[Reproducible-builds] Bug#806891: Multi-file HTML output
Package: diffoscope Version: 42 Severity: wishlist Tags: patch -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, in the branch "html-dir" in the main repository, you will find the following feature: Multi-file HTML output in order to make it feasible to create HTML output with large diffs, this adds a new presenter mode (--html-dir) which spreads the output over multiple files in a directory. In particular, the table presenting a diff that is larger than the number of bytes specified via --separate-file-diff-size (200kB by default) is writen to a separate file and loaded on demand using JavaScript/JQuery. By default, JQuery is symlinked into the output directory from /usr/share/javascript/jquery/jquery.js (install libjs-jquery!), but an alternative location (e.g. /javascript/jquery/jquery.js) can be specified using the --jquery command line parameter. http://anonscm.debian.org/cgit/reproducible/diffoscope.git/log/?h=html-dir Please review, and if you like it, merge it and remove the html-dir branch. Greetings, Joachim - -- System Information: Debian Release: stretch/sid APT prefers buildd-unstable APT policy: (500, 'buildd-unstable'), (500, 'unstable'), (101, 'experimental') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.2.0-1-amd64 (SMP w/4 CPU cores) Locale: LANG=de_DE.utf8, LC_CTYPE=de_DE.utf8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages diffoscope depends on: ii python3-libarchive-c 2.1-3 ii python3-magic 1:5.25-2 ii python3-pkg-resources 18.4-2 pn python3:any Versions of packages diffoscope recommends: ii acl2.2.52-2 pn binutils-multiarch ii bzip2 1.0.6-8 ii cpio 2.11+dfsg-4.1 ii default-jdk [java-sdk] 2:1.7-52.1 ii fontforge-extras 0.3-4 ii fp-utils 2.6.4+dfsg-8 ii fp-utils-2.6.4 [fp-utils] 2.6.4+dfsg-8 ii gcj-4.8-jdk [java-sdk] 4.8.5-2 ii gcj-4.9-jdk [java-sdk] 4.9.3-7 ii gcj-5-jdk [java-sdk] 5.2.1-23 ii gcj-jdk [java-sdk] 4:5.2.1-6 ii genisoimage9:1.1.11-3 ii gettext0.19.6-1 ii ghc7.8.4-9 ii gnupg 1.4.19-6 ii mono-utils 3.2.8+dfsg-10 ii openjdk-6-jdk [java-sdk] 6b37-1.13.9-1 ii openjdk-7-jdk [java-sdk] 7u91-2.6.3-2 ii pdftk 2.02-3 ii poppler-utils 0.26.5-4 ii python3-debian 0.1.27 ii python3-rpm4.12.0.1+dfsg1-3+b1 ii python3-tlsh 3.4.1+20151014-1 ii rpm2cpio 4.12.0.1+dfsg1-3+b1 ii sng1.0.6-2 ii sqlite33.9.2-1 ii squashfs-tools 1:4.3-3 ii unzip 6.0-20 ii vim-common 2:7.4.826-1+b1 ii xz-utils 5.1.1alpha+20120614-2.1 diffoscope suggests no packages. - -- no debconf information -BEGIN PGP SIGNATURE- Version: GnuPG v1 iEYEARECAAYFAlZfCmIACgkQ9ijrk0dDIGxJRgCfZzj7cMjbL+3kTBMs/fTjUcer 9Q8An0cIY9ylDdmHNKKQ818UdXAFgliN =YIRi -END PGP SIGNATURE- ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
[Reproducible-builds] Bug#806891: [Reproducible-commits] [diffoscope] 01/01: Multi-file HTML output
Hi,
Am Mittwoch, den 02.12.2015, 17:32 +0200 schrieb Jérémy Bobbio:
>
> > +parser.add_argument('--jquery', metavar='url', dest='jquery_url',
> > +help='link to the jquery url, with --html-dir. By
> > default, a symlink to /usr/share/javascript/jquery/jquery.js is created')
>
> To make the Suggests a reality, I think it would be better to only
> display “By default, a symlink…” if the file is already present on the
> filesystem. Otherwise, if `--html-dir` is specified, the software should
> exit with an error, asking users to specify `--jquery`.
You want --help output that depends on the system state?
> Related question: is the browser going to visit the subpage if I have
> JavaScript disabled? In that case, one other option is to only add the
> JavaScript code when we have access to jQuery. I know some people don't
> want JavaScript in their browser, so specifying something like
> `--no-jquery` or `--jquery=none` could also be a way to turn it off.
See my other mail, I can make it degrade gracefully. Those people will
have JavaScript disabled, so I don’t see a strong usecase for multi-
page html output with JavaScript completely disabled. Note that --html
is an option for those users.
> > +# no output desired? print text
> > +if not parsed_args.text_output and not
> > parsed_args.html_output and not parsed_args.html_output_directory:
> > +parsed_args.text_output = "-"
>
> Maybe it would be nicer to write:
>
> if not any(parsed_args.text_output, parsed_args.html_output,
> parsed_args.html_output_directory):
✓
> > +def output_unified_diff(print_func, directory, anchor,
> > unified_diff):
> > +if directory and len(unified_diff) >
> > Config.general.separate_file_diff_size:
> > +# open a new file for this table
> > +filename="%s.html" % hashlib.md5(anchor.encode('utf-
> > 8')).hexdigest()
>
> I'm not entirely sure the anchor as it's working right now will be
> unique…
Ok, I’ll ditch this and simply hash the unified_diff: There, clashes
will not matter.
> > +logger.debug('separate html output for diff of %s (size
> > %d)', anchor, len(unified_diff))
> > +with file_printer(directory, filename) as new_print_func:
> > +output_unified_diff_table(new_print_func,
> > unified_diff)
>
> So I think it would be great to crash here instead of overwrite if the file
> aleardy exists. What do you think?
Not with the above change.
Also, I want to be able to re-run diffoscope without having to manually
delete the output directory first.
Greetings,
Joachim
--
Joachim "nomeata" Breitner
Debian Developer
nome...@debian.org | ICQ# 74513189 | GPG-Keyid: F0FBF51F
JID: nome...@joachim-breitner.de | http://people.debian.org/~nomeata
signature.asc
Description: This is a digitally signed message part
___
Reproducible-builds mailing list
Reproducible-builds@lists.alioth.debian.org
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
[Reproducible-builds] Bug#806891: Bug#806891: Multi-file HTML output
Hi, Am Mittwoch, den 02.12.2015, 17:31 +0200 schrieb Holger Levsen: > On Mittwoch, 2. Dezember 2015, Joachim Breitner wrote: > > In particular, the table presenting a diff that is larger than the > > number of bytes specified via --separate-file-diff-size (200kB by > > default) is writen to a separate file and loaded on demand using > > JavaScript/JQuery. > > just using javascript for this I dislike, I'm very happy that > reproducible.debian.net doesnt need javascript for any functionality at all > (and I would like to use multi-html output there too), so how about adding > another option to create static links (or whatever non-JS solution) instead. the buttons to load the diffs are actually links to the HTML of the table, so it almost degrades gracefully. To make it degrade gracefully fully, I would have to add a proper HTML header around the table, and ignore that when inserting the fragment dynamically. Both is easily possible. Is that ok then? Geetings, Joachim -- Joachim "nomeata" Breitner Debian Developer nome...@debian.org | ICQ# 74513189 | GPG-Keyid: F0FBF51F JID: nome...@joachim-breitner.de | http://people.debian.org/~nomeata signature.asc Description: This is a digitally signed message part ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
[Reproducible-builds] Higher entropy notification mail subjects
[I’m not subscribed, please CC me on replies] Hi, my inbox is full of mails with subject [Pkg-haskell-maintainers] Reproducibility state of package haskell-hfuse change... as you can see the relevant part is cut off here. May I suggest to use a more compact subject, e.g. Reproducibility change of haskell-hfuse: FTBFS → unreproducible or maybe even R13y of haskell-hfuse: FTBFS → unreproducible Greetings, Joachim -- Joachim nomeata Breitner Debian Developer nome...@debian.org | ICQ# 74513189 | GPG-Keyid: F0FBF51F JID: nome...@joachim-breitner.de | http://people.debian.org/~nomeata signature.asc Description: This is a digitally signed message part ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] debug-sym packages?
Am 26. Mai 2015 14:57:13 MESZ, schrieb Jérémy Bobbio lu...@debian.org: I've scheduled another test for the version in experimental, and it now builds reproducibly! :) https://reproducible.debian.net/rb-pkg/experimental/amd64/ghc.html \o/ ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
