[bug#60358] [PATCH v9 3/4] gnu: Add gnulib.

[Liliana Marie Prikler]

Am Dienstag, dem 27.12.2022 um 17:23 +0100 schrieb Vivien Kraus:
> * gnu/packages/build-tools.scm (gnulib-checkout, gnulib): New
> variables
Missing the dot :)
> ---
>  gnu/packages/build-tools.scm | 201
> +++++++++++++++++++++++++++++++++++
>  1 file changed, 201 insertions(+)
> 
> diff --git a/gnu/packages/build-tools.scm b/gnu/packages/build-
> tools.scm
> index 6c1350c44f..3e6a16f3f5 100644
> --- a/gnu/packages/build-tools.scm
> +++ b/gnu/packages/build-tools.scm
> @@ -32,27 +32,37 @@
>  ;;; along with GNU Guix.  If not, see
> <http://www.gnu.org/licenses/>.
>  
>  (define-module (gnu packages build-tools)
> +  #:use-module (ice-9 optargs)
>    #:use-module ((guix licenses) #:prefix license:)
>    #:use-module (guix utils)
>    #:use-module (guix packages)
>    #:use-module (guix gexp)
>    #:use-module (guix download)
>    #:use-module (guix git-download)
> +  #:use-module (guix git)
>    #:use-module (guix build-system cmake)
> +  #:use-module (guix build-system copy)
> +  #:use-module (guix modules)
>    #:use-module (gnu packages)
>    #:use-module (gnu packages adns)
> +  #:use-module (gnu packages autotools)
>    #:use-module (gnu packages base)
>    #:use-module (gnu packages bash)
>    #:use-module (gnu packages check)
> +  #:use-module (gnu packages code)
>    #:use-module (gnu packages compression)
>    #:use-module (gnu packages cpp)
> +  #:use-module (gnu packages cppi)
>    #:use-module (gnu packages elf)
> +  #:use-module (gnu packages gcc)
>    #:use-module (gnu packages linux)
> +  #:use-module (gnu packages lisp)
>    #:use-module (gnu packages logging)
>    #:use-module (gnu packages lua)
>    #:use-module (gnu packages ninja)
>    #:use-module (gnu packages package-management)
>    #:use-module (gnu packages pcre)
> +  #:use-module (gnu packages perl)
>    #:use-module (gnu packages pkg-config)
>    #:use-module (gnu packages pretty-print)
>    #:use-module (gnu packages protobuf)
> @@ -65,6 +75,7 @@ (define-module (gnu packages build-tools)
>    #:use-module (gnu packages rpc)
>    #:use-module (gnu packages sqlite)
>    #:use-module (gnu packages tls)
> +  #:use-module (gnu packages unicode)
>    #:use-module (gnu packages version-control)
>    #:use-module (guix build-system gnu)
>    #:use-module (guix build-system python))
> @@ -803,3 +814,193 @@ (define-public genie
>  same settings to multiple projects.  It supports generating projects
> using GNU
>  Makefiles, JSON Compilation Database, and experimentally Ninja.")
>        (license license:bsd-3))))
> +
> +(define*-public (gnulib-checkout #:key
> +                                 version
> +                                 (revision "1")
> +                                 commit
> +                                 hash)
> +  "Return as a package the exact gnulib checkout."
> +  (package
> +    (name "gnulib")
> +    (version (git-version version revision commit))
> +    (source
> +     (origin
> +       (method git-fetch)
> +       (uri (git-reference
> +             (url "https://git.savannah.gnu.org/git/gnulib.git/";)
> +             (commit commit)))
> +       (file-name (git-file-name name version))
> +       (sha256 hash)
> +       (snippet
> +        (with-imported-modules (source-module-closure '((guix build
> utils)))
> +          #~(begin
> +              (use-modules (guix build utils)
> +                           (ice-9 ftw)
> +                           (ice-9 rdelim))
> +              ;; .c, .h and .gperf files whose first line is /* DO
> NOT EDIT!
> +              ;; GENERATED AUTOMATICALLY! */ are generated
> automatically based
> +              ;; on the unicode database. Since we replace the
> unicode
> +              ;; database with our own, we need to regenerate them.
> So, they
> +              ;; are removed from the source. They are sprinkled all
> over the
> +              ;; place unfortunately, so we can’t exclude whole
> directories.
> +              (let ((generated-automatically?
> +                     (lambda (filename . unused)
> +                       (and (or (string-suffix? ".c" filename)
> +                                (string-suffix? ".h" filename)
> +                                (string-suffix? ".gperf" filename))
> +                            (call-with-input-file filename
> +                              (lambda (port)
> +                                (let ((first-line (read-line port)))
> +                                  (equal?
> +                                   first-line
> +                                   "/* DO NOT EDIT! GENERATED
> AUTOMATICALLY! */"))))))))
> +                (for-each delete-file (find-files (getcwd)
> generated-automatically?)))
> +              ;; Other files are copied from UCD.
> +              (for-each delete-file
> +                        '("tests/unigbrk/GraphemeBreakTest.txt"
> +                          "tests/uninorm/NormalizationTest.txt"
> +                          "tests/uniname/UnicodeData.txt"
> +                          "tests/uniname/NameAliases.txt"
> +                          ;; FIXME:
> tests/uniname/HangulSyllableNames.txt
> +                          ;; seems like a UCD file but it is not
> distributed
> +                          ;; with UCD.
> +                          "tests/uniwbrk/WordBreakTest.txt")))))))
> +    (build-system copy-build-system)
> +    (arguments
> +     (list
> +      #:install-plan
> +      #~'(("./gnulib-tool" "bin/")
> +          ("." "src/gnulib" #:exclude-regexp ("\\.git.*")))
> +      #:modules '((ice-9 match)
> +                  (guix build utils)
> +                  (guix build copy-build-system)
> +                  ((guix build gnu-build-system) #:prefix gnu:))
> +      #:phases
> +      #~(modify-phases %standard-phases
> +          (add-before 'install 'gnu:check
> +            (assoc-ref gnu:%standard-phases 'check)
> )
Use just "check" for the phase name.
> +          (add-before 'gnu:check 'fix-tests
> +            (lambda _
> +              (substitute* "Makefile"
> +                (("-f maint.mk syntax-check")
> +                 "_gl-Makefile=yes -f maint.mk syntax-check"))
> +              (invoke "git" "init")
> +              (invoke "git" "config" "user.name" "Guix")
> +              (invoke "git" "config" "user.email" "guix@localhost")
> +              (invoke "git" "add" ".")
> +              ;; Syntax checks are only run against committed files.
> +              (invoke "git" "commit" "-m" "Prepare for tests.")))
> +          (add-before 'gnu:check 'disable-failing-tests
> +            (lambda _
> +              (substitute* "cfg.mk"
> +                (("local-checks-to-skip =")
> +                 ;; sc_copyright_check fails because the fake commit
> date may
> +                 ;; be later than the copyright year.
> +                 "local-checks-to-skip = \\
> +  sc_Wundef_boolean \\
> +  sc_copyright_check \\
> +  sc_file_system \\
> +  sc_indent \\
> +  sc_keep_gnulib_texi_files_mostly_ascii \\
> +  sc_prohibit_assert_without_use \\
> +  sc_prohibit_close_stream_without_use \\
> +  sc_prohibit_defined_have_decl_tests \\
> +  sc_prohibit_doubled_word \\
> +  sc_prohibit_empty_lines_at_EOF \\
> +  sc_prohibit_intprops_without_use \\
> +  sc_prohibit_openat_without_use \\
> +  sc_prohibit_test_minus_ao \\
> +  sc_unportable_grep_q"))
Can we provide these via #:make-flags (assuming they get forwarded
through copy-build-system)?
> +              (substitute* "Makefile"
> +                (("sc_(\
> +check_(sym_list|copyright)\
> +|cpp_indent_check\
> +|prohibit_(AC_LIBOBJ_in_m4|leading_TABs)\
> +|prefer_ac_check_funcs_once)" rule)
I do think splitting this into 4 separate rules is a comfy middle
ground between obivously splitting a regexp over multiple lines and
listing 10 rules separately.
> +                 (string-append "disabled_" rule)))))
> +          (add-before 'gnu:check 'regenerate-unicode
> +            (lambda* (#:key inputs #:allow-other-keys)
> +              (define (find-ucd-file name)
> +                (search-input-file inputs (string-append
> "share/ucd/" name)))
> +              (define (find-ucd-files . names)
> +                (map find-ucd-file names))
> +              (with-directory-excursion "lib"
> +                ;; See the compile-command buffer-local variable in
> +                ;; lib/gen-uni-tables.c
> +                (invoke #+(file-append gcc "/bin/gcc") "-O" "-Wall"
> "gen-uni-tables.c"
> +                        "-Iunictype" "-o" "gen-uni-tables")
> +                (apply invoke
> +                       "./gen-uni-tables"
> +                       (append
> +                        (find-ucd-files "UnicodeData.txt"
> +                                        "PropList.txt"
> +                                        "DerivedCoreProperties.txt"
> +                                        "emoji/emoji-data.txt"
> +                                        "ArabicShaping.txt"
> +                                        "Scripts.txt"
> +                                        "Blocks.txt")
> +                        (list
> +                         #$(origin
> +                             (method url-fetch)
> +                             (uri (string-append
> +                                   "https://www.unicode.org/Public/";
> +                                   "3.0-Update1/PropList-
> 3.0.1.txt"))
> +                             (sha256
> +                              (base32
> +                              
> "0k6wyijyzdl5g3nibcwfm898kfydx1pqaz28v7fdvnzdvd5fz7lh"))))
> +                        (find-ucd-files "EastAsianWidth.txt"
> +                                        "LineBreak.txt"
> +                                       
> "auxiliary/WordBreakProperty.txt"
> +                                       
> "auxiliary/GraphemeBreakProperty.txt"
> +                                        "CompositionExclusions.txt"
> +                                        "SpecialCasing.txt"
> +                                        "CaseFolding.txt")
> +                         (list #$(package-version (this-package-
> native-input "ucd")))))
> +                (invoke "clisp" "-C" "uniname/gen-uninames.lisp"
> +                        (search-input-file inputs
> "share/ucd/UnicodeData.txt")
> +                        "uniname/uninames.h"
> +                        (search-input-file inputs
> "share/ucd/NameAliases.txt"))
Why not (find-ucd-file "UnicodeData.txt") and (find-ucd-file
"NameAliases.txt") respectively?
> +                (for-each
> +                 (match-lambda
> +                  ((ucd-file . directory)
> +                   (copy-file (find-ucd-file ucd-file)
> +                              (string-append "../tests/" directory
> "/"
> +                                             (basename ucd-file)))))
> +                 '(("NameAliases.txt" . "uniname")
> +                   ("UnicodeData.txt" . "uniname")
> +                   ("NormalizationTest.txt" . "uninorm")
> +                   ("auxiliary/GraphemeBreakTest.txt" . "unigbrk")
> +                   ("auxiliary/WordBreakTest.txt" . "uniwbrk")))
LGTM.

Cheers