RE: [Htmlparser-developer] changes.txt and changes.html

[Marc N. <ma...@ke...>]

Works for me.  I usually just copy my commit message into the text file =
anyways!  :)

-----Original Message-----
From: Derrick Oswald [mailto:Der...@ro...]
Sent: Wednesday, May 07, 2003 6:28 PM
To: htm...@li...
Subject: Re: [Htmlparser-developer] changes.txt and changes.html



The out of sync part is my fault.  I didn't realize it was a manual=20
process to update changes.html too.
I agree, it should be dropped.

Maybe we shouldn't be manually updating changes.txt either.  From=20
"Introduction to SourceForge.net Project CVS Services for Developers"=20
(https://sourceforge.net/docman/display_doc.php?docid=3D768&group_id=3D1)=


    Commit Messages and Automatic ChangeLog Generation
    If your project team has been careful to provide meaningful commit
    messages when they have committed changes to your repository, you
    will have the option of using an automated tool to generate
    ChangeLog files for inclusion on your web site or in your file
    releases. The
    =
<http://cvsbook.red-bean.com/cvsbook.html#cvs2cl_--_Generate_GNU-Style_Ch=
angeLogs>cvs2cl
    =
(http://cvsbook.red-bean.com/cvsbook.html#cvs2cl_--_Generate_GNU-Style_Ch=
angeLogs)
    script may be used from your workstation in order to generate a
    ChangeLog file based on the commit messages for your repository.

I've tried this for the htmlparser project, getting the changes commited =

this week so far
(./cvs2cl.pl -l "-d'>May 4, 2003'") and it works pretty well:


2003-05-07 18:04  derrickoswald
=20
        * docs/: changes.txt, changes.html: update changelog
=20
2003-05-07 18:00  derrickoswald
=20
        * src/org/htmlparser/: tests/utilTests/NodeListTest.java,
        util/NodeList.java: added removeAll() to NodeList - Dhaval
=20
2003-05-07 16:16  polarys
=20
        * docs/changes.html, docs/changes.txt,
        src/org/htmlparser/scanners/ScriptScanner.java,
        src/org/htmlparser/tests/scannersTests/ScriptScannerTest.java:
        Fixed NPE in ScriptScanner when a script tag was not ended =
before
        the end of document
=20
2003-05-06 07:44  derrickoswald
=20
        * docs/changes.txt: update changelog
=20
2003-05-06 07:37  derrickoswald
=20
        * src/org/htmlparser/parserHelper/ParserHelper.java: Fix #732517
        Paser(String) c'tor not handling relative path local file
=20
2003-05-05 20:45  derrickoswald
=20
        * docs/changes.txt: update changelog
=20
2003-05-05 20:43  derrickoswald
=20
        * src/org/htmlparser/: scanners/LabelScanner.java,
        scanners/SelectTagScanner.java, tags/SelectTag.java: nodelist
        modifications from Dhaval
=20
2003-05-04 23:12  derrickoswald
=20
        * src/org/htmlparser/: Node.java, NodeReader.java, Parser.java,
        RemarkNode.java, RemarkNodeParser.java, StringNode.java,
        parserHelper/AttributeParser.java,
        parserHelper/CompositeTagScannerHelper.java,
        parserHelper/StringParser.java, parserHelper/TagParser.java,
        parserapplications/LinkExtractor.java,
<snip - it's every file in the source tree>
        visitors/StringFindingVisitor.java,
        visitors/TagFindingVisitor.java,
        visitors/TextExtractingVisitor.java,
        visitors/UrlModifyingVisitor.java: update version headers to
        1.3-20030504

If this kind of thing is acceptable, then all we need to do is convince=20
all developers that the commit messages are really important and need to =

be carefully constructed=20
(http://www.red-bean.com/cvs2cl/changelogs.html), since they will then=20
be automatically added to the changelog when a release happens.  I'm=20
willing to add that task to the list of things done at release time.

The only other thing the changes.txt is probably used for is=20
notification of commits, and that should be handled by the new=20
htmlparser-cvs mailing list.  Generating the changelog from commit=20
messages would get rid of all the 'update changelog' emails too.

There is also a cvs2html script (http://cvs.sslug.dk/cvs2html/) that=20
provides not only the changelog entries but also provides a html diff =
view.

Derrick

Marc Novakowski wrote:

>I just checked in a fix for a NPE I was getting in ScriptScanner when I =
parsed a document that had a <script> tag that wasn't closed before the =
end of document.  As usual, I also updated the changes.txt and =
changes.html files.  I noticed these two files are a little out of sync =
with each other, which begs the question of why do we have two files =
that have (or are supposed to have) the exact same content, only with =
different file extensions?
>
>It seems silly that commiters have to edit both of these files.  Maybe =
we should get rid of "changes.html" since it's just a text file anyways =
(no markup).
>
>On another note, there are currently 6 tests failing (5 of which =
weren't failing as of a few weeks ago).  Anyone who's checked in changes =
recently might want to take a look at the failures.
>
>Marc
>
>
> =20
>




-------------------------------------------------------
Enterprise Linux Forum Conference & Expo, June 4-6, 2003, Santa Clara
The only event dedicated to issues related to Linux enterprise solutions
www.enterpriselinuxforum.com

_______________________________________________
Htmlparser-developer mailing list
Htm...@li...
https://lists.sourceforge.net/lists/listinfo/htmlparser-developer