#105 extended import - inspired by feature request [2163626]


Hello developers,

our patch extends the functionality of "Import into new database" and "Import into current database" in that way, that the user could now choose multiple files and although directories. All directories will be recursivly searched for files.

All files will be grouped by their import format corresponding to the import file filter. Currently their will be only one file format retrieved by the choosen file type, but our code is also usable with multiple file formats.

All groups of files will be imported together by the use of the existing method "automatedImport", which needs an Array of Strings representing the filenames. Before the patch, there was only one filename within the Array, but this was easy to change.

The user gets then the normal dialog with all BibTeX-entries found within the files, where he could also choose the "deselect all duplicates"-option. The results of all files of one group are summarized within one dialog.

We was inspired for this patch by the feature request 2163626 [0], but we have only extended the existing import possibilities, so that the user could do something like that by the use of this option.
One point, where we have different opinions is, if it is requested that also PDF-files without any meta data should also be imported (e.g. as entry of type "Other" with only the file link). We did not change the behavior of the handling of those files, but if it is requested, it could be easyly solved by an new import format for that type of files.

Best regards, pjhimself

[0] http://sourceforge.net/tracker/index.php?func=detail&aid=2163626&group_id=92314&atid=600309


  • Hi,

    thank you for this patch! What do you mean by "all files will be grouped by their import format corresponding to the
    import file filter"?

  • PJhimself

    Hi mortenalver,

    as we wrote this patch, we want to group the imported files by their import format, so that the user could get a summary for each import format. But after some test, I noticed that their would only be one import format in every case (depends on the file filter) and that the further differentiation would be performed by trial-and-error method at the importing itself. Therefore, this could currently only be a later possibility, if this behavior would be changed.
    Because of that, I changed the patch a little bit, so that now all files (filenames) are only collected and then send to the automated import.
    The used files depends on the file filter in every case.

    You could delete the old patch whenever you want.

    I hope, I could clear all of your open questions.

    Best regards, pjhimself
    File Added: jabref_patch_ImportFormats.diff

  • PJhimself

    patch (2nd edition)