
 ║│ • printf                                                   │║
 ║│ • rm                                                       │║
 ║│ • sed                                                      │║
@@ -31,6 +33,7 @@ override define makefileinfo
 ║│ • touch                                                    │║
 ║│ • tr (requires support for `-d´)                           │║
 ║│ • uuencode (requires support for `-m´ and `-r´)            │║
+║│ • xargs (requires support for `-0´)                        │║
 ║│ • xmlcatalog (provided by libxml2)                         │║
 ║│ • xmllint (provided by libxml2)                            │║
 ║│ • xsltproc (provided by libxslt)                           │║
@@ -79,6 +82,7 @@ endef
 #
 # If these are not installed on your computer, or you need to use a
 # different implementation, you can override the appropriate variable.
+AWK := awk
 CAT := cat
 CP := cp
 DATE := date
@@ -87,6 +91,7 @@ FILE := file
 FIND := find
 MKDIR := mkdir
 MV := mv
+OD := od
 PRINTF := printf
 RM := rm
 SED := sed
@@ -96,6 +101,7 @@ TEST := test
 TOUCH := touch
 TR := tr
 UUENCODE := uuencode
+XARGS := xargs
 XMLCATALOG := xmlcatalog
 XMLLINT := xmllint
 XSLTPROC := xsltproc
@@ -203,6 +209,21 @@ override silent := $(if $(VERBOSE),,@)
 # expressions.
 override sedesc = $(subst /,[/],$(subst $$,\$$,$(subst *,\*,$(subst .,\.,$(subst [,\[,$(subst ^,\^,$(subst \,\\,$1)))))))
 
+# (callable) Percent‐decode the given strings.
+override perdec = $(foreach encoded,$1,$(shell $(PRINTF) '%s\n' $(call quote,$(encoded)) | $(SED) 's/`/`%60`/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",$$0)}/%/{sub("%","0x");printf("\\%04o",$$0)}' | $(XARGS) -0 $(PRINTF) '%b'))
+
+# (callable) Percent‐encode the given strings.
+#
+# This singly‐encodes u·r·i characters and doubly‐encodes other
+# characters, then calls `perdec´ to decode back to a single encoding.
+# The encoding assumes the input is a “u·r·i component”; e·g that the
+# resulting string should only contain `pchar´ (but can contain any
+# `sub-delim´).
+override perenc = $(foreach unencoded,$1,$(call perdec,$(shell $(PRINTF) '%s' $(call quote,$(unencoded)) | $(OD) -t x1 | $(SED) 's/^[0-9]*//;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/7[Ee]/%&/g;s/[ ][0-9A-Fa-f]\{2\}/%25&/g' | $(TR) -d ' \n' | $(TR) 'abcdef' 'ABCDEF')))
+
+# (callable) Percent‐encode each component in the given paths.
+override pathenc = $(foreach path,$1,$(subst $(space),/,$(call perenc,$(subst /, ,$(path)))))
+
 # (overridable) Collect all of the applicable includes from the
 # includes directory.
 sourceincludes := $(shell $(FIND) $(FINDOPTS) $(INCLUDEDIR) -type f '(' $(FINDRULES) ')')
@@ -234,7 +255,7 @@ override assetfiles := $(filter-out $(xmlfiles) $(plaintextfiles),$(sourcefiles)
 override typeoffile = $(patsubst $(foreach file,$1,$(file):%),%,$(filter $(foreach file,$1,$(file):%),$(types)))
 
 # (callable) Get the identifier for the given transform.
-override id = $(or $(shell $(XMLLINT) --xpath '/*/*[local-name()="id" and namespace-uri()="urn:fdc:ladys.computer:20231231:Shu1She4"]/text()[1]' $1),$(basename $(notdir $1)))
+override id = $(or $(shell $(XMLLINT) --xpath '/*/*[local-name()="id" and namespace-uri()="urn:fdc:ladys.computer:20231231:Shu1She4"]/text()[1]' $1),example:$(call pathenc,$(basename $(notdir $1))))
 
 # (callable) Get the local path for the given source file.
 override sourcepath = $(firstword $(foreach directory,$(SRCDIR),$(if $(filter $(directory)/%,$1),$(patsubst $(directory)/%,%,$1),)))
@@ -246,7 +267,7 @@ override includepath = $(firstword $(foreach directory,$(INCLUDEDIR),$(if $(filt
 override datauri = $(foreach file,$1,data:$(call typeoffile,$(file));base64,$(shell $(UUENCODE) -m -r $(call quote,$(file)) _ | tr -d ' \n'))
 
 # Pair each source file and include with its local u·r·i.
-override sourcelocalpair = $(foreach file,$(sourcefiles) $(sourceincludes),$(if $(filter $(file),$(sourceincludes)),$(file):about:shushe?include=$(call includepath,$(file)),$(file):about:shushe?source=$(call sourcepath,$(file))))
+override sourcelocalpair = $(foreach file,$(sourcefiles) $(sourceincludes),$(if $(filter $(file),$(sourceincludes)),$(file):about:shushe?include=$(call pathenc,$(call includepath,$(file))),$(file):about:shushe?source=$(call pathenc,$(call sourcepath,$(file)))))
 
 # (callable) Get local uris for the given files.
 override localuri = $(foreach file,$1,$(patsubst $(file):%,%,$(filter $(file):%,$(sourcelocalpair))))
@@ -410,7 +431,7 @@ $(BUILDDIR)/magic.mgc: $(wildcard $(MAGICDIR)/*)
 $(BUILDDIR)/parser.catalog: $(PARSERS)
        @$(ECHO) "Generating catalog of parsers…"
        $(silent)$(XMLCATALOG) --create --noout $(call quote,$@)
-       $(foreach parser,$(PARSERS),$(silent)$(XMLCATALOG) --add uri $(call quote,$(call id,$(parser))) $(call quote,../$(parser)) --noout $(call quote,$@)$(newline))
+       $(foreach parser,$(PARSERS),$(silent)$(XMLCATALOG) --add uri $(call quote,$(call id,$(parser))) $(call quote,$(call pathenc,../$(parser))) --noout $(call quote,$@)$(newline))
 $(BUILDDIR)/parser.xslt: $(BUILDDIR)/parser.catalog $(THISDIR)/lib/catalog2parser.xslt
        @$(ECHO) "Generating main parser…"
        $(silent)$(XSLTPROC) -o $(call quote,$@) $(call quote,$(THISDIR)/lib/catalog2parser.xslt) $(call quote,$<)
@@ -425,14 +446,14 @@ $(call parsed,$(sourcefiles) $(sourceincludes)): %: $$(call unparsed,$$@) $(type
        $(silent)$(call ensuredirectory,$(dir $@))
        $(silent)$(if $(filter $<,$(assetfiles)),$(PRINTF) '%s\n' $(call quote,<object xmlns="http://www.w3.org/1999/xhtml" type="$(call typeoffile,$<)" data="$(call datauri,$<)"/>) > $(call quote,$@),$(if $(filter $<,$(plaintextfiles)),$(call wrapplaintext,$<),$(CAT) $(call quote,$<)) | $(XSLTPROC) -o $(call quote,$@) $(call quote,$(BUILDDIR)/parser.xslt) -)
 
-# Generate a catalog of all transformed files, for use when processing
+# Generate a catalog of all parsed files, for use when processing
 # includes. This does not depend on actually transforming the files.
 $(BUILDDIR)/catalog: $(sourcefiles) $(sourceincludes) $(typeupdates)
        @$(ECHO) "Generating catalog of parsed files…"
        $(silent)$(XMLCATALOG) --create --noout $(call quote,$@)
-       $(foreach source,$(sourcefiles) $(sourceincludes),$(silent)$(XMLCATALOG) --add uri $(call quote,$(call localuri,$(source))) $(call quote,$(patsubst $(BUILDDIR)/%,%,$(call parsed,$(source)))#$(if $(filter $(source),$(assetfiles)),asset,xml)) --noout $(call quote,$@)$(newline))
+       $(foreach source,$(sourcefiles) $(sourceincludes),$(silent)$(XMLCATALOG) --add uri $(call quote,$(call localuri,$(source))) $(call quote,$(call pathenc,$(patsubst $(BUILDDIR)/%,%,$(call parsed,$(source))))#$(if $(filter $(source),$(assetfiles)),asset,xml)) --noout $(call quote,$@)$(newline))
 
-# Build a list of dependencies for each transformed file.
+# Build a list of dependencies for each parsed file.
 $(BUILDDIR)/dependencies: $(BUILDDIR)/catalog $(call parsed,$(plaintextfiles) $(xmlfiles)) $(THISDIR)/lib/catalog2dependencies.xslt
        @$(ECHO) "Identifying dependencies…"
        $(silent)$(XSLTPROC) -o $(call quote,$@) $(call quote,$(THISDIR)/lib/catalog2dependencies.xslt) $(call quote,$<)
@@ -441,7 +462,7 @@ $(BUILDDIR)/dependencies: $(BUILDDIR)/catalog $(call parsed,$(plaintextfiles) $(
 $(BUILDDIR)/transform.catalog: $(TRANSFORMS)
        @$(ECHO) "Generating catalog of transforms…"
        $(silent)$(XMLCATALOG) --create --noout $(call quote,$@)
-       $(foreach transform,$(TRANSFORMS),$(silent)$(XMLCATALOG) --add uri $(call quote,$(call id,$(transform))) $(call quote,../$(transform)) --noout $(call quote,$@)$(newline))
+       $(foreach transform,$(TRANSFORMS),$(silent)$(XMLCATALOG) --add uri $(call quote,$(call id,$(transform))) $(call quote,$(call pathenc,../$(transform))) --noout $(call quote,$@)$(newline))
 $(BUILDDIR)/transform.xslt: $(BUILDDIR)/transform.catalog $(THISDIR)/lib/catalog2transform.xslt
        @$(ECHO) "Generating main transform…"
        $(silent)$(XSLTPROC) -o $(call quote,$@) $(call quote,$(THISDIR)/lib/catalog2transform.xslt) $(call quote,$<)

diff --git a/README.markdown b/README.markdown
index 8e56f29ac40561d77954a326db73c763f4f1654c..dec092a5d7c59905bf4070c1bae96f8fa9febff4 100644 (file)
--- a/README.markdown
+++ b/README.markdown
@@ -91,6 +91,7 @@ In every case, you may supply your own implementation by overriding the
   corresponding (allcaps) variable (e·g, set `MKDIR` to supply your own
   `mkdir` implementation).
 
+- `awk`
 - `cat`
 - `cp`
 - `date`
@@ -99,6 +100,7 @@ In every case, you may supply your own implementation by overriding the
 - `find`
 - `mkdir` (requires support for `-p`)
 - `mv`
+- `od` (requires support for `-t x1`)
 - `printf`
 - `rm`
 - `sed`
@@ -108,6 +110,7 @@ In every case, you may supply your own implementation by overriding the
 - `touch`
 - `tr` (requires support for `-d`)
 - `uuencode` (requires support for `-m` and `-r`)
+- `xargs` (requires support for `-0`)
 - `xmlcatalog` (provided by `libxml2`)
 - `xmllint` (provided by `libxml2`)
 - `xsltproc` (provided by `libxslt`)