Michiel Kamermans wrote: > like nearly every XML editor I've found, your product seems to miss an > option that is so obviously useful in XML editing that people tend to > forget it's something that speeds up the creation and updating process: > search and replace in a node path context. > > Not just "find text [ ....] and replace with [....]" but "find text [ > ....] and replace with [....] in [...]". > > As an example, I am currently faced with poorly designed XML describing > Japanese ideographs, and have a few thousand entries of the following > XML (with japanese content in ISO-2022-JP) format: > > <kanji value="83"> > <written>"?"</written> > > <...> > > <readings> > <onyomi> > <pronunciation>?</pronunciation> > <meaning>non-existance</meaning> > <example> > <japanese>???<ruby><rb>??</rb><rt>??< /rt></ruby> > ???</japanese> > <english>That is not possible.</english> > </example> > </onyomi> > > <kunyomi> > <pronunciation>?[?]</pronunciation> > <meaning>negation marker</meaning> > <example> > <japanese><ruby><rb>??</rb><rt>??< /rt></ruby>? > <ruby><rb>?</rb><rt>?< /rt></ruby>??<ruby><rb>?</rb><rt>< /rt>? > </ruby>???????</japanese> > <english>I told you there was no meaning to it.</english> > </example> > </kunyomi> > > <special> > <example> > <japanese><ruby><rb>???</rb><rt>????< > /rt></ruby></japanese> > <english>fig</english> > </example> > </special> > </readings> > > <...> > </kanji> > > This is poor XML for several reasons, but one of the is that the > examples inside <reading> tags are sentence examples while examples > inside <special> tags are single word examples. What I would expect is > to be able to quickly do a blind search and replace on a subset on these > tags. XSLT or Xtransform are out of the question because it'd take too > long to write out in comparison to the following rather simple solution: > > find [<example> ] > replace [<example type="sentence"> ] > in path [document/kanji/readings/*yomi] > > I hit "Replace All", the program races through the few thousand entries > and only updates <example> tags that conform to this path. The tags in > <special> are untouched and I can do a second search/replace for them by > updating them to [<example type="word">] now using the same principle, > replacing "*yomi" with "special" instead. The whole operation only takes > me a few seconds, and never shifts my focus away from editing the XML, > unlike programming XSLT or Xtransforms would. > > This is a very good thing, it allows me to stay on the job. > > It's odd that no XML editor seems to have this function (yet?); it's > intuitive, and bloody handy for updating large quantities of relatively > small errors like these, for which writing XML conversion scripts is > really an unnecessary task. Adding it into your product, especially > since it's JAVA based, would be reasonably easy to do (after all, the > SAX2 model essentially natively allows path checking) and would result > in an increase in functionality that would most likely be highly > appreciated by many XML designers and editors. > > I hope to find this feature in a future version of your XML editor.
We totally agree with everything you say. That's why this feature is planned for XXE V3 (next year). Thanks for suggesting a simple and powerful user interface for it. See also http://www.xmlmind.com/xmleditor/wish_list.html
