> Since there is a "maybe need" wouldn't it be a good idea (or > maybe someone > allready does ?) to create an xml version of the map feature ?
That is sort of what the perl script does, or tries to, I think. Once it has the XML file the program can see whether a given feature is in that file or not. It can then do further validation on the values (as well as the keys). Here is a bit out of the middle of the generated not-in-map_features.xml created by not-in-map_features.pl (which if you run the [EMAIL PROTECTED] client is in the subfolder maplint\tests\strict) <xsl:when test="@k='boundary'"> <xsl:choose> <xsl:when test="@v='administrative'" /> <xsl:when test="@v='civil'" /> <xsl:when test="@v='national_park'" /> <xsl:when test="@v='political'" /> <xsl:otherwise> <maplint:result> <xsl:value-of select="concat('Value not in map features: ', @k, '=', @v)" /> </maplint:result> </xsl:otherwise> </xsl:choose> </xsl:when> I think this then gets built into the tests.xml or tests.xsl files in the maplint folder but I've not looked into it in quite that detail to see how and when. There are both advantages and disadvantages to maintaining an XML version. In my opinion the advantages of parsing the Map Features page include that I suspect it is easier for many users to add new features to a wiki page than in an XML file. The disadvantages I see are that users won't know that the page is parsed to extract the information, and so might not know the rules the parsing uses (one example might be that "User defined" allows any value in the Maplint validation, but a user might just put a * in the value column to allow anything. It might be, and I haven't checked, that the maplint script will then check for a value of '*' as valid, and otherwise report not in Map Features). I'm not saying either that the not-in-map_features.pl script can't be amended, but the structure of Map Features was standardised (as I understand it) to allow the automatic parsing (such as a set number of columns in a table row, the correct tags for node/way/area in the relevant column to know which tags are valid on which items, etc). If a new column were to be added to one of the sections though it would need adding to all of them and the line next unless $#columns == 5; in sub parse_featuretemplate would need updating in the Perl script. Ed _______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk