Download dataset: better filename

[aostrow]

problem: Filename is made of dataset title only solution: add id + version number at the beginning of filename

also the long titles are somehow bad to read - should they be snipped?

[geofranzi]

Downloads

• [ ] Titel yes/no, if yes how long?

• [ ] allow primary data download without metadata?

• [ ] unstructured/files: new: ID_VERSIONNUMBER_filename -> ok

• [ ] attachments: filename -> ?

• [ ] Download metadata HTML: ID_metadata.html ->? (+ VERSIONNUMBER, TITLE?)

• [ ] Download metadata XML: ID_VERSIONNUMBER_metadata_TITLE.xml ->?

• [ ] Download dataset unstructured: ID_VERSIONNUMEBR_dataset.zip -> add readme, title?

---

• [ ] Download xlsm/xlsx: TITLE.xlsm/xlsx -> ?
  • [ ] Dowload csv, tsv, txt : TITLE.csv / tsv / txt -> ?

---

• [ ] download data structure template: ID_DATASTRUCTURENAME.xlsm -> ?

[geofranzi]

Filenames in BEXIS2 (Protocol 31.7.2010 Roman&Franzi)

• current handling, see ticket in Github #445

• in general:

  • it would be good to download primary data always along with metadata/data structure
  • no spaces, only underscores are allowed

• if we want to have a title in the file name, it must be shortened (around 8 characters), if there is a (collection) year, it should be included potential approaches:

  • user/data manager identifies the most important term of the title, which will be included in the file name
    • use other MD elements, e.g. project name/ID, collection year,

Case Primary Data (unstructured)

• files uploaded to BEXIS2 are not changed
• filename should contain: application name (web.config), dataset ID (BEXIS2 dataset ID), and version ID (BEXIS2 version number)
• add a code to identify primary data (e.g. PD, pdat, prim,...)

Case Primary Data (structured)

• files uploaded to BEXIS2 are not changed
• filename should contain: application name (web.config), dataset ID (BEXIS2 dataset ID), and version ID (BEXIS2 version number)
• add a code to identify primary data (e.g. PD, pdat, prim,...)

Case Attachments

• files uploaded to BEXIS2 are not changed
• filename should contain: application name (web.config), dataset ID (BEXIS2 dataset ID), and version ID (BEXIS2 version number), if available
• add a code to identify attachments (e.g. att, atch, attm,...)

Case Metadata

• filename should contain: application name (web.config), dataset ID (BEXIS2 dataset ID), and version ID (BEXIS2 version number)
• add a code to identifiy attachments (e.g. metadata, meta, md,...) --> New feature: in addition to XML, html, metadata should be available for download as JSON, TXT (formated)

Case Data Structure

• filename should contain: application name (web.config), dataset ID (BEXIS2 dataset ID), and data structure ID
• add a code to identify attachments (e.g. data_structure, datastrct,...)

Case Dataset (Bundle) -filename should contain: application name (web.config), dataset ID (BEXIS2 dataset ID), and version ID (BEXIS2 version number)

• add date (e.g. download date, version date)

Special cases:

• what happened, if data has a DOI already?

examples:

ATTO_54_v301_data.csv
BExIS_987_v12_metadata.xml AquaDiva_27_datastructure_34.xlsm iDiv_3_v5_20200731.zip

Todo

• [ ] decide on code to identify xyz
• [ ] implementation Primary Data (unstructured)
• [ ] implementation Primary Data (structured)
• [ ] implementation Case Attachments
• [ ] implementation Metadata
• [ ] implementation Data structure
• [ ] implementation Dataset

[aostrow]

We got also requests from users that the dataset title should be in the filename. It is easier for them to overlook the downloaded data and to know wich files contains which data (e.g. because soil is in the name). In B1 we had "ID_Title_version":

27846_Cover of all vascular plant species in representative 2x2 quadrats of the major surrounding homogeneous vegetation zones in a 75-m radius of the 150 grassland EPs, 2017-2018_1.1.5.zip

I like your suggestion above but would like to see the title included.