11

I would like to add in my long CV (not my short CV) the list of package I maintain in CTAN.

Is there somewhere a file where I could make a request, to:

  • obtain the list of package I am associated
  • have some information for each package : maintenance status, name, short description.

For example an XML file?

7

I have just published an experimental feature on the CTAN portal to retrieve information about the catalogue in XML form. This includes cross references for authors. For instance: http://www.ctan.org/xml/au­thor/knuth?ref=true

The online help provides some details.

  • 1
    A warm welcome! Can you change your user name to something fancier than user43912? ;-) – egreg Jan 15 '14 at 23:24
  • that look exactly what I need ! a great thank ! – Maïeul Jan 16 '14 at 18:10
  • just an other question ? I was thinking that the ctan should be ask by mirror.ctan.org. Is it true for this also ? – Maïeul Jan 16 '14 at 19:59
  • CTAN provides different services. The file download is distributed to the mirrors. Others like searching and access to the catalogue (HTML or XML) are provided by the central server only. – Gerd Neugebauer Jan 16 '14 at 20:33
  • thanks. For your information, I was able to use your file to answer to my problem. Cf contrib.spip.net/Donnees-CTAN-4487 (in french, sorry) – Maïeul Jan 16 '14 at 22:22
17

Great, I should be sleeping by now, but no, I had to code something! :) Somebody please call the authorities.

Here goes my pathetic attempt to provide the raw data you want. Since I had no idea of a good name (namely bird names, of course), I wrote a small application called ctaninfo. Here it comes:

paulo@alexandria ~$ java -jar ctaninfo.jar 

And then we have:

Quack 1

The author ID is the short name CTAN gives us, for example, cereda for me, gregorio for egreg and carlisle for David. You guys probably know what I'm talking about, right? :) Now let's try:

Quack 2

Then we click the OK button (don't mind the localization stuff, since my Linux is in Portuguese; you guys will probably get the window in your current locale). And wait. And wait a lot. Did I mention we have to wait? :)

Specially if you are David. Oh boy, he has a lot of packages! :)

If things don't get messed up, we will probably reach this final screen

Quack 3

and we will have a file named <author ID>.xml lying somewhere in our computer, hopefully in the same directory the application is. :) In my example, we have rouquette.xml:

<ctan>
  <packages>
    <package>
      <title>bibleref-french – French translations for bibleref</title>
      <description>The package provides translations and alternative typesetting conventions for use of bibleref in French.</description>
      <sources>
        <string>/macros/latex/contrib/bibleref-french</string>
      </sources>
      <documentation>
        <string>Package Manual</string>
        <string>Package Manual</string>
      </documentation>
      <versions>
        <string>2.3.1</string>
      </versions>
      <licenses>
        <string>The LaTeX Project Public License 1.3</string>
      </licenses>
      <copyright>
        <string>2012 Maïeul Rouquette and Raphaël Pinson</string>
      </copyright>
      <maintainers>
        <string>Maïeul Rouquette</string>
        <string>Raphaël Pinson</string>
      </maintainers>
      <distributions>
        <string>TeXLive as bibleref-french</string>
        <string>MiKTeX as bibleref-french</string>
      </distributions>
      <topics>
        <string>typeset devotional texts and other material of interest to theologians</string>
      </topics>
    </package>

    ...

  </packages>
</ctan>

I created a GitHub repository for ctaninfo, and the binary can be found here, under the releases link. Phew. :)

Up the wooden hill to Bedfordshire!

  • 1
    that is a great think. Thanks very much. But, as I could see in the code, you take the information from the .html webpage, which structure could change. So I think it don't solve the main problem. A second point is that, in my case, I would like to make request automatically with PHP... so your tools is useful for local use, but not if we want to make automatically request. – Maïeul Jan 7 '14 at 12:00
  • @Maïeul: Thanks! I wrote this humble code in 10 minutes, I guess. :) You are right, I retrieve data from the HTML page, parse and then serialize into an XML file, so if things change in the future, I might get into trouble. :) I think the best option is then kindly ask the CTAN gentlemen to include some sort of web service or even statically generated .json file with the author metadata. – Paulo Cereda Jan 7 '14 at 12:51
  • yes, that is why I was thinking. But I don't know if the CTAN gentlemen have time ... – Maïeul Jan 7 '14 at 13:00
10

there is a read-only copy of the catalogue source, omitting author details at rsync://rsync.tex.ac.uk for use with rsync. (the author details contain email addresses, so we decided simply to omit them. for the same reason, we can't simply publish our svn pointer -- that remains protected.)

the copy is updated daily.

(nb, i tried to use the facilities provided for urls in answers, but the formatting insisted on putting http: before the rsync: )

  • 1
    Yay, that's great, R... I mean, wasteofspace! :) – Paulo Cereda Jan 9 '14 at 15:12
  • it's great ... but there is still the problem of no information about author :-( – Maïeul Jan 9 '14 at 22:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.